code.vuplus.com Git - vuplus_webkit/blob - Tools/ChangeLog

   1 2011-09-15  Anders Carlsson  <andersca@apple.com>
   2
   3         Fix Mac build by including config.h where needed.
   4
   5         * TestWebKitAPI/Tests/TestWebKitAPI/mac/InstanceMethodSwizzler.mm:
   6         * TestWebKitAPI/Tests/WTF/MetaAllocator.cpp:
   7         * TestWebKitAPI/Tests/WTF/RedBlackTree.cpp:
   8         * TestWebKitAPI/Tests/WebKit2/WebArchive.cpp:
   9         * TestWebKitAPI/Tests/WebKit2/WebArchive_Bundle.cpp:
  10         * TestWebKitAPI/Tests/mac/DOMRangeOfString.mm:
  11         * TestWebKitAPI/Tests/mac/DeviceScaleFactorOnBack.mm:
  12         * TestWebKitAPI/Tests/mac/DynamicDeviceScaleFactor.mm:
  13         * TestWebKitAPI/Tests/mac/StringByEvaluatingJavaScriptFromString.mm:
  14         * TestWebKitAPI/mac/InstanceMethodSwizzler.mm:
  15         * TestWebKitAPI/mac/JavaScriptTestMac.mm:
  16         * TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
  17         * TestWebKitAPI/mac/PlatformWebViewMac.mm:
  18         * TestWebKitAPI/mac/SyntheticBackingScaleFactorWindow.m:
  19         * TestWebKitAPI/mac/WebKitAgnosticTest.mm:
  20         * TestWebKitAPI/mac/main.mm:
  21
  22 2011-09-15  Xianzhu Wang  <wangxianzhu@chromium.org>
  23
  24         Let TestWebKitAPI work for chromium
  25         https://bugs.webkit.org/show_bug.cgi?id=67756
  26
  27         Reviewed by Sam Weinig.
  28
  29         * Scripts/build-api-tests:
  30         * Scripts/run-api-tests:
  31         (isSupportedPlatform):
  32         (runTest):
  33         (populateTests):
  34         * Scripts/webkitdirs.pm:
  35         (determineBaseProductDir):
  36         * TestWebKitAPI/Configurations/InjectedBundle.xcconfig:
  37         * TestWebKitAPI/Configurations/TestWebKitAPI.xcconfig:
  38         * TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
  39         * TestWebKitAPI/InjectedBundleController.cpp:
  40         * TestWebKitAPI/InjectedBundleMain.cpp:
  41         * TestWebKitAPI/JavaScriptTest.cpp:
  42         * TestWebKitAPI/PlatformUtilities.cpp:
  43         * TestWebKitAPI/TestWebKitAPI.gypi: Added.
  44         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
  45         * TestWebKitAPI/TestWebKitAPIPrefix.h: Removed.
  46         * TestWebKitAPI/Tests/WTF/CheckedArithmeticOperations.cpp:
  47         * TestWebKitAPI/Tests/WTF/StringOperators.cpp:
  48         * TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
  49         * TestWebKitAPI/Tests/WTF/VectorReverse.cpp:
  50         * TestWebKitAPI/Tests/WebKit2/AboutBlankLoad.cpp:
  51         * TestWebKitAPI/Tests/WebKit2/CanHandleRequest.cpp:
  52         * TestWebKitAPI/Tests/WebKit2/CanHandleRequest_Bundle.cpp:
  53         * TestWebKitAPI/Tests/WebKit2/CookieManager.cpp:
  54         * TestWebKitAPI/Tests/WebKit2/DocumentStartUserScriptAlertCrash.cpp:
  55         * TestWebKitAPI/Tests/WebKit2/DocumentStartUserScriptAlertCrash_Bundle.cpp:
  56         * TestWebKitAPI/Tests/WebKit2/DownloadDecideDestinationCrash.cpp:
  57         * TestWebKitAPI/Tests/WebKit2/EvaluateJavaScript.cpp:
  58         * TestWebKitAPI/Tests/WebKit2/FailedLoad.cpp:
  59         * TestWebKitAPI/Tests/WebKit2/Find.cpp:
  60         * TestWebKitAPI/Tests/WebKit2/ForceRepaint.cpp:
  61         * TestWebKitAPI/Tests/WebKit2/FrameMIMETypeHTML.cpp:
  62         * TestWebKitAPI/Tests/WebKit2/FrameMIMETypePNG.cpp:
  63         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle.cpp:
  64         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle_Bundle.cpp:
  65         * TestWebKitAPI/Tests/WebKit2/InjectedBundleBasic.cpp:
  66         * TestWebKitAPI/Tests/WebKit2/InjectedBundleBasic_Bundle.cpp:
  67         * TestWebKitAPI/Tests/WebKit2/LoadCanceledNoServerRedirectCallback.cpp:
  68         * TestWebKitAPI/Tests/WebKit2/LoadCanceledNoServerRedirectCallback_Bundle.cpp:
  69         * TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash.cpp:
  70         * TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash_Bundle.cpp:
  71         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
  72         * TestWebKitAPI/Tests/WebKit2/PageLoadDidChangeLocationWithinPageForFrame.cpp:
  73         * TestWebKitAPI/Tests/WebKit2/PreventEmptyUserAgent.cpp:
  74         * TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly.cpp:
  75         * TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly_Bundle.cpp:
  76         * TestWebKitAPI/Tests/WebKit2/RestoreSessionStateContainingFormData.cpp:
  77         * TestWebKitAPI/Tests/WebKit2/SpacebarScrolling.cpp:
  78         * TestWebKitAPI/Tests/WebKit2/WKPreferences.cpp:
  79         * TestWebKitAPI/Tests/WebKit2/WKString.cpp:
  80         * TestWebKitAPI/Tests/WebKit2/WKStringJSString.cpp:
  81         * TestWebKitAPI/Tests/WebKit2/win/AltKeyGeneratesWMSysCommand.cpp:
  82         * TestWebKitAPI/Tests/WebKit2/win/DoNotCopyANullCFURLResponse.cpp:
  83         * TestWebKitAPI/Tests/WebKit2/win/HideFindIndicator.cpp:
  84         * TestWebKitAPI/Tests/WebKit2/win/ResizeViewWhileHidden.cpp:
  85         * TestWebKitAPI/Tests/WebKit2/win/TranslateMessageGeneratesWMChar.cpp:
  86         * TestWebKitAPI/Tests/WebKit2/win/WMCloseCallsUIClientClose.cpp:
  87         * TestWebKitAPI/Tests/WebKit2/win/WMPrint.cpp:
  88         * TestWebKitAPI/TestsController.cpp:
  89         * TestWebKitAPI/config.h: Copied from TestWebKitAPI/TestWebKitAPIPrefix.h.
  90         * TestWebKitAPI/win/InjectedBundleControllerWin.cpp:
  91         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp:
  92         * TestWebKitAPI/win/PlatformWebViewWin.cpp:
  93         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
  94         * TestWebKitAPI/win/main.cpp:
  95
  96 2011-09-15  Jochen Eisinger  <jochen@chromium.org>
  97
  98         [chromium] extract normalization of LayoutTests URLs to a method of TestShell
  99         https://bugs.webkit.org/show_bug.cgi?id=68145
 100
 101         This will allow for other classes in DumpRenderTree to reuse this functionality
 102
 103         Reviewed by David Levin.
 104
 105         * DumpRenderTree/chromium/TestShell.cpp:
 106         (normalizeLayoutTestURLInternal):
 107         (dumpHistoryItem):
 108         (TestShell::normalizeLayoutTestURL):
 109         * DumpRenderTree/chromium/TestShell.h:
 110
 111 2011-09-14  Ada Chan  <adachan@apple.com>
 112
 113         Add test for WKBundleFrameCopyWebArchive().
 114         https://bugs.webkit.org/show_bug.cgi?id=67857
 115
 116         Reviewed by Anders Carlsson.
 117
 118         * TestWebKitAPI/Tests/WebKit2/WebArchive.cpp: Added.
 119         (TestWebKitAPI::didReceiveMessageFromInjectedBundle): Do some basic sanity checks to make sure
 120         the returned data is a valid webarchive.
 121         (TestWebKitAPI::setInjectedBundleClient):
 122         (TestWebKitAPI::didFinishLoadForFrame):
 123         (TestWebKitAPI::TEST): Load simple.html in a WebView.  Wait for it to finish loading, and then ask the
 124         injected bundle for the webarchive and check the returned webarchive.
 125         * TestWebKitAPI/Tests/WebKit2/WebArchive_Bundle.cpp: Added.
 126         (TestWebKitAPI::WebArchiveTest::WebArchiveTest):
 127         (TestWebKitAPI::WebArchiveTest::didReceiveMessage): Handle the "GetWebArchive" message and return
 128         the web archive.
 129
 130         Add files to project.
 131         * TestWebKitAPI/win/TestWebKitAPI.vcproj: Also let Visual Studio reorder a file.
 132         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
 133         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
 134
 135 2011-09-14  Ryosuke Niwa  <rniwa@webkit.org>
 136
 137         committers.py should support multiple IRC nicknames
 138         https://bugs.webkit.org/show_bug.cgi?id=68110
 139
 140         Reviewed by Eric Seidel.
 141
 142         Add multiple IRC nickname support to committers.py
 143
 144         * Scripts/webkitpy/common/config/committers.py:
 145         * Scripts/webkitpy/common/config/committers_unittest.py:
 146         * Scripts/webkitpy/tool/bot/irc_command.py:
 147         * Scripts/webkitpy/tool/bot/sheriff.py:
 148
 149 2011-09-14  Jon Lee  <jonlee@apple.com>
 150
 151         Implement globalFlag for WebKitTestRunner
 152         https://bugs.webkit.org/show_bug.cgi?id=68033
 153         <rdar://problem/10119204>
 154
 155         Reviewed by Sam Weinig.
 156
 157         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl: Porting globalFlag from DRT
 158         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
 159         (WTR::LayoutTestController::LayoutTestController): initialize m_globalFlag to false
 160         * WebKitTestRunner/InjectedBundle/LayoutTestController.h: add m_globalFlag
 161         (WTR::LayoutTestController::globalFlag): simple getter
 162         (WTR::LayoutTestController::setGlobalFlag): simple setter
 163
 164 2011-09-13  Eric Seidel  <eric@webkit.org>
 165
 166         Remove ENABLE_SVG_USE as <use> is required by HTML5
 167         https://bugs.webkit.org/show_bug.cgi?id=68019
 168
 169         Reviewed by Ryosuke Niwa.
 170
 171         * Scripts/build-webkit:
 172         * waf/build/settings.py:
 173
 174 2011-09-14  Ryosuke Niwa  <rniwa@webkit.org>
 175
 176         Remove Jan Erik Hanssen (added in my r95078) from the list for now
 177         since I can't confirm acceptable his committership.
 178
 179         * Scripts/webkitpy/common/config/committers.py:
 180
 181 2011-09-14  Jochen Eisinger  <jochen@chromium.org>
 182
 183         [chromium] Remove obsolete WebViewClient methods from DumpRenderTree.
 184         https://bugs.webkit.org/show_bug.cgi?id=68066
 185
 186         Reviewed by Adam Barth.
 187
 188         * DumpRenderTree/chromium/WebViewHost.cpp:
 189         * DumpRenderTree/chromium/WebViewHost.h:
 190
 191 2011-09-13  Ryosuke Niwa  <rniwa@webkit.org>
 192
 193         committers.py is missing some contributors
 194         https://bugs.webkit.org/show_bug.cgi?id=68059
 195
 196         Reviewed by Adam Barth.
 197
 198         Add David Barr, Jan Erik Hanssen, and Kim Grönholm.
 199
 200         Also add nicks of Alejandro G. Castro, Collin Jackson, and Matt Lilek.
 201
 202         * Scripts/webkitpy/common/config/committers.py:
 203
 204 2011-09-13  Ryosuke Niwa  <rniwa@webkit.org>
 205
 206         Fix webkit python test after r95061.
 207
 208         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
 209
 210 2011-09-13  Xianzhu Wang  <wangxianzhu@chromium.org>
 211
 212         Add myself as a contributor. No review needed.
 213
 214         * Scripts/webkitpy/common/config/committers.py:
 215
 216 2011-09-13  Adam Barth  <abarth@webkit.org>
 217
 218         garden-o-matic has unfriendly error messages when the local server is not available
 219         https://bugs.webkit.org/show_bug.cgi?id=68042
 220
 221         Reviewed by Dimitri Glazkov.
 222
 223         This patch moves us from using alert to showing the error inline in the
 224         progress dialog.
 225
 226         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout.js:
 227         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout_unittests.js:
 228         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
 229         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
 230         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
 231
 232 2011-09-13  Adam Barth  <abarth@webkit.org>
 233
 234         FailureGrid in garden-o-matic should link to the builder's waterfall page
 235         https://bugs.webkit.org/show_bug.cgi?id=68036
 236
 237         Reviewed by Dimitri Glazkov.
 238
 239         One of the users in the user study thought the configurations names in
 240         the failure grid would link to the underlying bots.  That makes sense,
 241         so this patch adds the behavior.
 242
 243         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures.js:
 244         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js:
 245         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
 246         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css:
 247
 248 2011-09-13  Adam Barth  <abarth@webkit.org>
 249
 250         garden-o-matic should be able to mark failures as expected
 251         https://bugs.webkit.org/show_bug.cgi?id=68027
 252
 253         Reviewed by Dimitri Glazkov.
 254
 255         All the back-end infrastructure exists for this operation.  This patch
 256         just adds some UI to the front-end and wires that UI up to the back
 257         end.  This exact UI will likely need some more polish, but this should
 258         be a reasonable first cut.
 259
 260         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
 261         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions.js:
 262         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
 263         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
 264         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
 265
 266 2011-09-13  Adam Barth  <abarth@webkit.org>
 267
 268         garden-o-matic should tell you about the progress of your rebaseline
 269         https://bugs.webkit.org/show_bug.cgi?id=67967
 270
 271         Reviewed by Dimitri Glazkov.
 272
 273         This patch adds ui.MessageBox, which is a light wrapper around jQuery
 274         UI's dialog widget.  We then use ui.MessageBox to show status messages
 275         about in-progress rebaselines.  As each baseline is downloaded, the
 276         MessageBox updates.  This UI replaces our temporary "alert"-based UI.
 277
 278         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout.js:
 279         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
 280         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
 281         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions.js:
 282         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
 283         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
 284
 285 2011-09-13  Leandro Pereira  <leandro@profusion.mobi>
 286
 287         GTK DumpRenderTree uses inefficient idioms to iterate over G[S]Lists
 288         https://bugs.webkit.org/show_bug.cgi?id=68024
 289
 290         Reviewed by Gustavo Noronha Silva.
 291
 292         Using g_list_count() and g_list_nth_data() together on a loop is
 293         inneficient since they're both O(n). Iterate over lists in a saner
 294         way.
 295
 296         * DumpRenderTree/gtk/DumpRenderTree.cpp:
 297         (dumpHistoryItem): Reduce the scope for the 'kids' variable, and
 298         iterate on it using g_list_next(). Free the list after done with it.
 299         (dumpBackForwardListForWebView): Instead of appending (which is
 300         expensive in GLists) history items and then iterating from the tail
 301         of the itemsToPrint list, prepend items and walk forwards as usual.
 302         (dumpBackForwardListForAllWebViews): Walk the list in a saner way,
 303         remove the (unneeded) viewList variable.
 304
 305 2011-09-13  Ryosuke Niwa  <rniwa@webkit.org>
 306
 307         Add Eric's IRC nick.
 308
 309         * Scripts/webkitpy/common/config/committers.py:
 310
 311 2011-09-13  Anders Carlsson  <andersca@apple.com>
 312
 313         Disable C++ exceptions when building with clang
 314         https://bugs.webkit.org/show_bug.cgi?id=68031
 315         <rdar://problem/9556880>
 316
 317         Reviewed by Mark Rowe.
 318
 319         * MiniBrowser/Configurations/Base.xcconfig:
 320         * TestWebKitAPI/Configurations/Base.xcconfig:
 321         * WebKitTestRunner/Configurations/Base.xcconfig:
 322
 323 2011-09-13  Ojan Vafai  <ojan@chromium.org>
 324
 325         [chromium] fix flakiness dashboard for chromium interactive tests
 326         https://bugs.webkit.org/show_bug.cgi?id=68029
 327
 328         Reviewed by Tony Chang.
 329
 330         * TestResultServer/static-dashboards/flakiness_dashboard.html:
 331         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
 332
 333 2011-09-13  Ryosuke Niwa  <rniwa@webkit.org>
 334
 335         Add Alice and Aryeh to the list of contributors.
 336
 337         * Scripts/webkitpy/common/config/committers.py:
 338
 339 2011-09-13  Ryosuke Niwa  <rniwa@webkit.org>
 340
 341         sheriffbot whois should also tell us email addresses
 342         https://bugs.webkit.org/show_bug.cgi?id=67975
 343
 344         Reviewed by Eric Seidel and David Levin.
 345
 346         Taught sheriffbot how to tell us contributor's email addresses.
 347
 348         * Scripts/webkitpy/tool/bot/irc_command.py:
 349         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
 350
 351 2011-09-13  Eric Seidel  <eric@webkit.org>
 352
 353         Remove ENABLE_SVG_FOREIGN_OBJECT as it is a required part of HTML5
 354         https://bugs.webkit.org/show_bug.cgi?id=68018
 355
 356         Reviewed by Ryosuke Niwa.
 357
 358         * Scripts/build-webkit:
 359         * waf/build/settings.py:
 360
 361 2011-09-13  Chang Shu  <cshu@webkit.org>
 362
 363         [WK2] [Mac] Implement KeyDown function for WebKit2 EventSender.
 364         https://bugs.webkit.org/show_bug.cgi?id=57515
 365
 366         The code change in WebKitTestRunner first sends keydown event from InjectedBundle to
 367         TestController synchronously. The latter then sends simulated event to WebKit2 module,
 368         which is synchronous, too.
 369
 370         Reviewed by Darin Adler.
 371
 372         * WebKitTestRunner/EventSenderProxy.h: Added.
 373         (WTR::EventSenderProxy::EventSenderProxy):
 374         * WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
 375         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
 376         (WTR::EventSendingController::keyDown):
 377         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
 378         * WebKitTestRunner/PlatformWebView.h:
 379         (WTR::PlatformWebView::platformWindow):
 380         * WebKitTestRunner/TestController.cpp:
 381         (WTR::TestController::TestController):
 382         (WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
 383         * WebKitTestRunner/TestController.h:
 384         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
 385         * WebKitTestRunner/mac/EventSenderProxy.mm: Added.
 386         (WTR::buildModifierFlags):
 387         (WTR::EventSenderProxy::keyDown):
 388
 389 2011-09-12  Leandro Pereira  <leandro@profusion.mobi>
 390
 391         [EFL] DRT: Add DumpRenderTreeEfl.h
 392         https://bugs.webkit.org/show_bug.cgi?id=63993
 393
 394         Reviewed by Eric Seidel.
 395
 396         This header contains some global variables used by EFL's
 397         DumpRenderTree implementation.
 398
 399         * DumpRenderTree/DumpRenderTree.h: Include DumpRenderTreeEfl.h when
 400         appropriate.
 401         * DumpRenderTree/efl/DumpRenderTreeEfl.h: Added.
 402
 403 2011-09-12  Raphael Kubo da Costa  <kubo@profusion.mobi>
 404
 405         [EFL] Send the right key names for PageUp and PageDown.
 406         https://bugs.webkit.org/show_bug.cgi?id=67044
 407
 408         Reviewed by Kenneth Rohde Christiansen.
 409
 410         PlatformKeyboardEventEfl uses "Prior" and "Next" to refer to PageUp and
 411         PageDown (which is close to the defines in WindowsKeyboardCodes.h).
 412         However, EventSender was sending "Page_Up" and "Page_Down", so tests
 413         such as editing/input/page-up-down-scrolls.html were timing out due to
 414         the right key down events not being sent.
 415
 416         * DumpRenderTree/efl/EventSender.cpp:
 417         (keyPadNameFromJSValue):
 418         (keyNameFromJSValue):
 419
 420 2011-09-12  Eric Seidel  <eric@webkit.org>
 421
 422         sheriffbot whois is case sensitive
 423         https://bugs.webkit.org/show_bug.cgi?id=67972
 424
 425         Reviewed by Adam Barth.
 426
 427         I also fixed a typo in the "More than 5 contributors" message
 428         from my previous commit.
 429
 430         * Scripts/webkitpy/common/config/committers.py:
 431         * Scripts/webkitpy/common/config/committers_unittest.py:
 432         * Scripts/webkitpy/tool/bot/irc_command.py:
 433         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
 434
 435 2011-09-12  Eric Seidel  <eric@webkit.org>
 436
 437         sheriffbot whois should be smarter
 438         https://bugs.webkit.org/show_bug.cgi?id=67971
 439
 440         Reviewed by Adam Barth.
 441
 442         This makes sheriffbot whois basically grep the committers list
 443         making it much easier to find a contributor by name.
 444
 445         Unfortunately the current unittests depend on the real committers.py
 446         file, so as we add more eric's to the project, this unittest result will
 447         change.  I think that's OK for now.  We'll fix it when it breaks.
 448
 449         * Scripts/webkitpy/common/config/committers.py:
 450         * Scripts/webkitpy/tool/bot/irc_command.py:
 451         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
 452
 453 2011-09-12  Sheriff Bot  <webkit.review.bot@gmail.com>
 454
 455         Unreviewed, rolling out r94976.
 456         http://trac.webkit.org/changeset/94976
 457         https://bugs.webkit.org/show_bug.cgi?id=67953
 458
 459         Broke a bunch of security/load& tests (Requested by rniwa on
 460         #webkit).
 461
 462         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
 463         (LayoutTestController::pathToLocalResource):
 464
 465 2011-09-12  Eric Seidel  <eric@webkit.org>
 466
 467         Rubber-stamped by Adam Barth.
 468
 469         Remove the "the tree is burning!" warning previously shown
 470         when landing patches.  It was a dream.  But it's just noise now.
 471
 472         Some bots will require a restart to notice the removal of the option
 473         so I've left the --ignore-builders option in for now.
 474
 475         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
 476         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
 477         * Scripts/webkitpy/tool/commands/download.py:
 478         * Scripts/webkitpy/tool/commands/download_unittest.py:
 479         * Scripts/webkitpy/tool/commands/queues_unittest.py:
 480         * Scripts/webkitpy/tool/steps/__init__.py:
 481         * Scripts/webkitpy/tool/steps/commit.py:
 482         * Scripts/webkitpy/tool/steps/ensurebuildersaregreen.py: Removed.
 483         * Scripts/webkitpy/tool/steps/postdiffforrevert.py:
 484
 485 2011-09-12  Eric Seidel  <eric@webkit.org>
 486
 487         [NRWT] REGRESSION: Local loader tests are failing on machines that lost /tmp/LayoutTests symlink
 488         https://bugs.webkit.org/show_bug.cgi?id=65781
 489
 490         Reviewed by Ryosuke Niwa.
 491
 492         Instead of making NRWT create the symlink, I just made DumpRenderTree smart enough
 493         to resolve the passed in url relative to the absolute url for the test.
 494
 495         I believe this is a better approach than the on used in the Qt and Chromium DRT's
 496         (which resolves the path relative to the built location of the DRT executable)
 497         and we should move this new code into a shared location in a follow-up patch.
 498
 499         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
 500         (LayoutTestController::pathToLocalResource):
 501
 502 2011-09-12  Eric Seidel  <eric@webkit.org>
 503
 504         Reshuffle some code in WebKitDriver._read_block in preparation for reading stderr/stdout separately
 505         https://bugs.webkit.org/show_bug.cgi?id=67530
 506
 507         Reviewed by Adam Barth.
 508
 509         I believe I've fixed the bug in the original patch which prompted the rollout.
 510         The previous patch was using the wrong deadline for the initial read,
 511         subtracting time.time() twice from the deadline value.
 512
 513         * Scripts/webkitpy/layout_tests/port/webkit.py:
 514         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
 515
 516 2011-09-12  Balazs Kelemen  <kbalazs@webkit.org>
 517
 518         [Qt][WK2] WebKitTestRunner does not produce crash logs
 519         https://bugs.webkit.org/show_bug.cgi?id=67714
 520
 521         Reviewed by Chang Shu.
 522
 523         Added a simple way of generating backtrace on crash
 524         to the web process. The implementation is similar what
 525         we have in DRT. It depends on GNU libc functionality
 526         so it is only enabled where we are running in such an environment.
 527
 528         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
 529         (WTR::printBacktrace):
 530         (WTR::crashHandler):
 531         (WTR::InjectedBundle::platformInitialize):
 532
 533 2011-09-11  Filip Pizlo  <fpizlo@apple.com>
 534
 535         Added my IRC nick to the committers file.
 536
 537         * Scripts/webkitpy/common/config/committers.py:
 538
 539 2011-09-01  Filip Pizlo  <fpizlo@apple.com>
 540
 541         The executable allocator makes it difficult to free individual
 542         chunks of executable memory
 543         https://bugs.webkit.org/show_bug.cgi?id=66363
 544
 545         Reviewed by Oliver Hunt.
 546
 547         Introduced a best-fit, balanced-tree based allocator. The allocator
 548         required a balanced tree that does not allocate memory and that
 549         permits the removal of individual nodes directly (as opposed to by
 550         key); neither AVLTree nor WebCore's PODRedBlackTree supported this.
 551         Changed all references to executable code to use a reference counted
 552         handle.
 553
 554         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
 555         * TestWebKitAPI/Tests/WTF/MetaAllocator.cpp: Added.
 556         (TestWebKitAPI::TEST_F):
 557         * TestWebKitAPI/Tests/WTF/RedBlackTree.cpp: Added.
 558         (TestWebKitAPI::Pair::findExact):
 559         (TestWebKitAPI::Pair::remove):
 560         (TestWebKitAPI::Pair::findLeastGreaterThanOrEqual):
 561         (TestWebKitAPI::Pair::assertFoundAndRemove):
 562         (TestWebKitAPI::Pair::assertEqual):
 563         (TestWebKitAPI::Pair::assertSameValuesForKey):
 564         (TestWebKitAPI::Pair::testDriver):
 565         (TestWebKitAPI::TEST_F):
 566
 567 2011-09-10  Andy Estes  <aestes@apple.com>
 568
 569         Move myself from committers_unable_to_review to reviewers_list.
 570
 571         * Scripts/webkitpy/common/config/committers.py:
 572
 573 2011-09-10  Kevin Ollivier  <kevino@theolliviers.com>
 574
 575         [wx] Unreviewed build fix. MSW build fixes.
 576
 577         * DumpRenderTree/config.h:
 578         * waf/build/wxpresets.py:
 579
 580 2011-09-10  Andreas Kling  <kling@webkit.org>
 581
 582         Crash beneath WKURLIsEqual() when reloading in MiniBrowser.
 583         https://bugs.webkit.org/show_bug.cgi?id=59328
 584
 585         Reviewed by Kenneth Rohde Christiansen.
 586
 587         * MiniBrowser/mac/BrowserWindowController.m:
 588         (-[BrowserWindowController updateProvisionalURLForFrame:]): Add missing null check.
 589
 590 2011-09-10  Adam Barth  <abarth@webkit.org>
 591
 592         Build fix for Android.  DumpRenderTree is a client of the WebKit API
 593         and does not have access to notImplemented().
 594
 595         * DumpRenderTree/chromium/TestShellAndroid.cpp:
 596         (openStartupDialog):
 597
 598 2011-09-09  Alexey Proskuryakov  <ap@apple.com>
 599
 600         Add a script to create a monolithic script-test
 601         https://bugs.webkit.org/show_bug.cgi?id=67747
 602
 603         Reviewed by Darin Adler.
 604
 605         * Scripts/make-new-script-test: Added.
 606
 607 2011-09-09  Luiz Agostini  <luiz@webkit.org>
 608
 609         Adding myself to the reviewers list.
 610
 611         * Scripts/webkitpy/common/config/committers.py:
 612
 613 2011-09-09  Ryosuke Niwa  <rniwa@webkit.org>
 614
 615         Buildbot may use RunWebKitTest's evaluateCommand on bots that run nrwt
 616         https://bugs.webkit.org/show_bug.cgi?id=67855
 617
 618         Reviewed by Dirk Pranke.
 619
 620         The bug was caused by the fact some ports use RunWebKitTests instead of NewRunWebKitTests to run tests,
 621         and using RunWebKitTests.evaluateCommand after calling _parseNewRunWebKitTestsOutput.
 622
 623         Fixed the bug by merging evaluateCommand of RunWebKitTests and NewRunWebKitTests.
 624
 625         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
 626
 627 2011-09-09  Mark Hahnenberg  <mhahnenberg@apple.com>
 628
 629         Add myself to committers.py
 630         https://bugs.webkit.org/show_bug.cgi?id=67853
 631
 632         Reviewed by Darin Adler.
 633
 634         * Scripts/webkitpy/common/config/committers.py:
 635
 636 2011-09-09  Laszlo Gombos  <laszlo.1.gombos@nokia.com>
 637
 638         [Qt] Remove common.pri
 639         https://bugs.webkit.org/show_bug.cgi?id=67814
 640
 641         Reviewed by Andreas Kling.
 642
 643         * Tools.pro:
 644
 645 2011-09-09  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
 646
 647         [NRWT] the nrwt check httpd with --no-http option
 648         https://bugs.webkit.org/show_bug.cgi?id=67835
 649
 650         Reviewed by Tony Chang.
 651
 652         * Scripts/webkitpy/layout_tests/controllers/manager.py:
 653         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
 654
 655 2011-09-09  Hao Zheng  <zhenghao@chromium.org>
 656
 657         [chromium] Implement DRT support for Android.
 658         https://bugs.webkit.org/show_bug.cgi?id=67590
 659
 660         Reviewed by Tony Chang.
 661
 662         Add 2 mechanism to make DRT interaction more robust on Android:
 663         - Send 'QUIT' to DRT to make it stop.
 664         - Wait until receiving '#READY' from DRT, so that DRT won't miss any
 665         input commands before it starts.
 666         And LayoutTestController.cpp needs to include <cctype> for 'tolower'
 667         and <clocale> for 'LC_ALL' to compile on Android.
 668
 669         Also change to use sigaction instead of signal for TestShellGtk.cpp,
 670         as signal() man page recommends using sigaction().
 671
 672         * DumpRenderTree/DumpRenderTree.gypi:
 673         * DumpRenderTree/chromium/DumpRenderTree.cpp:
 674         (main):
 675         * DumpRenderTree/chromium/LayoutTestController.cpp:
 676         * DumpRenderTree/chromium/TestShellAndroid.cpp: Added.
 677         (AlarmHandler):
 678         (TestShell::waitTestFinished):
 679         (platformInit):
 680         (openStartupDialog):
 681         (checkLayoutTestSystemDependencies):
 682         * DumpRenderTree/chromium/TestShellGtk.cpp:
 683         (AlarmHandler):
 684         (TestShell::waitTestFinished):
 685
 686 2011-09-09  Kevin Ollivier  <kevino@theolliviers.com>
 687
 688         [wx] Unreviewed build fix. Add Source/JavaScriptCore/dfg directory to the build.
 689
 690         * waf/build/settings.py:
 691
 692 2011-09-09  Sheriff Bot  <webkit.review.bot@gmail.com>
 693
 694         Unreviewed, rolling out r94851.
 695         http://trac.webkit.org/changeset/94851
 696         https://bugs.webkit.org/show_bug.cgi?id=67847
 697
 698         It broke a webkitpy test (Requested by ossy on #webkit).
 699
 700         * Scripts/webkitpy/layout_tests/controllers/manager.py:
 701
 702 2011-09-09  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
 703
 704         [NRWT] the nrwt check httpd with --no-http option
 705         https://bugs.webkit.org/show_bug.cgi?id=67835
 706
 707         Reviewed by Csaba Osztrogonác.
 708
 709         * Scripts/webkitpy/layout_tests/controllers/manager.py:
 710
 711 2011-09-08  Brian Weinstein  <bweinstein@apple.com>
 712
 713         WebKit2: Assertion when calling didPerform*Redirect on null source/destination URL string
 714         https://bugs.webkit.org/show_bug.cgi?id=67794
 715         <rdar://problem/9892024>
 716
 717         Add a test that we don't send a didPerformServerRedirect when a load is cancelled from willSendRequest.
 718
 719         Reviewed by Brady Eidson.
 720
 721         * TestWebKitAPI/Tests/WebKit2/LoadCanceledNoServerRedirectCallback.cpp: Added.
 722         (TestWebKitAPI::didFinishLoadForFrame): Set flags on whether we've loaded the main frame, iframe, or both.
 723         (TestWebKitAPI::didPerformServerRedirect): Set a flag for whether we've performed a server redirect.
 724         (TestWebKitAPI::TEST): Set up a WKPageLoaderClient and WKContextHistoryClient, and load simple-iframe.html.
 725             Once both frames have loaded (main frame and blocked iframe), make sure that we have never called
 726             didPerformServerRedirect.
 727
 728         * TestWebKitAPI/Tests/WebKit2/LoadCanceledNoServerRedirectCallback_Bundle.cpp: Added.
 729         (TestWebKitAPI::LoadCanceledNoServerRedirectCallbackTest::LoadCanceledNoServerRedirectCallbackTest): Initialize
 730             our InjectedBundleTest.
 731         (TestWebKitAPI::LoadCanceledNoServerRedirectCallbackTest::willSendRequestForFrame): If we are dealing with a request
 732             from the main frame (the main resource in simple-iframe.html), allow it. If not (dealing with the iframe), block
 733             the request.
 734         (TestWebKitAPI::LoadCanceledNoServerRedirectCallbackTest::didCreatePage): Set up the WKBundlePageResourceLoadClient.
 735
 736         * TestWebKitAPI/Tests/WebKit2/simple-iframe.html: Added.
 737
 738         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Add new files.
 739         * TestWebKitAPI/win/TestWebKitAPI.vcproj: Ditto.
 740         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj: Ditto.
 741         * TestWebKitAPI/win/copy-resources.cmd: Ditto.
 742
 743 2011-09-08  David Levin  <levin@chromium.org>
 744
 745         check-webkit-style: Ref|Own Ptr& should be allowed as a parameter.
 746         https://bugs.webkit.org/show_bug.cgi?id=67797
 747
 748         Reviewed by Eric Seidel.
 749
 750         * Scripts/webkitpy/style/checkers/cpp.py: Allow Ref/Own Ptr when it ends with &.
 751         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added test cases for this.
 752
 753 2011-09-08  Fady Samuel  <fsamuel@chromium.org>
 754
 755         Fixed Layout Mode should be adjustable from layoutTestController for testing on Chromium platforms
 756         https://bugs.webkit.org/show_bug.cgi?id=67723
 757
 758         Reviewed by Adam Barth.
 759
 760         * DumpRenderTree/chromium/EventSender.h:
 761         * DumpRenderTree/chromium/LayoutTestController.cpp:
 762         (LayoutTestController::LayoutTestController):
 763         (LayoutTestController::enableFixedLayoutMode):
 764         (LayoutTestController::setFixedLayoutSize):
 765         * DumpRenderTree/chromium/LayoutTestController.h:
 766
 767 2011-09-07  Matthew Delaney  <mdelaney@apple.com>
 768
 769         Add standalone script that filters the output of build-webkit to be more human-readable
 770         https://bugs.webkit.org/show_bug.cgi?id=44081
 771
 772         Reviewed by Daniel Bates.
 773
 774         * Scripts/filter-build-webkit: Added. Just need to feed it the output of
 775         a build script (e.g. build-webkit) and it spits out a prettier version.
 776         * Scripts/VCSUtils.pm: Refactored "possiblyColored" out of run-api-tests into VCSUtils.
 777         * Scripts/run-api-tests: Now includes VCSUtils to use "possiblyColored"
 778
 779 2011-09-07  Noel Gordon  <noel.gordon@gmail.com>
 780
 781         Add myself to committers.py
 782         https://bugs.webkit.org/show_bug.cgi?id=67696
 783
 784         Reviewed by Kent Tamura.
 785
 786         * Scripts/webkitpy/common/config/committers.py:
 787
 788 2011-09-07  Adam Barth  <abarth@webkit.org>
 789
 790         rebaseline button in garden-o-matic details view should work
 791         https://bugs.webkit.org/show_bug.cgi?id=67738
 792
 793         Reviewed by Dimitri Glazkov.
 794
 795         This patch wires up the rebaseline button in garden-o-matic results
 796         details view.  The button is current per-test, so there isn't a way to
 797         rebaseline individual bots.  This is a place to start.  If we decide we
 798         want per-bot rebaselining, we'll need to think of some better UI.
 799
 800         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
 801         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions.js:
 802         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
 803         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
 804
 805 2011-09-07  Eric Seidel  <eric@webkit.org>
 806
 807         Teach the PortFactory object how to pass along executive/user/filesystem to port objects (which fixes the failing rebaseline_test on bots)
 808         https://bugs.webkit.org/show_bug.cgi?id=67737
 809
 810         Reviewed by Adam Barth.
 811
 812         I got a little over-eager when deprecating host.port_factory in the previous commit.
 813         Turns out that Host.port_factory was already the "modern" port/factory.py
 814         it was just the Module object.  I un-deprecated the variable, and changed it from
 815         being the module to being an actual PortFactory, passing it a host object.
 816         I also taught PortFactory how to set user/executive/filesystem in the kwargs
 817         when instantiating a new Port object.
 818         This magically fixed the test_rebaseline test, since its now no longer talking
 819         to the real filesystem.
 820
 821         * Scripts/webkitpy/common/host.py:
 822         * Scripts/webkitpy/layout_tests/port/factory.py:
 823         * Scripts/webkitpy/tool/commands/queries.py:
 824         * Scripts/webkitpy/tool/mocktool.py:
 825
 826 2011-09-07  Eric Seidel  <eric@webkit.org>
 827
 828         Add a new PortFactory class for creating new-style ports and deprecate the old PortFactory
 829         https://bugs.webkit.org/show_bug.cgi?id=67734
 830
 831         Reviewed by Adam Barth.
 832
 833         Right now every place that creates a Port object has to be very careful
 834         to pass in a MockSCM/MockUser/MockFilesystem when appropriate.
 835         The correct fix is for port/factory.py (which creates most port objects)
 836         to know how to pass these in automatically.  However, currently all port/factory.py
 837         functions are free functions instead of being on an object.  This patch
 838         adds a new PortFactory object and deprecates the PortFactory for the old-style Port objects.
 839
 840         In a follow-up patch I will add PortFactory to the Host object and teach PortFactory how to
 841         pass along the proper scm/user/filesystem to the Port object.
 842
 843         * Scripts/webkitpy/common/host.py:
 844         * Scripts/webkitpy/layout_tests/port/factory.py:
 845         * Scripts/webkitpy/tool/commands/queries.py:
 846         * Scripts/webkitpy/tool/mocktool.py:
 847
 848 2011-09-07  Sheriff Bot  <webkit.review.bot@gmail.com>
 849
 850         Unreviewed, rolling out r94441.
 851         http://trac.webkit.org/changeset/94441
 852         https://bugs.webkit.org/show_bug.cgi?id=67731
 853
 854         "MacEWSes are backed up due to new machines not able to reach
 855         git.webkit.org.  Disabling tests for now" (Requested by
 856         eseidel2 on #webkit).
 857
 858         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
 859
 860 2011-09-07  Elliot Poger  <epoger@google.com>
 861
 862         filter test_expectations properly for chromium-mac vs chromium-cg-mac
 863         https://bugs.webkit.org/show_bug.cgi?id=67324
 864
 865         Reviewed by Dirk Pranke.
 866
 867         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
 868         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
 869         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
 870         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
 871         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
 872
 873 2011-09-07  Eric Boren  <borenet@gmail.com>
 874
 875         [Chromium] Modify WebTouchEvent structure to match WebCore::TouchEvent
 876         https://bugs.webkit.org/show_bug.cgi?id=66800
 877
 878         Reviewed by Darin Fisher.
 879
 880         * DumpRenderTree/chromium/EventSender.cpp:
 881         (EventSender::sendCurrentTouchEvent):
 882
 883 2011-09-07  Jarred Nicholls  <jarred@sencha.com>
 884
 885         [Qt][DRT] Normalize file:///tmp/LayoutTests in LayoutTestController::pathToLocalResource()
 886         https://bugs.webkit.org/show_bug.cgi?id=67254
 887
 888         Reviewed by Csaba Osztrogonác.
 889
 890         Translate file:///tmp/LayoutTests/* urls into the repository LayoutTests directory, which is
 891         derived from the running location of DumpRenderTree binary.
 892
 893         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
 894         (LayoutTestController::pathToLocalResource):
 895         (LayoutTestController::setUserStyleSheetLocation): pass the url through pathToLocalResource
 896
 897 2011-09-07  Csaba Osztrogonác  <ossy@webkit.org>
 898
 899         REGRESSION(r94288) build-webkit doesn't work on Symbian
 900         https://bugs.webkit.org/show_bug.cgi?id=67706
 901
 902         * Scripts/webkitdirs.pm: Disable automatic clean build on Symbian, because it doesn't support shadow build.
 903         (buildQMakeProject):
 904
 905 2011-09-07  Csaba Osztrogonác  <ossy@webkit.org>
 906
 907         REGRESSION(r94288) build-webkit doesn't work on Windows
 908         https://bugs.webkit.org/show_bug.cgi?id=67646
 909
 910         Roll back r94655 with a small fix.
 911
 912         * Scripts/webkitdirs.pm:
 913         (buildQMakeProject):
 914
 915 2011-09-07  Sheriff Bot  <webkit.review.bot@gmail.com>
 916
 917         Unreviewed, rolling out r94635.
 918         http://trac.webkit.org/changeset/94635
 919         https://bugs.webkit.org/show_bug.cgi?id=67704
 920
 921         Linux Touch build is failing to compile this change.
 922         (Requested by loislo on #webkit).
 923
 924         * DumpRenderTree/chromium/EventSender.cpp:
 925         (EventSender::sendCurrentTouchEvent):
 926
 927 2011-09-07  Sheriff Bot  <webkit.review.bot@gmail.com>
 928
 929         Unreviewed, rolling out r94655.
 930         http://trac.webkit.org/changeset/94655
 931         https://bugs.webkit.org/show_bug.cgi?id=67703
 932
 933         It broke Leopard debug bot (Requested by ossy on #webkit).
 934
 935         * Scripts/webkitdirs.pm:
 936
 937 2011-09-07  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
 938
 939         REGRESSION(r94288) build-webkit doesn't work on Windows
 940         https://bugs.webkit.org/show_bug.cgi?id=67646
 941
 942         Reviewed by Csaba Osztrogonác.
 943
 944         * Scripts/webkitdirs.pm:
 945
 946 2011-09-06  Eric Boren  <borenet@gmail.com>
 947
 948         [Chromium] Modify WebTouchEvent structure to match WebCore::TouchEvent
 949         https://bugs.webkit.org/show_bug.cgi?id=66800
 950
 951         Reviewed by Darin Fisher.
 952
 953         * DumpRenderTree/chromium/EventSender.cpp:
 954         (EventSender::sendCurrentTouchEvent):
 955
 956 2011-09-06  Adam Barth  <abarth@webkit.org>
 957
 958         garden-o-matic details view should having working rebaseline and next/previous buttons
 959         https://bugs.webkit.org/show_bug.cgi?id=67659
 960
 961         Reviewed by Dimitri Glazkov.
 962
 963         This patch wires up basic back/forward buttons that let you traverse
 964         through the results we're examining in the details view.  This ended up
 965         being more code than I expected, but I wanted to keep all the state
 966         information in the DOM itself.
 967
 968         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
 969         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
 970         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
 971
 972 2011-09-01  Dirk Pranke  <dpranke@chromium.org>
 973
 974         fix MockFileSystem.glob(), refactor filesystem tests
 975         https://bugs.webkit.org/show_bug.cgi?id=67462
 976
 977         Reviewed by Eric Seidel.
 978
 979         In debugging why my proposed change for MockFileSystem.glob()
 980         in bug 66228 didn't work, I stumbled across two real bugs
 981         that cancelled each other out for the unit tests. I fixed those
 982         (glob shouldn't recurse into subdirs, and we weren't populating
 983         self.dirs from self.files properly).
 984
 985         I have also created a "mixin" class for tests that can be shared
 986         between the real filesystem and the mock filesystem - MockFileSystem
 987         is being used enough it needs its own tests.
 988
 989         * Scripts/webkitpy/common/system/filesystem_mock.py:
 990         * Scripts/webkitpy/common/system/filesystem_mock_unittest.py: Added.
 991         * Scripts/webkitpy/common/system/filesystem_unittest.py:
 992
 993 2011-09-06  Anders Carlsson  <andersca@apple.com>
 994
 995         Move NPAPI headers in bridge to plugins
 996         https://bugs.webkit.org/show_bug.cgi?id=67661
 997
 998         Update forwarding headers.
 999
1000         * DumpRenderTree/unix/TestNetscapePlugin/ForwardingHeaders/WebKit/npapi.h:
1001         * DumpRenderTree/unix/TestNetscapePlugin/ForwardingHeaders/WebKit/npruntime.h:
1002
1003 2011-09-06  Tony Chang  <tony@chromium.org>
1004
1005         [chromium] Check in 'build' for compiled binaries first
1006         https://bugs.webkit.org/show_bug.cgi?id=67549
1007
1008         Reviewed by Dirk Pranke.
1009
1010         I'm working on moving the windows build output to a common location
1011         rather than relative to the sln file.  Because of that, we need to
1012         check those build locations first for DumpRenderTree.exe.
1013
1014         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1015
1016 2011-09-06  Adam Barth  <abarth@webkit.org>
1017
1018         Use a hierarchial structure for the garden-o-matic results view
1019         https://bugs.webkit.org/show_bug.cgi?id=67620
1020
1021         Reviewed by Dimitri Glazkov.
1022
1023         This patch re-organizes the results view to use a hierachial structure
1024         for selection.  At the top-level are sections for each test.  Expanding
1025         the test section shows a tab selector for builders.
1026
1027         This structure prepares us for wiring up various actions to this view.
1028
1029         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
1030         (.):
1031         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
1032         ():
1033         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
1034         (.):
1035         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
1036         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
1037         (.results-view>.toolbar):
1038         (.results-view>.toolbar ul.actions):
1039         (.results-view>.toolbar ul.actions li):
1040         (.results-grid table):
1041         (.results-grid table td, .results-view table th):
1042         (.results-grid table th):
1043         (.results-grid .text-result):
1044         (.results-grid .image-result):
1045
1046 2011-09-06  Adam Barth  <abarth@webkit.org>
1047
1048         garden-o-matic's base.probe generates spam in the JavaScript console
1049         https://bugs.webkit.org/show_bug.cgi?id=67622
1050
1051         Reviewed by Dimitri Glazkov.
1052
1053         Now that we have CORS set up on the buildbot, we can use XMLHttpRequest
1054         instead of HTMLScriptElement to test whether results exist.  This
1055         reduces the console.log spam and saves bandwidth because we can use a
1056         HEAD request instead of a GET request.
1057
1058         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/net.js:
1059
1060 2011-09-05  Adam Barth  <abarth@webkit.org>
1061
1062         Add a tab to garden-o-matic for viewing expected failures
1063         https://bugs.webkit.org/show_bug.cgi?id=67619
1064
1065         Reviewed by Dimitri Glazkov.
1066
1067         This patch adds a basic tab to garden-o-matic that shows all the tests
1068         failing on the bots, including the expected failures.  This patch
1069         mostly involves refactoring existing classes to allow new subclasses.
1070
1071         This view is useful for when gardeners mark tests as expected to fail
1072         and then rebaseline them later.
1073
1074         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
1075         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
1076         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
1077         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary-mock.js:
1078         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
1079         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
1080         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
1081
1082 2011-09-05  Adam Barth  <abarth@webkit.org>
1083
1084         Fix global variable leak in garden-o-matic
1085         https://bugs.webkit.org/show_bug.cgi?id=67617
1086
1087         Reviewed by Dimitri Glazkov.
1088
1089         Caught by running the unit tests with noglobals turned on.
1090
1091         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
1092
1093 2011-09-05  Balazs Kelemen  <kbalazs@webkit.org>
1094
1095         [Qt][WK2] Add pixel test support
1096         https://bugs.webkit.org/show_bug.cgi?id=66283
1097
1098         Reviewed by Andreas Kling.
1099
1100         Implemented pixel result generation by reusing the logic
1101         that we have in DumpRenderTree.
1102         * WebKitTestRunner/qt/TestInvocationQt.cpp:
1103         (WTR::dumpImage):
1104         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
1105
1106 2011-09-04  Robin Dunn  <robin@alldunn.com>
1107
1108         [wx] Enable wxWebKit to run using the wxGC Cairo backend on platforms other than GTK.
1109         https://bugs.webkit.org/show_bug.cgi?id=67577
1110
1111         Reviewed by Kevin Ollivier.
1112
1113         * waf/build/settings.py:
1114
1115 2011-09-03  Kevin Ollivier  <kevino@theolliviers.com>
1116
1117         [wx] Unreviewed build fix. Add wtf/dtoa directory to build.
1118
1119         * waf/build/settings.py:
1120
1121 2011-09-03  Dan Bernstein  <mitz@apple.com>
1122
1123         32-bit build fix.
1124
1125         * TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
1126         (TestWebKitAPI::Util::toSTD):
1127
1128 2011-09-03  Dan Bernstein  <mitz@apple.com>
1129
1130         Move testStringByEvaluatingJavaScriptFromString() from DumpRenderTree to TestWebKitAPI
1131         https://bugs.webkit.org/show_bug.cgi?id=67559
1132
1133         Reviewed by Darin Adler.
1134
1135         * DumpRenderTree/mac/DumpRenderTree.mm:
1136         (dumpRenderTree): Moved testStringByEvaluatingJavaScriptFromString() from here and removed
1137         the call to it.
1138         * TestWebKitAPI/PlatformUtilities.h: Declared Util::toSTD(NSString *)
1139         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added StringByEvaluatingJavaScriptFromString.mm.
1140         * TestWebKitAPI/Tests/mac/StringByEvaluatingJavaScriptFromString.mm: Added.
1141         (TestWebKitAPI::TEST): Added. Moved testStringByEvaluatingJavaScriptFromString() to here.
1142         * TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
1143         (TestWebKitAPI::Util::toSTD): Added.
1144
1145 2011-09-03  Adam Barth  <abarth@webkit.org>
1146
1147         garden-o-matic results view should have a better results selector
1148         https://bugs.webkit.org/show_bug.cgi?id=67542
1149
1150         Reviewed by Dimitri Glazkov.
1151
1152         This patch adds a basic grid-based results selector to the results
1153         view.  Currently, this grid just shows how the tests fail on the
1154         various bots.  A future patch will let you select which failure to view
1155         and to mark the failures for rebaselining.
1156
1157         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
1158         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js:
1159           - Add some more basic functions for manipulating dictionaries.
1160         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
1161         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
1162         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
1163
1164 2011-09-03  Adam Barth  <abarth@webkit.org>
1165
1166         garden-o-matic results view should show Actual and Expected results for text failures
1167         https://bugs.webkit.org/show_bug.cgi?id=67536
1168
1169         Reviewed by Eric Seidel.
1170
1171         In the past, we didn't show actual and expected text results (we just
1172         showed the diff) because we didn't have the awesome six-by-two results
1173         grid.  Now that we do, we can move show these results.
1174
1175         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
1176         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js:
1177           - Fixed some typos that were preventing these unit tests from running.
1178         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
1179         (.):
1180         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results_unittests.js:
1181           - Report -actual.txt and -expected.txt as possible results URLs.
1182
1183 2011-09-03  Adam Barth  <abarth@webkit.org>
1184
1185         garden-o-matic shows "ghost" failures that are already fixed
1186         https://bugs.webkit.org/show_bug.cgi?id=67535
1187
1188         Reviewed by Eric Seidel.
1189
1190         About 10% of writes fail to stick on test-results.appspot.com.  This
1191         patch causes us to fetch the full_results.json blob directly from
1192         build.chromium.org rather than relying upon AppEngine.
1193
1194         Unforuntately, we still need to get historical information from
1195         AppEngine.  That will take several patches to fix becuase I need to
1196         change the bots to write the full_results.json file to the
1197         build-specific directory in addition to the zip file.
1198
1199         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
1200         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
1201         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
1202
1203 2011-09-03  Sheriff Bot  <webkit.review.bot@gmail.com>
1204
1205         Unreviewed, rolling out r94458.
1206         http://trac.webkit.org/changeset/94458
1207         https://bugs.webkit.org/show_bug.cgi?id=67558
1208
1209         Caused appcache test to fail on various bots (Requested by
1210         rniwa on #webkit).
1211
1212         * Scripts/webkitpy/layout_tests/port/webkit.py:
1213         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1214
1215 2011-09-03  Sheriff Bot  <webkit.review.bot@gmail.com>
1216
1217         Unreviewed, rolling out r94453.
1218         http://trac.webkit.org/changeset/94453
1219         https://bugs.webkit.org/show_bug.cgi?id=67557
1220
1221         Caused appcache test to fail on various bots (Requested by
1222         rniwa on #webkit).
1223
1224         * Scripts/webkitpy/layout_tests/port/webkit.py:
1225
1226 2011-09-02  Michael Saboff  <msaboff@apple.com>
1227
1228         Add JSC:RegExp functional tests
1229         https://bugs.webkit.org/show_bug.cgi?id=67339
1230
1231         New perl script to build JavaScriptCore and run the RegExp tests.
1232
1233         Reviewed by Gavin Barraclough.
1234
1235         * Scripts/run-regexp-tests: Added.
1236
1237 2011-09-02  Michael Saboff  <msaboff@apple.com>
1238
1239         Add JSC:RegExp functional tests
1240         https://bugs.webkit.org/show_bug.cgi?id=67339
1241
1242         New perl script to build JavaScriptCore and run the RegExp tests.
1243
1244         Reviewed by Gavin Barraclough.
1245
1246         * Scripts/run-regexp-tests: Added.
1247
1248 2011-09-02  Eric Seidel  <eric@webkit.org>
1249
1250         Reshuffle some code in WebKitDriver._read_block in preparation for reading stderr/stdout separately
1251         https://bugs.webkit.org/show_bug.cgi?id=67530
1252
1253         Unreviewed.  Fixing typo from previous commit.
1254
1255         Turns out there was *no* unittesting of WebKitDriver.  Added a basic test of _read_block
1256         which exercises the code I previously made a typo in.
1257
1258         * Scripts/webkitpy/layout_tests/port/webkit.py:
1259         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1260
1261 2011-09-02  Eric Seidel  <eric@webkit.org>
1262
1263         Reshuffle some code in WebKitDriver._read_block in preparation for reading stderr/stdout separately
1264         https://bugs.webkit.org/show_bug.cgi?id=67530
1265
1266         Reviewed by Adam Barth.
1267
1268         No functional change, just reshuffling code.
1269
1270         * Scripts/webkitpy/layout_tests/port/webkit.py:
1271
1272 2011-09-02  Darin Adler  <darin@apple.com>
1273
1274         Added Subversion ignore for .pyc generated files.
1275
1276         * Scripts/webkitpy/layout_tests/models: Added property svn:ignore.
1277         * Scripts/webkitpy/layout_tests/servers: Added property svn:ignore.
1278         * Scripts/webkitpy/tool/servers: Added property svn:ignore.
1279
1280 2011-09-02  Adam Barth  <abarth@webkit.org>
1281
1282         Enable testing on mac-ews
1283         https://bugs.webkit.org/show_bug.cgi?id=67505
1284
1285         Reviewed by Eric Seidel.
1286
1287         This patch enables testing on the Mac EWS, which should help
1288         contributors not turn the Mac bots red.
1289
1290         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
1291
1292 2011-09-01  Beth Dakin  <bdakin@apple.com>
1293
1294         https://bugs.webkit.org/show_bug.cgi?id=67451
1295         Add a test for going back to a cached page after changing the device scale factor
1296
1297         Reviewed by Adam Roben.
1298
1299         New test DeviceScaleFactorOnBack.mm inherits from WebKitAgnosticTest.
1300         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1301         * TestWebKitAPI/Tests/mac/DeviceScaleFactorOnBack.mm: Added.
1302         (TestWebKitAPI::DeviceScaleFactorOnBack::url):
1303         (TestWebKitAPI::DeviceScaleFactorOnBack::didLoadURL):
1304         (TestWebKitAPI::DeviceScaleFactorOnBack::createWindow):
1305         (TestWebKitAPI::DeviceScaleFactorOnBack::initializeView):
1306         (TestWebKitAPI::DeviceScaleFactorOnBack::runTest):
1307         (TestWebKitAPI::TEST_F):
1308
1309         Added new function to WebKitAgnosticTest to goBack. Separated
1310         waitForLoadToFinish() into its own function (instead of being a part of
1311         loadAndWaitUntilFinished()) so that it can be called from goBack as
1312         well. Also added initializeView() so we can set the cache model.
1313         * TestWebKitAPI/mac/WebKitAgnosticTest.h:
1314         (TestWebKitAPI::WebKitAgnosticTest::initializeView):
1315         * TestWebKitAPI/mac/WebKitAgnosticTest.mm:
1316         (TestWebKitAPI::WebKitAgnosticTest::WebKitAgnosticTest):
1317         (TestWebKitAPI::WebKitAgnosticTest::runWebKit1Test):
1318         (TestWebKitAPI::WebKitAgnosticTest::runWebKit2Test):
1319         (TestWebKitAPI::WebKitAgnosticTest::loadURL):
1320         (TestWebKitAPI::WebKitAgnosticTest::goBack):
1321         (TestWebKitAPI::WebKitAgnosticTest::waitForLoadToFinish):
1322
1323 2011-09-02  Anders Carlsson  <andersca@apple.com>
1324
1325         Build fix.
1326
1327         * DumpRenderTree/TestNetscapePlugIn/Tests/PrivateBrowsing.cpp:
1328         (PrivateBrowsing::ScriptableObject::pluginTest):
1329
1330 2011-09-02  Anders Carlsson  <andersca@apple.com>
1331
1332         Move private browsing test to a PluginTest subclass
1333         https://bugs.webkit.org/show_bug.cgi?id=67498
1334
1335         Reviewed by Darin Adler.
1336
1337         * DumpRenderTree/DumpRenderTree.gypi:
1338         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
1339         * DumpRenderTree/TestNetscapePlugIn/PluginObject.cpp:
1340         (pluginGetProperty):
1341         * DumpRenderTree/TestNetscapePlugIn/PluginObject.h:
1342         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
1343         (PluginTest::NPP_SetValue):
1344         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
1345         * DumpRenderTree/TestNetscapePlugIn/Tests/PrivateBrowsing.cpp: Added.
1346         (PrivateBrowsing::PrivateBrowsing):
1347         (PrivateBrowsing::privateBrowsingEnabled):
1348         (PrivateBrowsing::cachedPrivateBrowsingEnabled):
1349         (PrivateBrowsing::ScriptableObject::hasProperty):
1350         (PrivateBrowsing::ScriptableObject::getProperty):
1351         (PrivateBrowsing::ScriptableObject::pluginTest):
1352         (PrivateBrowsing::NPP_New):
1353         (PrivateBrowsing::NPP_GetValue):
1354         (PrivateBrowsing::NPP_SetValue):
1355         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
1356         (NPP_New):
1357         (NPP_SetValue):
1358         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
1359         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
1360         * GNUmakefile.am:
1361
1362 2011-09-02  Csaba Osztrogonác  <ossy@webkit.org>
1363
1364         [Qt] Build dependency problems
1365         https://bugs.webkit.org/show_bug.cgi?id=38054
1366
1367         Trivial fix after r94288.
1368
1369         Rubber-stamped by Andras Becsi.
1370
1371         * Scripts/webkitdirs.pm: We should read the whole defaults.txt, not only the first two lines.
1372         (buildQMakeProject):
1373
1374 2011-09-02  Andras Becsi  <andras.becsi@nokia.com>
1375
1376         Remove my old e-mail address and add my new one.
1377
1378         Rubber-stamped by Tor Arne Vestbø.
1379
1380         * Scripts/webkitpy/common/config/committers.py:
1381
1382 2011-09-02  Philippe Normand  <pnormand@igalia.com>
1383
1384         [GTK] run-gtk-tests leaves a zombie Xvfb process
1385         https://bugs.webkit.org/show_bug.cgi?id=67389
1386
1387         Reviewed by Martin Robinson.
1388
1389         * Scripts/run-gtk-tests: Properly close STDIN and STDERR on the
1390         child process instead of doing shell redirection.
1391
1392 2011-09-01  Mark Rowe  <mrowe@apple.com>
1393
1394         Fix the build by adding some more missing return types.
1395
1396         * DumpRenderTree/mac/ResourceLoadDelegate.mm:
1397         (-[ResourceLoadDelegate webView:identifierForInitialRequest:fromDataSource:]):
1398
1399 2011-09-01  Eric Seidel  <eric@webkit.org>
1400
1401         REGRESSION (NRWT): Leaks Viewer can't load leaks from test runs that used NRWT
1402         https://bugs.webkit.org/show_bug.cgi?id=66228
1403
1404         Reviewed by Dirk Pranke.
1405
1406         ORWT used $(PROCESS)$(NUMBER)-leaks.txt for leaks filenames.
1407         The LeaksViewer tool globs for all *-leaks.txt files.
1408         This patch makes NRWT output $(PROCESS)-$(PID)-leaks.txt which should match.
1409
1410         In order to test this I had to fix a limitation in our MockFileSystem.glob
1411         method.  However, doing so uncovered a typo and bug in the integration tests.
1412         I've disabled the offending integration tests.
1413
1414         * Scripts/webkitpy/common/system/filesystem_mock.py:
1415         * Scripts/webkitpy/layout_tests/port/leakdetector.py:
1416         * Scripts/webkitpy/layout_tests/port/leakdetector_unittest.py:
1417
1418 2011-09-01  Dominic Mazzoni  <dmazzoni@google.com>
1419
1420         Adds a LayoutTestController method to make it possible to test
1421         starting speech input via an API call.
1422         https://bugs.webkit.org/show_bug.cgi?id=60170
1423
1424         Reviewed by Dimitri Glazkov.
1425
1426         * DumpRenderTree/LayoutTestController.h:
1427         * DumpRenderTree/chromium/LayoutTestController.cpp:
1428         (LayoutTestController::LayoutTestController):
1429         (LayoutTestController::startSpeechInput):
1430         * DumpRenderTree/chromium/LayoutTestController.h:
1431         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1432         (LayoutTestController::startSpeechInput):
1433         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1434         (LayoutTestController::startSpeechInput):
1435         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
1436         (LayoutTestController::startSpeechInput):
1437         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1438         (LayoutTestController::startSpeechInput):
1439         * DumpRenderTree/qt/LayoutTestControllerQt.h:
1440         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1441         (LayoutTestController::startSpeechInput):
1442         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
1443         (LayoutTestController::startSpeechInput):
1444
1445 2011-09-01  Tim Horton  <hortont424@gmail.com>
1446
1447         Fix TestWebKitAPI build on clang-2.0
1448
1449         * TestWebKitAPI/Tests/WTF/CheckedArithmeticOperations.cpp:
1450
1451 2011-09-01  Adam Barth  <abarth@webkit.org>
1452
1453         Remove unused "extension" parameter from rebaseline gardeningserver handler
1454         https://bugs.webkit.org/show_bug.cgi?id=67384
1455
1456         Reviewed by James Robinson.
1457
1458         This parameter hasn't done anything in a while.
1459
1460         * Scripts/webkitpy/tool/commands/rebaseline.py:
1461         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1462         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
1463
1464 2011-09-01  Eric Seidel  <eric@webkit.org>
1465
1466         Unreviewed.
1467
1468         Fix a typo in unittests.  Unfortunately fixing it
1469         caused several assertions.  Just disabling the test
1470         instead, since it was providing no value before due to the typo.
1471
1472         * Scripts/webkitpy/layout_tests/port/test.py:
1473
1474 2011-09-01  Adam Barth  <abarth@webkit.org>
1475
1476         Add a "rebaseline" button to the garden-o-matic summary page
1477         https://bugs.webkit.org/show_bug.cgi?id=67382
1478
1479         Reviewed by Dimitri Glazkov.
1480
1481         As much as we might wish folks to not rebaseline entire test groups at
1482         once, the truth is that doing so is a common operation.  Adding a
1483         button for this operation is likely to make folks happy.
1484
1485         This patch also cleans up a bunch of plumbing surrounding failureInfo
1486         objects.
1487
1488         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
1489         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js:
1490         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout.js:
1491         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout_unittests.js:
1492         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
1493         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
1494         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
1495         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results_unittests.js:
1496         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
1497         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
1498
1499 2011-09-01  Adam Barth  <abarth@webkit.org>
1500
1501         garden-o-matic should let the gardener assign blame for a group of test failures
1502         https://bugs.webkit.org/show_bug.cgi?id=67379
1503
1504         Reviewed by Dimitri Glazkov.
1505
1506         If a group of test failures has a large revision range, the gardener
1507         might wish to blame a specific revision for the failure.  Currently,
1508         this only effect of assigning blame in this way is to hide the other
1509         suspect commits.
1510
1511         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
1512         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions.js:
1513         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
1514
1515 2011-09-01  Adam Barth  <abarth@webkit.org>
1516
1517         Hide garden-o-matic failure table until hover
1518         https://bugs.webkit.org/show_bug.cgi?id=67376
1519
1520         Reviewed by Dimitri Glazkov.
1521
1522         This patch removes the failure table for build failures (because it
1523         does not make sense) and hides the failure table for test groups until
1524         the user hovers over the test failure block, reducing visual noise.
1525
1526         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
1527         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
1528         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css:
1529
1530 2011-09-01  Adam Barth  <abarth@webkit.org>
1531
1532         Switch garden-o-matic from commitData.title to commitData.summary
1533         https://bugs.webkit.org/show_bug.cgi?id=67375
1534
1535         Reviewed by Dimitri Glazkov.
1536
1537         commitData.title is a string provided by Trac, which isn't great in all
1538         cases.  commitData.summary is my attempt to find the most important
1539         one-line summary of the commit.  This patch switches garden-o-matic
1540         over to summary.
1541
1542         Also, this patch tweaks a bunch of CSS to tighten up the summary page
1543         and reduce visual noise.
1544
1545         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
1546         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
1547         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css:
1548
1549 2011-09-01  Adam Barth  <abarth@webkit.org>
1550
1551         Group failing tests in garden-o-matic if there are more than three in a folder
1552         https://bugs.webkit.org/show_bug.cgi?id=67373
1553
1554         Reviewed by Dimitri Glazkov.
1555
1556         Sometimes a failure will cause many tests in a given directory to fail.
1557         When that happens, it's better to list the directory as a single item
1558         (with a count) rather than spamming all the test names.
1559
1560         (This patch also updates a bunch of unit test expected results after my last commit.)
1561
1562         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
1563         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js:
1564         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js:
1565         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
1566         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
1567
1568 2011-09-01  Adam Barth  <abarth@webkit.org>
1569
1570         Lower the visual noise from the failing builders display
1571         https://bugs.webkit.org/show_bug.cgi?id=67369
1572
1573         Reviewed by Dimitri Glazkov.
1574
1575         This patch ensures that the "Building" row in the failing builders
1576         display is always last and dims it by 50% to reduce visual noise.  This
1577         lets the gardener focus on the actual failures.
1578
1579         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
1580         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures.js:
1581         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/failures.css: Added.
1582
1583 2011-09-01  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
1584
1585         [Qt] Build dependency problems
1586         https://bugs.webkit.org/show_bug.cgi?id=38054
1587
1588         Make clean build after the defines are changed.
1589
1590         Reviewed by Csaba Osztrogonác.
1591
1592         * Scripts/webkitdirs.pm:
1593         (buildQMakeProject):
1594
1595 2011-08-31  Adam Barth  <abarth@webkit.org>
1596
1597         Move DumpRenderTree to WebKitPlatformSupport
1598         https://bugs.webkit.org/show_bug.cgi?id=67331
1599
1600         Reviewed by Darin Fisher.
1601
1602         This patch update DumpRenderTree to account for the WebKitClient =>
1603         WebKitPlatformSupport rename.
1604
1605         * DumpRenderTree/chromium/NotificationPresenter.cpp:
1606         (NotificationPresenter::show):
1607         * DumpRenderTree/chromium/Task.cpp:
1608         (postTask):
1609
1610 2011-08-31  Eric Seidel  <eric@webkit.org>
1611
1612         Add queues.webkit.org/active-bots page to make it easy to see which bots are down
1613         https://bugs.webkit.org/show_bug.cgi?id=67314
1614
1615         Reviewed by Adam Barth.
1616
1617         Very simple page.  Just lists all the bots, what queue they belong to and what we last heard from them.
1618         If we haven't heard from a bot in the last 500 messages (last day or so), then it won't show up here.
1619         This makes it very easy for me to see which of my cr-jail bots might be down at any one time.
1620
1621         I also moved a bunch of code into dashboard.css which belonged there.
1622
1623         * QueueStatusServer/handlers/activebots.py: Added.
1624         * QueueStatusServer/main.py:
1625         * QueueStatusServer/stylesheets/dashboard.css:
1626         (.queue_bubble):
1627         (.queue_name):
1628         (.last_heard_from):
1629         (.status_text):
1630         (.alive):
1631         (.behind):
1632         (.dead):
1633         * QueueStatusServer/templates/activebots.html: Added.
1634         * QueueStatusServer/templates/dashboard.html:
1635         * QueueStatusServer/templates/recentstatus.html:
1636
1637 2011-08-31  Oliver Hunt  <oliver@apple.com>
1638
1639         Move CheckedArithmeticOperations.cpp to the intended location
1640
1641         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1642         * TestWebKitAPI/Tests/WTF/CheckedArithmeticOperations.cpp: Renamed from Tools/TestWebKitAPI/Tests/CheckedArithmeticOperations.cpp.
1643
1644 2011-08-31  Elliot Poger  <epoger@google.com>
1645
1646         add unittest to validate the complete set of Chromium configurations
1647         https://bugs.webkit.org/show_bug.cgi?id=67211
1648
1649         Reviewed by Adam Barth.
1650
1651         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
1652
1653 2011-08-31  Nico Weber  <thakis@chromium.org>
1654
1655         Make ScrollAnimator(Chromium)?Mac check scrollAnimatorEnabled()
1656         https://bugs.webkit.org/show_bug.cgi?id=67297
1657
1658         Reviewed by Tony Chang.
1659
1660         Explicitly disable smooth scrolling in chromium's DRT.
1661
1662         * DumpRenderTree/chromium/WebPreferences.cpp:
1663         (WebPreferences::applyTo):
1664
1665 2011-08-31  Oliver Hunt  <oliver@apple.com>
1666
1667         Add Checked<> tests to windows build
1668
1669         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
1670
1671 2011-08-30  Oliver Hunt  <oliver@apple.com>
1672
1673         Add support for checked arithmetic
1674         https://bugs.webkit.org/show_bug.cgi?id=67095
1675
1676         Reviewed by Sam Weinig.
1677
1678         Add test cases for Checked<>
1679
1680         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1681         * TestWebKitAPI/Tests/CheckedArithmeticOperations.cpp: Added.
1682
1683 2011-08-31  Adam Barth  <abarth@webkit.org>
1684
1685         Move summary.html to garden-o-matic.html.
1686
1687         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html.
1688         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js.
1689         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js: Removed.
1690         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html: Removed.
1691
1692 2011-08-31  Noel Gordon  <noel.gordon@gmail.com>
1693
1694         [chromium] DRT linux should pass nativeKeyCode to plugins
1695         https://bugs.webkit.org/show_bug.cgi?id=66679
1696
1697         Reviewed by Tony Chang.
1698
1699         * DumpRenderTree/chromium/EventSender.cpp:
1700         (EventSender::keyDown):  Call a Chromium Linux GTK helper routine that returns the
1701         (simulated) nativeKeyCode for the given input keyboard code.
1702
1703 2011-08-31  Adam Barth  <abarth@webkit.org>
1704
1705         Tighen up UI for garden-o-matic tab strip
1706         https://bugs.webkit.org/show_bug.cgi?id=67266
1707
1708         Reviewed by Dimitri Glazkov.
1709
1710         This patch add CSS to move the tab strip to the edges of the page and
1711         to remove some gratuitous border-radius.
1712
1713         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/common.css:
1714         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html:
1715
1716 2011-08-31  Adam Barth  <abarth@webkit.org>
1717
1718         garden-o-matic results view shouldn't be horrifically ugly
1719         https://bugs.webkit.org/show_bug.cgi?id=67264
1720
1721         Reviewed by Dimitri Glazkov.
1722
1723         Minor tweak to the results view CSS to make it fit in better with the new tabbed UI.
1724
1725         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
1726         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
1727         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
1728
1729 2011-08-31  Adam Barth  <abarth@webkit.org>
1730
1731         garden-o-matic needs a tabstrip
1732         https://bugs.webkit.org/show_bug.cgi?id=67263
1733
1734         Reviewed by Dimitri Glazkov.
1735
1736         This patch adds a basic tabstrip to garden-o-matic.  The results
1737         details view now opens in the "Results" tab, and the user can flip back
1738         and forth between the Summary and the Results tab quickly.  There's
1739         also a stub for a "Commits" tab, which is disabled.
1740
1741         This patch isn't the most beautiful in the world, but it gets the ball
1742         rolling on a multi-modal UI.
1743
1744         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
1745         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js:
1746         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
1747         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/common.css:
1748         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html:
1749
1750 2011-08-31  Ai Makabi  <makabi@google.com>
1751
1752         Extract reference links from reftest test file.
1753         https://bugs.webkit.org/show_bug.cgi?id=66838
1754
1755         Reviewed by Shinichiro Hamaji.
1756
1757         * Scripts/webkitpy/layout_tests/reftests/__init__.py: Added.
1758         * Scripts/webkitpy/layout_tests/reftests/extract_reference_link.py: Added.
1759         * Scripts/webkitpy/layout_tests/reftests/extract_reference_link_unittest.py: Added.
1760
1761 2011-08-03  Philippe Normand  <pnormand@igalia.com>
1762
1763         [webkitpy] missing log import in common/net/resultsjsonparser.py
1764         https://bugs.webkit.org/show_bug.cgi?id=65606
1765
1766         Reviewed by Martin Robinson.
1767
1768         * Scripts/webkitpy/common/net/resultsjsonparser.py: Import the logging module.
1769
1770 2011-08-30  Philippe Normand  <pnormand@igalia.com>
1771
1772         [stylechecker] complains about NULL usage in some gst_structure_ calls
1773         https://bugs.webkit.org/show_bug.cgi?id=67194
1774
1775         Reviewed by Martin Robinson.
1776
1777         Don't warn about NULL in some gst_structure_ calls. 0 can't be
1778         used for these, a proper sentinel is required.
1779
1780         * Scripts/webkitpy/style/checkers/cpp.py:
1781         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1782
1783 2011-08-30  Adam Barth  <abarth@webkit.org>
1784
1785         Remove old-style UI from garden-o-matic
1786         https://bugs.webkit.org/show_bug.cgi?id=67257
1787
1788         Reviewed by Dimitri Glazkov.
1789
1790         This code is nowhere near as awesome as the new hotness.  Let's remove
1791         it so it doesn't weigh us down.
1792
1793         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html: Removed.
1794         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/main.js: Removed.
1795         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
1796         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
1797         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/unexpected-passes.js: Removed.
1798         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/unexpected-passes.html: Removed.
1799
1800 2011-08-30  Takashi Toyoshima  <toyoshim@chromium.org>
1801
1802         [WebSocket] update pywebsocket to 0.6b5.
1803         https://bugs.webkit.org/show_bug.cgi?id=66924
1804
1805         Reviewed by Kent Tamura.
1806
1807         pywebsocket 0.6b2 supported minimum features of Hybi-08 specification.
1808         Now, pywebsocket 0.6b5 supports full features of Hybi-10 specification.
1809         0.6b3 or later supports a closing handshake handler and it is useful to
1810         verify client's closing handshake implementation.
1811         0.6b4 supports AbortedByUserException. It is useful to eliminate
1812         unnecessary server warning messages on intended abnormal closures.
1813         0.6b5 supports Python 2.5.1 or Mac 10.5.
1814
1815         * Scripts/webkitpy/thirdparty/__init__.py:
1816         Update download URL and directory path.
1817
1818 2011-08-30  Dan Bernstein  <mitz@apple.com>
1819
1820         Test for <rdar://problem/9281695> Add text search API for getting the DOM range of a text match
1821         https://bugs.webkit.org/show_bug.cgi?id=67230
1822
1823         Reviewed by Darin Adler.
1824
1825         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1826         * TestWebKitAPI/Tests/mac/DOMRangeOfString.html: Added.
1827         * TestWebKitAPI/Tests/mac/DOMRangeOfString.mm: Added.
1828         (-[DOMRangeOfStringFrameLoadDelegate webView:didFinishLoadForFrame:]):
1829         (TestWebKitAPI::TEST):
1830
1831 2011-08-30  Aaron Colwell  <acolwell@chromium.org>
1832
1833         Add MediaSource API to HTMLMediaElement
1834         https://bugs.webkit.org/show_bug.cgi?id=64731
1835
1836         Reviewed by Eric Carlson.
1837
1838         * Scripts/build-webkit:
1839
1840 2011-08-30  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
1841
1842         Emit last progress notification before calling dispatchDidFinishLoad
1843         https://bugs.webkit.org/show_bug.cgi?id=28851
1844
1845         Reviewed by Adam Barth.
1846
1847         Implement dumpProgressFinishedCallback() in the layoutTestController for Qt,
1848         Chromium and GTK ports. The first two were simple changes.
1849
1850         Some refactoring was needed in GTK port to assure that the dumps are emitted
1851         correctly and the processing after the page load finished is in the right place.
1852
1853         To detect the load finished, we use the callback for the "load-status" notifier
1854         signal of the top loading frame. And we use the deprecated "load-finished" signal
1855         for dumping (if enabled) the progress finished expected output.
1856
1857         * DumpRenderTree/LayoutTestController.cpp:
1858         (LayoutTestController::LayoutTestController):
1859         (dumpProgressFinishedCallbackCallback):
1860         (LayoutTestController::staticFunctions):
1861         * DumpRenderTree/LayoutTestController.h:
1862         (LayoutTestController::dumpProgressFinishedCallback):
1863         (LayoutTestController::setDumpProgressFinishedCallback):
1864         * DumpRenderTree/chromium/LayoutTestController.cpp:
1865         (LayoutTestController::LayoutTestController):
1866         (LayoutTestController::dumpProgressFinishedCallback):
1867         (LayoutTestController::reset):
1868         * DumpRenderTree/chromium/LayoutTestController.h:
1869         (LayoutTestController::shouldDumpProgressFinishedCallback):
1870         (LayoutTestController::setShouldDumpProgressFinishedCallback):
1871         * DumpRenderTree/chromium/WebViewHost.cpp:
1872         (WebViewHost::didStopLoading):
1873         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1874         (webViewLoadFinished):
1875         (topLoadingFrameLoadFinished):
1876         (webFrameLoadStatusNotified):
1877         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1878         (LayoutTestController::reset):
1879         (LayoutTestController::dumpProgressFinishedCallback):
1880         * DumpRenderTree/qt/LayoutTestControllerQt.h:
1881
1882 2011-08-30  Filip Pizlo  <fpizlo@apple.com>
1883
1884         JavaScriptCore does not have a good mechanism for performance
1885         regression testing
1886         https://bugs.webkit.org/show_bug.cgi?id=67177
1887
1888         Reviewed by Darin Adler.
1889
1890         Added a new script, which allows for testing one, two, or more VMs
1891         at once on SunSpider, V8, and Kraken benchmarks.  Benchmark/VM runs
1892         are interleaved at random to minimize systematics.  The results
1893         that bencher prints includes all possible relevant information about
1894         how bencher was invoked.  Bencher can be easily used to measure
1895         performance changes down to fractions of a percent.
1896
1897         Bencher also includes experimental support for running benchmarks in
1898         a DumpRenderTree release build.  This works in SunSpider and V8, but
1899         not yet in Kraken.  Running benchmarks in DumpRenderTree more
1900         closely mimics the performance when running in browser.
1901
1902         * Scripts/bencher: Added.
1903
1904 2011-08-29  Alexey Proskuryakov  <ap@apple.com>
1905
1906         DumpRenderTree should begin each test with an empty cookie store
1907         https://bugs.webkit.org/show_bug.cgi?id=63545
1908         <rdar://problem/5666907>
1909
1910         Reviewed by Darin Adler.
1911
1912         * DumpRenderTree/mac/DumpRenderTree.mm:
1913         (resetDefaultsToConsistentValues):
1914         (setDefaultsToConsistentValuesForTesting):
1915         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
1916         (LayoutTestController::setAlwaysAcceptCookies):
1917         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1918         (WTR::InjectedBundle::beginTesting):
1919         Use new SPIs to manipulate sessions.
1920
1921 2011-08-30  Sheriff Bot  <webkit.review.bot@gmail.com>
1922
1923         Unreviewed, rolling out r94055 and r94057.
1924         http://trac.webkit.org/changeset/94055
1925         http://trac.webkit.org/changeset/94057
1926         https://bugs.webkit.org/show_bug.cgi?id=67191
1927
1928         Many failures on Mac 10.5 (Requested by toyoshim on #webkit).
1929
1930         * Scripts/webkitpy/thirdparty/__init__.py:
1931
1932 2011-08-30  Adam Barth  <abarth@webkit.org>
1933
1934         Chromium Android should not be a core builder
1935         https://bugs.webkit.org/show_bug.cgi?id=67184
1936
1937         Reviewed by Ryosuke Niwa.
1938
1939         This bot hasn't been consistently green yet because it has never actually run!
1940
1941         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
1942         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
1943
1944 2011-08-30  Takashi Toyoshima  <toyoshim@chromium.org>
1945
1946         [WebSocket] update pywebsocket to 0.6b4.
1947         https://bugs.webkit.org/show_bug.cgi?id=66924
1948
1949         Reviewed by Kent Tamura.
1950
1951         pywebsocket 0.6b2 supported minimum features of Hybi-08 specification.
1952         Now, pywebsocket 0.6b4 supports full features of Hybi-10 specification.
1953         0.6b3 or later supports a closing handshake handler and it is useful to
1954         verify client's closing handshake implementation.
1955         0.6b4 supports AbortedByUserException. It is useful to eliminate
1956         unnecessary server warning messages on intended abnormal closures.
1957
1958         * Scripts/webkitpy/thirdparty/__init__.py:
1959         Update download URL and directory path.
1960
1961 2011-08-29  Tom Zakrajsek  <tomz@codeaurora.org>
1962
1963         sheriffbot "usage" help statements are wrong
1964         https://bugs.webkit.org/show_bug.cgi?id=67170
1965
1966         Reviewed by Adam Barth.
1967
1968         * Scripts/webkitpy/tool/bot/irc_command.py:
1969         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
1970         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
1971
1972 2011-08-29  Adam Barth  <abarth@webkit.org>
1973
1974         Add support for chomium-mac (non-CoreGraphics) baselines
1975         https://bugs.webkit.org/show_bug.cgi?id=67163
1976
1977         Reviewed by Eric Seidel.
1978
1979         This patch teaches the non-CG chromium-mac bots to look at non-CG
1980         baselines.  However, the patch still leaves CG as the default set of
1981         baselines.
1982
1983         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
1984         * Scripts/webkitpy/common/config/build.py:
1985         * Scripts/webkitpy/common/config/build_unittest.py:
1986         * Scripts/webkitpy/layout_tests/port/builders.py:
1987         * Scripts/webkitpy/layout_tests/port/chromium.py:
1988         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1989         * Scripts/webkitpy/layout_tests/port/factory.py:
1990         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
1991         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
1992         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
1993         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1994
1995 2011-08-29  Shawn Singh  <shawnsingh@chromium.org>
1996
1997         Small fix in chromium port of new-run-webkit-tests.
1998         https://bugs.webkit.org/show_bug.cgi?id=67078
1999
2000         Reviewed by Dirk Pranke.
2001
2002         * Scripts/webkitpy/layout_tests/port/chromium.py:
2003
2004 2011-08-29  Sheriff Bot  <webkit.review.bot@gmail.com>
2005
2006         Unreviewed, rolling out r93987, r93992, r93995, r93998, and
2007         r93999.
2008         http://trac.webkit.org/changeset/93987
2009         http://trac.webkit.org/changeset/93992
2010         http://trac.webkit.org/changeset/93995
2011         http://trac.webkit.org/changeset/93998
2012         http://trac.webkit.org/changeset/93999
2013         https://bugs.webkit.org/show_bug.cgi?id=67147
2014
2015         Many failing tests (Requested by ap on #webkit).
2016
2017         * DumpRenderTree/mac/DumpRenderTree.mm:
2018         (setDefaultsToConsistentValuesForTesting):
2019         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
2020         (WTR::InjectedBundle::beginTesting):
2021
2022 2011-08-26  Alexey Proskuryakov  <ap@apple.com>
2023
2024         DumpRenderTree should begin each test with an empty cookie store
2025         https://bugs.webkit.org/show_bug.cgi?id=63545
2026         <rdar://problem/5666907>
2027
2028         Reviewed by Darin Adler.
2029
2030         * DumpRenderTree/mac/DumpRenderTree.mm:
2031         (setDefaultsToConsistentValuesForTesting):
2032         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
2033         (WTR::InjectedBundle::beginTesting):
2034         Tell WebCore to use a temporary session, starting with a clean cookie store.
2035
2036 2011-08-23  Chris Marrin  <cmarrin@apple.com>
2037
2038         [mac] requestAnimationFrame support for mac port
2039         https://bugs.webkit.org/show_bug.cgi?id=59146
2040
2041         Reviewed by Simon Fraser.
2042
2043         Add support for request-animation-callback flag
2044
2045         * Scripts/build-webkit:
2046
2047 2011-08-27  Peter Beverloo  <peter@chromium.org>
2048
2049         Add a build bot for Chromium with OS(ANDROID)
2050         https://bugs.webkit.org/show_bug.cgi?id=66687
2051
2052         Reviewed by Adam Barth.
2053
2054         * BuildSlaveSupport/build.webkit.org-config/config.json:
2055
2056 2011-08-27  Peter Beverloo  <peter@chromium.org>
2057
2058         Update the master build bot configuration script to properly
2059         deal with the --chromium-android switch.
2060         https://bugs.webkit.org/show_bug.cgi?id=66687
2061
2062         The --chromium-android switch is needed when running
2063         update-webkit-chromium and build-webkit, as it'll tell the
2064         scripts to pull in the right dependencies and cross-compile
2065         WebKit using the Android NDK.
2066
2067         Reviewed by Adam Barth.
2068
2069         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2070
2071 2011-08-27  Peter Beverloo  <peter@chromium.org>
2072
2073         Teach the update and build scripts how to handle dependencies
2074         and cross-compiling rules for the Chromium port on OS(ANDROID).
2075         https://bugs.webkit.org/show_bug.cgi?id=66687
2076
2077         Reviewed by Adam Barth.
2078
2079         * Scripts/build-webkit:
2080         * Scripts/update-webkit:
2081         * Scripts/update-webkit-chromium:
2082         * Scripts/webkitdirs.pm:
2083         (argumentsForConfiguration):
2084         (isChromium):
2085         (isChromiumAndroid):
2086         (determineIsChromiumAndroid):
2087         (buildChromium):
2088
2089 2011-08-26  Adam Barth  <abarth@webkit.org>
2090
2091         garden-o-matic should let me know that I need to run the local server
2092         https://bugs.webkit.org/show_bug.cgi?id=66985
2093
2094         Reviewed by Dimitri Glazkov.
2095
2096         The patch triggers an alert dialog whenever the user tries to use a
2097         feature that requires the local server.
2098
2099         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout.js:
2100         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout_unittests.js:
2101         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js:
2102
2103 2011-08-26  Dimitri Glazkov  <dglazkov@chromium.org>
2104
2105         garden-o-matic summary view should have pretty commit data information.
2106         https://bugs.webkit.org/show_bug.cgi?id=67069
2107
2108         Reviewed by Adam Barth.
2109
2110         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Added support for styling commit data.
2111         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Changed tests.
2112         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/common.css: Defined global anchor color.
2113         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css: Styled commit data.
2114
2115 2011-08-26  Dimitri Glazkov  <dglazkov@chromium.org>
2116
2117         Tweak the use of Open Sans in garden-o-matic summary view to make it look decent on Windows and Linux.
2118         https://bugs.webkit.org/show_bug.cgi?id=67066
2119
2120         Reviewed by Adam Barth.
2121
2122         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary-mock.html: Tweaked.
2123         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html: Ditto.
2124
2125 2011-08-26  Raphael Kubo da Costa  <kubo@profusion.mobi>
2126
2127         [DRT] Include cairo.h in PixelDumpSupportCairo.h for EFL too.
2128         https://bugs.webkit.org/show_bug.cgi?id=67041
2129
2130         Reviewed by Martin Robinson.
2131
2132         This is a follow-up to r89053: EFL's PixelDumpSupportEfl.cpp includes
2133         PixelDumpSupportCairo.h, but the latter was #including cairo.h only
2134         when the GTK+ port was being built.
2135
2136         * DumpRenderTree/cairo/PixelDumpSupportCairo.h:
2137
2138 2011-08-26  Dimitri Glazkov  <dglazkov@chromium.org>
2139
2140         Clean up and line up layout in garden-o-matic summary view.
2141         https://bugs.webkit.org/show_bug.cgi?id=67057
2142
2143         1) Renamed ui.notifications.TestFailures to ui.notifications.TestsFailing to match BuildersFailing;
2144         2) Rearranged DOM structure a bit to accommodate recent changes;
2145         3) Added minimal CSS to line things up.
2146
2147         Reviewed by Adam Barth.
2148
2149         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js: Renamed TestFailures to TestsFailing.
2150         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary-mock.js: Added mocks for BuildersFailing.
2151         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures.js: Renamed TestFailures to TestsFailing.
2152         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js: Updated tests.
2153         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Changed DOM structure.
2154         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Updated tests.
2155         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/common.css: Tweaked common appearance.
2156         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css: Added minimal layout/style.
2157         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary-mock.html: Added Google Web Fonts hook.
2158         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html: Ditto.
2159
2160 2011-08-26  Alice Liu  <alice.liu@apple.com>
2161
2162         https://bugs.webkit.org/show_bug.cgi?id=66823
2163         Add the request info to the WebPageProxy::CreateNewPage message
2164
2165         Reviewed by Anders Carlsson.
2166
2167         Update function declarations with additional parameter:
2168         * MiniBrowser/gtk/BrowserWindow.c:
2169         (createNewPage):
2170         * MiniBrowser/mac/BrowserWindowController.m:
2171         (createNewPage):
2172         * MiniBrowser/win/BrowserView.cpp:
2173         (createNewPage):
2174         * WebKitTestRunner/TestController.cpp:
2175         (WTR::TestController::createOtherPage):
2176         * WebKitTestRunner/TestController.h:
2177
2178 2011-08-26  Noel Gordon  <noel.gordon@gmail.com>
2179
2180         Make show-pretty-diff work on linux
2181         https://bugs.webkit.org/show_bug.cgi?id=67017
2182
2183         Reviewed by Adam Barth.
2184
2185         * Scripts/show-pretty-diff:
2186
2187 2011-08-25  Dimitri Glazkov  <dglazkov@chromium.org>
2188
2189         Reorganize garden-o-matic CSS files to match JS-built views.
2190         https://bugs.webkit.org/show_bug.cgi?id=66998
2191
2192         Reviewed by Adam Barth.
2193
2194         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/common.css: Added.
2195         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css: Added.
2196         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css: Added.
2197         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/summary.css: Removed.
2198         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary-mock.html: Changed to use new CSS files.
2199         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html: Ditto.
2200
2201 2011-08-25  Adam Barth  <abarth@webkit.org>
2202
2203         garden-o-matic should be able to roll out patches
2204         https://bugs.webkit.org/show_bug.cgi?id=66982
2205
2206         Reviewed by Dimitri Glazkov.
2207
2208         This patch just wires the rollout event through the controller to the
2209         local server, which is already set up to perform rollouts.
2210
2211         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout.js:
2212         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
2213         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
2214         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
2215
2216 2011-08-25  Viatcheslav Ostapenko  <ostapenko.viatcheslav@nokia.com>
2217
2218         [Qt] [WK2] Assert in computeViewportAttributes when MiniBrowser started with -touch
2219         https://bugs.webkit.org/show_bug.cgi?id=66951
2220
2221         Reviewed by Andreas Kling.
2222
2223         Function computeViewportAttributes() gets called when height is still 0 because MiniBrowser sets
2224         width and height one by one and setWidth cause geometryChange event in QTouchWebView.
2225
2226         * MiniBrowser/qt/BrowserView.cpp:
2227         (BrowserView::resizeEvent):
2228
2229 2011-08-25  Dimitri Glazkov  <dglazkov@chromium.org>
2230
2231         garden-o-matic summary view should provide information on builders still in flight.
2232         https://bugs.webkit.org/show_bug.cgi?id=66992
2233
2234         Reviewed by Adam Barth.
2235
2236         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js: Added grabbing the list of in-flight builders.
2237         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js: Added buildersInFlightForRevision.
2238         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model_unittests.js: Added tests.
2239         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Split off updating builder results to own method.
2240         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Added tests.
2241
2242 2011-08-25  Adam Barth  <abarth@webkit.org>
2243
2244         webkit-patch rebaseline-expectations don't work when run outside the root
2245         https://bugs.webkit.org/show_bug.cgi?id=66966
2246
2247         Reviewed by Eric Seidel.
2248
2249         Turns out tool.path() used to be relative, which caused problems when
2250         running webkit-patch outside the root directory.  This patch makes it
2251         absolute.  I checked all callers, and they all appear to want this
2252         behavior.
2253
2254         It was unclear to me how to test this change because we only really
2255         have a testing strategy for webkitpy, not for the main script.
2256
2257         * Scripts/webkit-patch:
2258
2259 2011-08-25  Dimitri Glazkov  <dglazkov@chromium.org>
2260
2261         garden-o-matic summary needs a mock view that could be used for working on styling
2262         https://bugs.webkit.org/show_bug.cgi?id=66971
2263
2264         Reviewed by Adam Barth.
2265
2266         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary-mock.js: Added.
2267         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary-mock.html: Added.
2268
2269 2011-08-25  Anders Carlsson  <andersca@apple.com>
2270
2271         Return the correct value for NPNVsupportsCarbonBool
2272         https://bugs.webkit.org/show_bug.cgi?id=66964
2273
2274         Add a test.
2275
2276         Reviewed by Adam Roben.
2277
2278         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
2279         * DumpRenderTree/TestNetscapePlugIn/Tests/mac/SupportsCarbonEventModel.cpp: Added.
2280         (SupportsCarbonEventModel::SupportsCarbonEventModel):
2281         (SupportsCarbonEventModel::testConvert):
2282         (SupportsCarbonEventModel::runTest):
2283         (SupportsCarbonEventModel::NPP_New):
2284
2285 2011-08-25  Dimitri Glazkov  <dglazkov@chromium.org>
2286
2287         garden-o-matic summary view should not be broken.
2288
2289         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/summary.css:
2290         (table.comparison): Added more specificity.
2291
2292 2011-08-25  Adam Barth  <abarth@webkit.org>
2293
2294         This command has been broken for a while.  This patch updates the
2295         callsite to the new API.
2296
2297         * Scripts/webkitpy/tool/commands/rebaseline.py:
2298
2299 2011-08-25  Dimitri Glazkov  <dglazkov@chromium.org>
2300
2301         garden-o-matic needs a way to report where and how tests are failing in the summary view.
2302         https://bugs.webkit.org/show_bug.cgi?id=66955
2303
2304         Introduce FailureGrid, your one-stop failure-indicating destination.
2305
2306         Reviewed by Adam Barth.
2307
2308         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html: Added new scripts.
2309         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures.js: Added.
2310         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js: Added.
2311         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Updated to use FailureGrid.
2312         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Updated tests.
2313         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html: Added new scripts.
2314
2315 2011-08-25  Igor Oliveira  <igor.oliveira@openbossa.org>
2316
2317         Add myself to the committers list
2318         https://bugs.webkit.org/show_bug.cgi?id=66941
2319
2320         Reviewed by Andreas Kling.
2321
2322         * Scripts/webkitpy/common/config/committers.py:
2323
2324 2011-08-25  Sheriff Bot  <webkit.review.bot@gmail.com>
2325
2326         Unreviewed, rolling out r93767.
2327         http://trac.webkit.org/changeset/93767
2328         https://bugs.webkit.org/show_bug.cgi?id=66931
2329
2330         Broke Qt bot. (Requested by yutak on #webkit).
2331
2332         * Scripts/webkitpy/thirdparty/__init__.py:
2333
2334 2011-08-25  Takashi Toyoshima  <toyoshim@chromium.org>
2335
2336         [WebSocket] update pywebsocket to 0.6b3 which supports closing handshake test
2337         https://bugs.webkit.org/show_bug.cgi?id=66924
2338
2339         Reviewed by Kent Tamura.
2340
2341         * Scripts/webkitpy/thirdparty/__init__.py:
2342
2343 2011-08-24  Adam Barth  <abarth@webkit.org>
2344
2345         The user can't close the details view in garden-o-matic
2346         https://bugs.webkit.org/show_bug.cgi?id=66911
2347
2348         Reviewed by Dimitri Glazkov.
2349
2350         In addition to adding a close button, this patch changes the test
2351         selector to use a <select> element and refactors the test selector to
2352         share code with the builder selector.
2353
2354         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
2355         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions.js:
2356         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions_unittests.js:
2357         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2358
2359 2011-08-24  James Robinson  <jamesr@chromium.org>
2360
2361         [chromium] Stacktrace not in test output when a test crashes
2362         https://bugs.webkit.org/show_bug.cgi?id=66806
2363
2364         Reviewed by Dirk Pranke.
2365
2366         Appends the .error and .text output when a test crashes since on chromium the .text contains the actual stack
2367         trace.
2368
2369         * Scripts/webkitpy/layout_tests/port/chromium.py:
2370
2371 2011-08-24  Adam Barth  <abarth@webkit.org>
2372
2373         Remove empty directory.
2374
2375         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui: Removed.
2376
2377 2011-08-24  Adam Barth  <abarth@webkit.org>
2378
2379         Finish writing up details-view in garden-o-matic
2380         https://bugs.webkit.org/show_bug.cgi?id=66891
2381
2382         Reviewed by Dimitri Glazkov.
2383
2384         This patch mostly just copies a bunch of CSS from main.css to
2385         summary.css.  Presumably we'll delete main.css soon.
2386
2387         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
2388         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions.js:
2389         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions_unittests.js:
2390         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/summary.css:
2391         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html:
2392
2393 2011-08-24  James Robinson  <jamesr@chromium.org>
2394
2395         Some compositing LayoutTests flakily crashing on Mac 10.6
2396         https://bugs.webkit.org/show_bug.cgi?id=66740
2397
2398         Reviewed by Adam Barth.
2399
2400         Null-check obj->pluginTest before dereferencing it. This can be null in chromium if the plugin load fails, for
2401         example if the test attempts to load a PDF plugin the load will fail since we don't have such a thing in the
2402         chromium configuration.
2403
2404         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
2405         (NPP_Destroy):
2406
2407 2011-08-24  Anton D'Auria  <adauria@apple.com>
2408
2409         No review necessary.
2410
2411         Adding myself to list of committers.
2412
2413         * Scripts/webkitpy/common/config/committers.py:
2414
2415 2011-08-24  Avi Drissman  <avi@google.com>
2416
2417         Crash when restoring NSGraphicsContext when running the Chromium GPU layout tests on Mac
2418         https://bugs.webkit.org/show_bug.cgi?id=66875
2419
2420         Reviewed by James Robinson.
2421
2422         * DumpRenderTree/chromium/WebThemeEngineDRTMac.mm:
2423         (WebThemeEngineDRTMac::paintNSScrollerScrollbarThumb):
2424
2425 2011-08-23  Dimitri Glazkov  <dglazkov@chromium.org>
2426
2427         Make notifications appear in time order again.
2428         https://bugs.webkit.org/show_bug.cgi?id=66815
2429
2430         Reviewed by Adam Barth.
2431
2432         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Added setting index of the notification.
2433         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Updated the test.
2434
2435 2011-08-23  James Robinson  <jamesr@chromium.org>
2436
2437         [chromium] Need a way to test lost compositor context recovery
2438         https://bugs.webkit.org/show_bug.cgi?id=66820
2439
2440         Reviewed by Kenneth Russell.
2441
2442         Exposes a LayoutTestController interface in chromium to simulate a
2443         lost compositor context.
2444
2445         * DumpRenderTree/chromium/LayoutTestController.cpp:
2446         (LayoutTestController::LayoutTestController):
2447         (LayoutTestController::loseCompositorContext):
2448         * DumpRenderTree/chromium/LayoutTestController.h:
2449
2450 2011-08-23  Dimitri Glazkov  <dglazkov@chromium.org>
2451
2452         Change garden-o-matic frontend builder array to be a configuration map.
2453         https://bugs.webkit.org/show_bug.cgi?id=66816
2454
2455         Reviewed by Adam Barth.
2456
2457         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js: Made kBuilders a map, with value being a configuration object.
2458         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js: Adjusted callsites to expect a map.
2459         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/main.js: Ditto.
2460         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js: Ditto.
2461         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js: Ditto.
2462
2463 2011-08-23  Cary Clark  <caryclark@google.com>
2464
2465         Turn WebCanvas into CGContext before drawing scrollbar (DRT using Skia on Mac)
2466         https://bugs.webkit.org/show_bug.cgi?id=66803
2467
2468         This fixes the WebKit Canaries, in particular avoids a crash on tests
2469         like css1/basic/comments.html.
2470
2471         Reviewed by James Robinson.
2472
2473         * DumpRenderTree/chromium/WebThemeEngineDRTMac.mm:
2474         (WebThemeEngineDRTMac::paintNSScrollerScrollbarThumb):
2475         The NSGraphicsContext call graphicsContextWithGraphicsPort takes a void*
2476         parameter, so it took a WebCanvas* (mapped to SkCanvas* in Skia-land)
2477         and treated it as if it were a CGContextRef.
2478
2479 2011-08-23  Patrick Gansterer  <paroga@webkit.org>
2480
2481         Fix mergeChangeLogs test on native windows
2482         https://bugs.webkit.org/show_bug.cgi?id=66797
2483
2484         Reviewed by Adam Roben.
2485
2486         Windows has no /tmp/ directory and the "TMPDIR" environment variable isn't set in the default setup.
2487         Add the "TEMP" environment variable as additional source for the temporary directory used in writeTempFile.
2488
2489         * Scripts/webkitperl/VCSUtils_unittest/mergeChangeLogs.pl:
2490         (writeTempFile):
2491
2492 2011-08-23  Dimitri Glazkov  <dglazkov@chromium.org>
2493
2494         garden-o-matic relative time is wrong and updates too frequently.
2495         https://bugs.webkit.org/show_bug.cgi?id=66802
2496
2497         Reviewed by Adam Barth.
2498
2499         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js: Changed to account for rounding correctly.
2500         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js: Updated tests.
2501         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js: Upped the interval to 1 minute.
2502
2503 2011-08-23  Patrick Gansterer  <paroga@webkit.org>
2504
2505         Add missing quotes around filenames in mergeChangeLogs
2506         https://bugs.webkit.org/show_bug.cgi?id=66794
2507
2508         Reviewed by Adam Roben.
2509
2510         * Scripts/VCSUtils.pm:
2511         (mergeChangeLogs):
2512
2513 2011-08-23  Patrick Gansterer  <paroga@webkit.org>
2514
2515         webkit-perl tests fail on win32 Perl due to lack of list form of pipe open implementation
2516         https://bugs.webkit.org/show_bug.cgi?id=49080
2517
2518         Reviewed by Adam Roben.
2519
2520         Use the string form of pipe open instead (like we do at all other places).
2521
2522         * Scripts/VCSUtils.pm:
2523         (mergeChangeLogs):
2524
2525 2011-08-23  Adam Barth  <abarth@webkit.org>
2526
2527         garden-o-matic summary should be able to examine failures
2528         https://bugs.webkit.org/show_bug.cgi?id=66748
2529
2530         Reviewed by Dimitri Glazkov.
2531
2532         This patch refactors a bunch of code from summary.js to independent
2533         controllers in controllers.js.  This patch also contains the first
2534         iteration of binding the "examine" event to a the behavior showing the
2535         details view for the tests.  There's more work to do here to make this
2536         work properly (and elegantly), but this patch is a start.
2537
2538         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
2539         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js:
2540         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
2541         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers/rebaseline.js:
2542         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js:
2543         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
2544         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
2545         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
2546         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html:
2547
2548 2011-08-23  Adam Barth  <abarth@webkit.org>
2549
2550         webkitpy/common/config/build.py shouldn't mention non-existant Android directory
2551         https://bugs.webkit.org/show_bug.cgi?id=66737
2552
2553         Reviewed by Steve Block.
2554
2555         This directory doesn't exist anymore now that the Android port shares
2556         more code with Chromium.
2557
2558         * Scripts/webkitpy/common/config/build.py:
2559
2560 2011-08-22  Eric Seidel  <eric@webkit.org>
2561
2562         Hack start-queue.sh in an attempt to resurrect the cr-jail-1 cr-mac-ews bot.
2563         I won't have ssh access to the bot again for a couple days, so
2564         attempting to fix by changing this script.  If this doesn't
2565         work we'll just live with the it hung for now.
2566
2567         Unreviewed.
2568
2569         * EWSTools/start-queue.sh:
2570
2571 2011-08-22  Dimitri Glazkov  <dglazkov@chromium.org>
2572
2573         webkit-patch rebaseline-expectations command should not rearrange the entire test_expectations.txt file.
2574         https://bugs.webkit.org/show_bug.cgi?id=66727
2575
2576         Reviewed by Adam Barth.
2577
2578         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Made it possible to run TestExpectationSerializer without a TestConfigurationConverter,
2579             in which case the serializer treats all expectations as unparsed.
2580         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added unit tests.
2581
2582 2011-08-22  Dimitri Glazkov  <dglazkov@chromium.org>
2583
2584         Update unit test expectations after r93530.
2585
2586         * Scripts/webkitpy/layout_tests/port/factory_unittest.py: Updated.
2587         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py: Ditto.
2588
2589 2011-08-22  Dimitri Glazkov  <dglazkov@chromium.org>
2590
2591         [Chromium] Expand CPU/GPU/Skia/CG into a matrix in graphics_type.
2592         https://bugs.webkit.org/show_bug.cgi?id=66705
2593
2594         Reviewed by Adam Barth.
2595
2596         * Scripts/webkitpy/layout_tests/port/chromium.py: Added plumbing for CG graphics types.
2597         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Adjusted unit test.
2598
2599 2011-08-22  James Robinson  <jamesr@chromium.org>
2600
2601         [chromium] Update builder names for chromium gpu bots due to CG switchover
2602         https://bugs.webkit.org/show_bug.cgi?id=66693
2603
2604         Reviewed by Dirk Pranke.
2605
2606         Update the port and builder names for the chromium mac gpu bots to
2607         account for the CG transition.
2608
2609         * Scripts/webkitpy/layout_tests/port/builders.py:
2610
2611 2011-08-22  Dirk Pranke  <dpranke@chromium.org>
2612
2613         fix the GPU bot names on the layout test dashboard
2614         https://bugs.webkit.org/show_bug.cgi?id=66697
2615
2616         Reviewed by Dimitri Glazkov.
2617
2618         * TestResultServer/static-dashboards/builders.js:
2619
2620 2011-08-22  Dimitri Glazkov  <dglazkov@chromium.org>
2621
2622         garden-o-matic's relative time display should be updated regularly.
2623         https://bugs.webkit.org/show_bug.cgi?id=66682
2624
2625         Reviewed by Adam Barth.
2626
2627         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js: Moved Time to ui.Time, added code for regular updates.
2628         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Changed to use ui.Time.
2629         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js: Added tests.
2630
2631 2011-08-22  Xianzhu Wang  <wangxianzhu@chromium.org>
2632
2633         Pass --makeargs of build-webkit to chromium linux builder
2634         https://bugs.webkit.org/show_bug.cgi?id=66361
2635
2636         Reviewed by Tony Chang.
2637
2638         * Scripts/build-webkit:
2639         * Scripts/webkitdirs.pm:
2640         (buildChromiumMakefile):
2641         (buildChromium):
2642
2643 2011-08-22  Peter Beverloo  <peter@chromium.org>
2644
2645         Add myself to the contributor list.
2646         https://bugs.webkit.org/show_bug.cgi?id=66674
2647
2648         Reviewed by Tony Gentilcore.
2649
2650         * Scripts/webkitpy/common/config/committers.py:
2651
2652 2011-08-22  Balazs Kelemen  <kbalazs@webkit.org>
2653
2654         [Qt] ImageDiff does not consider --tolerance
2655         https://bugs.webkit.org/show_bug.cgi?id=66299
2656
2657         Reviewed by Csaba Osztrogonác.
2658
2659         * DumpRenderTree/qt/ImageDiff.cpp:
2660         (main):
2661
2662 2011-08-22  Csaba Osztrogonác  <ossy@webkit.org>
2663
2664         Update contributor list.
2665
2666         Reviewed by Zoltan Herczeg.
2667
2668         * Scripts/webkitpy/common/config/committers.py:
2669
2670 2011-08-22  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
2671
2672         [ORWT] Fix --reset-results --add-platform-exceptions combo
2673         https://bugs.webkit.org/show_bug.cgi?id=65464
2674
2675         Reviewed by Csaba Osztrogonác.
2676
2677         * Scripts/old-run-webkit-tests:
2678
2679 2011-08-20  Adam Barth  <abarth@webkit.org>
2680
2681         garden-o-matic summary.html should have UI for examining failures
2682         https://bugs.webkit.org/show_bug.cgi?id=66625
2683
2684         Reviewed by Dimitri Glazkov.
2685
2686         This patch introduces the notion of UI actions and converts our
2687         existing manual <button> manipulation with this new abstraction.  The
2688         patch also adds an Examine button to failure notifications.  In a
2689         future patch, I'll add a controller that wires this button up to
2690         something useful.
2691
2692         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
2693         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
2694         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions.js: Added.
2695         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions_unittests.js: Added.
2696         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
2697         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
2698         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2699         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html:
2700
2701 2011-08-20  Adam Barth  <abarth@webkit.org>
2702
2703         garden-o-matic summary page should indiciate compile failures
2704         https://bugs.webkit.org/show_bug.cgi?id=66624
2705
2706         Reviewed by Dimitri Glazkov.
2707
2708         This patch creates a notification-based UI for build failures.  The new
2709         UI displays the same information that used to be displayed in the
2710         infobar.  There's lots of room for improvement in surfacing more
2711         information, but this is a start.
2712
2713         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js:
2714         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
2715         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
2716
2717 2011-08-19  James Robinson  <jamesr@chromium.org>
2718
2719         [chromium] Default to experimental-fully-parallel mode for chromium gpu port
2720         https://bugs.webkit.org/show_bug.cgi?id=66606
2721
2722         Reviewed by Dirk Pranke.
2723
2724         The chromium GPU configuration does not (currently) contain any tests
2725         incompatible with --experimental-fully-parallel, and this flag is a
2726         significan speedup because of the directory canvas/philip/tests which
2727         contains 802 tests in the same directory. Setting this option by
2728         default on this configuration is a 2x speed in release, 44s to 22s,
2729         and a 3x speedup in debug, 223s to 77s, on linux on a 16 core machine.
2730
2731         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2732         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
2733         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2734
2735 2011-08-19  Sheriff Bot  <webkit.review.bot@gmail.com>
2736
2737         Unreviewed, rolling out r93426.
2738         http://trac.webkit.org/changeset/93426
2739         https://bugs.webkit.org/show_bug.cgi?id=66607
2740
2741         Broke the Mac build by not building gtest.framework anymore
2742         (Requested by andersca on #webkit).
2743
2744         * TestWebKitAPI/Tests/WTF/StringOperators.cpp:
2745
2746 2011-08-19  James Robinson  <jamesr@chromium.org>
2747
2748         [chromium] Chromium GPU layout tests should use the normal process number selection, not force 1 worker
2749         https://bugs.webkit.org/show_bug.cgi?id=66603
2750
2751         Reviewed by Dirk Pranke.
2752
2753         The various ChromiumGpu ports should use the same child worker
2754         selection logic as other ports so they can run with full
2755         parallelization on multicore systems.
2756
2757         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2758         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
2759
2760 2011-08-19  Alok Priyadarshi  <alokp@chromium.org>
2761
2762         [chromium] Remove dependency of WebThemeControlDRTWin on skia::PlatformCanvas
2763         https://bugs.webkit.org/show_bug.cgi?id=66570
2764
2765         Reviewed by Mihai Parparita.
2766
2767         * DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
2768         (WebThemeControlDRTWin::draw):
2769         (WebThemeControlDRTWin::drawTextField):
2770         (WebThemeControlDRTWin::drawProgressBar):
2771         * DumpRenderTree/chromium/WebThemeControlDRTWin.h:
2772
2773 2011-08-18  Dmitry Lomov  <dslomov@google.com>
2774
2775         TestWebKitAPI breaks on Windows due to fast malloc incompatibility.
2776         https://bugs.webkit.org/show_bug.cgi?id=66521
2777
2778         This patch ensures that gtest uses new and delete operators that are defined in JavaScriptCore.
2779
2780         Reviewed by David Levin.
2781
2782         * TestWebKitAPI/Tests/WTF/StringOperators.cpp: Removed redefinition of JS_EXPORTDATA.
2783
2784 2011-08-19  Adam Roben  <aroben@apple.com>
2785
2786         Add project file changes I left out in r93417
2787
2788         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2789
2790 2011-08-19  Adam Roben  <aroben@apple.com>
2791
2792         Add a helper class for swizzling ObjC instance methods to TestWebKitAPI
2793
2794         Fixes <http://webkit.org/b/66565> TestWebKitAPI needs a way to override instance methods of
2795         NSScreen
2796
2797         Reviewed by Anders Carlsson.
2798
2799         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added new files.
2800
2801         * TestWebKitAPI/Tests/TestWebKitAPI/mac/InstanceMethodSwizzler.mm: Added.
2802         (-[SimpleObject setValue:]): Original implementation.
2803
2804         (TestWebKitAPI::setValue2):
2805         (TestWebKitAPI::setValue3):
2806         Swizzled implementations.
2807
2808         (TestWebKitAPI::TEST): Test that we can swizzle and re-swizzle and un-swizzle an ObjC
2809         instance method.
2810
2811         * TestWebKitAPI/mac/InstanceMethodSwizzler.h: Added.
2812         * TestWebKitAPI/mac/InstanceMethodSwizzler.mm: Added.
2813         (TestWebKitAPI::InstanceMethodSwizzler::InstanceMethodSwizzler): Swizzle the method.
2814         (TestWebKitAPI::InstanceMethodSwizzler::~InstanceMethodSwizzler): Unswizzle the method.
2815
2816 2011-08-19  Tom Zakrajsek  <tomz@codeaurora.org>
2817
2818         Fix _path_to_apache() error reporting
2819         https://bugs.webkit.org/show_bug.cgi?id=66486
2820
2821         Reviewed by Ryosuke Niwa.
2822
2823         * Scripts/webkitpy/layout_tests/port/webkit.py:
2824
2825 2011-08-19  Adam Roben  <aroben@apple.com>
2826
2827         Windows build fix after r93404
2828
2829         * TestWebKitAPI/Tests/WebKit2/win/HideFindIndicator.cpp: Added missing #include.
2830
2831 2011-08-19  Tony Gentilcore  <tonyg@chromium.org>
2832
2833         Add leandrogracia to the committers list
2834         https://bugs.webkit.org/show_bug.cgi?id=66552
2835
2836         Also fix line endings in file.
2837
2838         No reviewed required.
2839
2840         * Scripts/webkitpy/common/config/committers.py:
2841
2842 2011-08-19  Amruth Raj  <amruthraj@motorola.com>
2843
2844         Fix compilation issue in WebKitTestRunner
2845         https://bugs.webkit.org/show_bug.cgi?id=66554
2846
2847         Reviewed by Martin Robinson.
2848
2849         * WebKitTestRunner/PixelDumpSupport.cpp:
2850
2851 2011-08-19  Adam Roben  <aroben@apple.com>
2852
2853         Pull some of DynamicDeviceScaleFactor's code up into a shareable base class
2854
2855         This will make it easier to write other WebKit-agnostic tests
2856
2857         Fixes <http://webkit.org/b/66558> Would like to be able to reuse DynamicDeviceScaleFactor's
2858         code in other WebKit-agnostic tests
2859
2860         Reviewed by Anders Carlsson.
2861
2862         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added new files.
2863
2864         * TestWebKitAPI/Tests/mac/DynamicDeviceScaleFactor.mm: Fixed copyright. Moved a bunch of
2865         code to WebKitAgnosticTest.h/mm. Changed to inherit from WebKitAgnosticTest.
2866         (TestWebKitAPI::DynamicDeviceScaleFactor::url): Simple getter.
2867         (TestWebKitAPI::DynamicDeviceScaleFactor::didLoadURL): Calls through to runTest.
2868         (TestWebKitAPI::DynamicDeviceScaleFactor::runTest): Removed loading code which
2869         WebKitAgnosticTest handles for us.
2870         (TestWebKitAPI::TEST_F): Removed view-creation code which WebKitAgnosticTest handles for us.
2871
2872         * TestWebKitAPI/mac/WebKitAgnosticTest.h: Added. Code came from DynamicDeviceScaleFactor.
2873         (TestWebKitAPI::WebKitAgnosticTest::loadAndWaitUntilDone): Code came from
2874         DynamicDeviceScaleFactor::runTest.
2875
2876         * TestWebKitAPI/mac/WebKitAgnosticTest.mm: Copied from Tools/TestWebKitAPI/Tests/mac/DynamicDeviceScaleFactor.mm.
2877         (TestWebKitAPI::WebKitAgnosticTest::runWebKit1Test): Code came from
2878         DynamicDeviceScaleFactor.WebKit.
2879         (TestWebKitAPI::WebKitAgnosticTest::runWebKit2Test): Code came from
2880         DynamicDeviceScaleFactor.WebKit2.
2881
2882 2011-08-19  Adam Roben  <aroben@apple.com>
2883
2884         Clean up #includes in TestWebKitAPI
2885
2886         Fixes <http://webkit.org/b/66556> TestWebKitAPI's prefix header isn't very helpful
2887
2888         Reviewed by Anders Carlsson.
2889
2890         * TestWebKitAPI/TestWebKitAPIPrefix.h: Added wtf/Platform.h. Added gtest/gtest.h for C++
2891         files and WebKit/WebKit.h for ObjC files.
2892
2893         * TestWebKitAPI/InjectedBundleController.cpp:
2894         * TestWebKitAPI/JavaScriptTest.h:
2895         * TestWebKitAPI/PlatformUtilities.h:
2896         * TestWebKitAPI/PlatformWebView.h:
2897         * TestWebKitAPI/Test.h:
2898         * TestWebKitAPI/Tests/WTF/StringOperators.cpp:
2899         * TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
2900         * TestWebKitAPI/Tests/WTF/VectorReverse.cpp:
2901         * TestWebKitAPI/Tests/WebKit2/AboutBlankLoad.cpp:
2902         * TestWebKitAPI/Tests/WebKit2/CanHandleRequest.cpp:
2903         * TestWebKitAPI/Tests/WebKit2/CookieManager.cpp:
2904         * TestWebKitAPI/Tests/WebKit2/DocumentStartUserScriptAlertCrash.cpp:
2905         * TestWebKitAPI/Tests/WebKit2/DocumentStartUserScriptAlertCrash_Bundle.cpp:
2906         * TestWebKitAPI/Tests/WebKit2/DownloadDecideDestinationCrash.cpp:
2907         * TestWebKitAPI/Tests/WebKit2/EvaluateJavaScript.cpp:
2908         * TestWebKitAPI/Tests/WebKit2/FailedLoad.cpp:
2909         * TestWebKitAPI/Tests/WebKit2/Find.cpp:
2910         * TestWebKitAPI/Tests/WebKit2/ForceRepaint.cpp:
2911         * TestWebKitAPI/Tests/WebKit2/FrameMIMETypeHTML.cpp:
2912         * TestWebKitAPI/Tests/WebKit2/FrameMIMETypePNG.cpp:
2913         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle.cpp:
2914         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle_Bundle.cpp:
2915         * TestWebKitAPI/Tests/WebKit2/InjectedBundleBasic.cpp:
2916         * TestWebKitAPI/Tests/WebKit2/InjectedBundleBasic_Bundle.cpp:
2917         * TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash.cpp:
2918         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
2919         * TestWebKitAPI/Tests/WebKit2/PageLoadDidChangeLocationWithinPageForFrame.cpp:
2920         * TestWebKitAPI/Tests/WebKit2/PreventEmptyUserAgent.cpp:
2921         * TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly.cpp:
2922         * TestWebKitAPI/Tests/WebKit2/RestoreSessionStateContainingFormData.cpp:
2923         * TestWebKitAPI/Tests/WebKit2/SpacebarScrolling.cpp:
2924         * TestWebKitAPI/Tests/WebKit2/WKPreferences.cpp:
2925         * TestWebKitAPI/Tests/WebKit2/WKString.cpp:
2926         * TestWebKitAPI/Tests/WebKit2/WKStringJSString.cpp:
2927         * TestWebKitAPI/Tests/WebKit2/win/AltKeyGeneratesWMSysCommand.cpp:
2928         * TestWebKitAPI/Tests/WebKit2/win/DoNotCopyANullCFURLResponse.cpp:
2929         * TestWebKitAPI/Tests/WebKit2/win/HideFindIndicator.cpp:
2930         * TestWebKitAPI/Tests/WebKit2/win/ResizeViewWhileHidden.cpp:
2931         * TestWebKitAPI/Tests/WebKit2/win/TranslateMessageGeneratesWMChar.cpp:
2932         * TestWebKitAPI/Tests/WebKit2/win/WMCloseCallsUIClientClose.cpp:
2933         * TestWebKitAPI/Tests/WebKit2/win/WMPrint.cpp:
2934         * TestWebKitAPI/Tests/mac/DynamicDeviceScaleFactor.mm:
2935         * TestWebKitAPI/TestsController.cpp:
2936         * TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
2937         Removed now-unnecessary #includes. Moved #includes of Test.h in with the rest of the
2938         #includes since it is no longer a special header.
2939
2940 2011-08-18  Dirk Pranke  <dpranke@chromium.org>
2941
2942         new-run-webkit-tests hung while acquiring http lock on snow leopard bots
2943         https://bugs.webkit.org/show_bug.cgi?id=64886
2944
2945         Temporarily disable the http locking to work around the issue.
2946         I'm not actually sure if this is going to work or improve things
2947         much.
2948
2949         Reviewed by Ryosuke Niwa.
2950
2951         * Scripts/webkitpy/layout_tests/port/mac.py:
2952
2953 2011-08-18  Tony Chang  <tony@chromium.org>
2954
2955         add embedded png checksums to WebKitTestRunner
2956         https://bugs.webkit.org/show_bug.cgi?id=66494
2957
2958         Looks like WebKitTestRunner has never supported embedded checksums.  This copies
2959         some code from DRT and adds it to the WebKitTestRunner.
2960
2961         Reviewed by Darin Adler.
2962
2963         * WebKitTestRunner/CyclicRedundancyCheck.cpp: Copied from Tools/DumpRenderTree
2964         * WebKitTestRunner/CyclicRedundancyCheck.h: Copied from Tools/DumpRenderTree
2965         * WebKitTestRunner/GNUmakefile.am: Add new files
2966         * WebKitTestRunner/PixelDumpSupport.cpp: Copied from Tools/DumpRenderTree
2967         * WebKitTestRunner/PixelDumpSupport.h: Copied from Tools/DumpRenderTree
2968         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj: Add new files
2969         * WebKitTestRunner/cairo/TestInvocationCairo.cpp:
2970         (WTR::dumpBitmap): Refactor to use PixelDumpSupport.
2971         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
2972         * WebKitTestRunner/cg/TestInvocationCG.cpp:
2973         (WTR::dumpBitmap): Refactor to use PixelDumpSupport.
2974         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
2975         * WebKitTestRunner/win/WebKitTestRunner.vcproj: Add new files
2976
2977 2011-08-18  Shawn Singh  <shawnsingh@chromium.org>
2978
2979         https://bugs.webkit.org/show_bug.cgi?id=47240
2980
2981         Fixed a cygwin path problem in the chromium port of diff_image;
2982         Also made the return values of the diff_image function more consistent.
2983
2984         Reviewed by Dirk Pranke.
2985
2986         * Scripts/webkitpy/layout_tests/port/chromium.py:
2987
2988 2011-08-17  Alejandro G. Castro  <alex@igalia.com>
2989
2990         [GTK] Fix compilation problems with deprecations in gtk+
2991         https://bugs.webkit.org/show_bug.cgi?id=66073
2992
2993         Reviewed by Martin Robinson.
2994
2995         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2996         (main): G_CONST_RETURN was deprecated
2997         * GtkLauncher/main.c:
2998         (createWindow): Added gtk_box_new conditional compilation for
2999         gtk+-3.
3000         * MiniBrowser/gtk/BrowserWindow.c:
3001         (browser_window_init): Replaced gtk_vbox_new with gtk_box_new, we
3002         are just supporting gtk+-3 for WebKit2.
3003
3004 2011-08-18  Anders Carlsson  <andersca@apple.com>
3005
3006         Fix libc++ C++0x build
3007         https://bugs.webkit.org/show_bug.cgi?id=66479
3008
3009         Reviewed by Adam Roben.
3010
3011         Add missing includes.
3012
3013         * DumpRenderTree/LayoutTestController.cpp:
3014         * DumpRenderTree/mac/CheckedMalloc.cpp:
3015
3016 2011-08-18  Adam Roben  <aroben@apple.com>
3017
3018         Test that WebKit updates style when a WebView is moved between differently-scaled windows
3019
3020         Test for <http://webkit.org/b/66412> <rdar://problem/9971958> WebKit doesn't react when a
3021         WebView is moved between windows with different backing scale factors
3022
3023         Reviewed by Anders Carlsson.
3024
3025         * TestWebKitAPI/JavaScriptTest.cpp:
3026         (TestWebKitAPI::runJSTest): Moved a little bit of code from here...
3027         (TestWebKitAPI::compareJSResult): ...to here. Also made the error message more similar to
3028         gtest's built-in error messages.
3029
3030         * TestWebKitAPI/JavaScriptTest.h: Added overloads of runJSTest that take a WebView * and
3031         WKView * for convenience on Mac. Added compareJSResult helper function for implementing
3032         those overloads.
3033
3034         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added new files.
3035
3036         * TestWebKitAPI/Tests/mac/DynamicDeviceScaleFactor.mm: Added.
3037         (-[FrameLoadDelegate initWithDidFinishLoadBoolean:]): Simple initializer.
3038         (-[FrameLoadDelegate webView:didFinishLoadForFrame:]): Record that the load finished.
3039         (TestWebKitAPI::didFinishLoadForFrame): Record that the load finished.
3040         (TestWebKitAPI::setPageLoaderClient): Set up the client.
3041         (TestWebKitAPI::DynamicDeviceScaleFactor::DynamicDeviceScaleFactor): Simple constructor.
3042         (TestWebKitAPI::DynamicDeviceScaleFactor::createWindow): Creates a
3043         SyntheticBackingScaleFactorWindow and returns it.
3044         (TestWebKitAPI::DynamicDeviceScaleFactor::runTest): Loads devicePixelRatio.html and checks
3045         that WebKit uses the correct scale factor when the WebView is not in a window, is put in a
3046         window, is moved to a differently-scaled window, and is taken out of the window.
3047         (TestWebKitAPI::DynamicDeviceScaleFactor::loadURL): Helper function with overloads for
3048         WebKit1 and WebKit2.
3049         (TestWebKitAPI::TEST_F): Runs the test, with overloads for WebKit1 and WebKit2.
3050
3051         * TestWebKitAPI/Tests/mac/devicePixelRatio.html: Added.
3052
3053         * TestWebKitAPI/mac/JavaScriptTestMac.mm: Added.
3054         (TestWebKitAPI::runJSTest): Fairly simple overloads for WebView * and WKView *.
3055
3056         * TestWebKitAPI/mac/SyntheticBackingScaleFactorWindow.h: Added.
3057         * TestWebKitAPI/mac/SyntheticBackingScaleFactorWindow.m: Added.
3058         (-[SyntheticBackingScaleFactorWindow initWithContentRect:styleMask:backing:defer:]): Simple
3059         initializer.
3060         (-[SyntheticBackingScaleFactorWindow setBackingScaleFactor:]): Simple setter.
3061
3062         (-[SyntheticBackingScaleFactorWindow backingScaleFactor]):
3063         (-[SyntheticBackingScaleFactorWindow userSpaceScaleFactor]):
3064         Overrides of NSWindow methods that WebKit uses to determine the device scale factor.
3065
3066 2011-08-18  Dmitry Lomov  <dslomov@google.com>
3067
3068         https://bugs.webkit.org/show_bug.cgi?id=66425
3069         check-webkit-style should detect incorrectly formatted file change descriptions.
3070         Description, if present, should be separated from colon with a space.
3071
3072         Reviewed by David Levin.
3073
3074         * Scripts/webkitpy/style/checkers/changelog.py: Fix.
3075         * Scripts/webkitpy/style/checkers/changelog_unittest.py: Unit tests.
3076
3077 2011-08-18  Adam Barth  <abarth@webkit.org>
3078
3079         webkit-patch pretty-diff should be shown in main help
3080         https://bugs.webkit.org/show_bug.cgi?id=66404
3081
3082         Reviewed by Dimitri Glazkov.
3083
3084         As requested by dglazkov.
3085
3086         * Scripts/webkitpy/tool/commands/prettydiff.py:
3087
3088 2011-08-17  Dmitry Lomov  <dslomov@google.com>
3089
3090         https://bugs.webkit.org/show_bug.cgi?id=66400
3091         Fix TestWebKitAPI build on Mac OS.
3092
3093         Reviewed by David Levin.
3094
3095         * TestWebKitAPI/Configurations/Base.xcconfig: Added ICU headers to the HEADER_SEARCH_PATH.
3096
3097 2011-08-17  Dimitri Glazkov  <dglazkov@chromium.org>
3098
3099         garden-o-matic spring cleaning: move scripts, styles, and images in the respective directories.
3100         https://bugs.webkit.org/show_bug.cgi?id=66427
3101
3102         Reviewed by Adam Barth.
3103
3104         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js: Removed.
3105         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js: Removed.
3106         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot_unittests.js: Removed.
3107         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js: Removed.
3108         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js: Removed.
3109         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm.js: Removed.
3110         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js: Removed.
3111         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js: Removed.
3112         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector_unittests.js: Removed.
3113         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm.js: Removed.
3114         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm_unittests.js: Removed.
3115         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js: Removed.
3116         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js: Removed.
3117         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NRWTResultsParser.js: Removed.
3118         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm.js: Removed.
3119         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm_unittests.js: Removed.
3120         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ORWTResultsParser.js: Removed.
3121         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/PersistentCache.js: Removed.
3122         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css: Removed.
3123         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm.js: Removed.
3124         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm_unittests.js: Removed.
3125         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js: Removed.
3126         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js: Removed.
3127         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities_unittests.js: Removed.
3128         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js: Removed.
3129         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBugzilla.js: Removed.
3130         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBuildbot.js: Removed.
3131         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js: Removed.
3132         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js: Removed.
3133         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/bugzilla_unittests.js: Removed.
3134         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders.js: Removed.
3135         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders_unittests.js: Removed.
3136         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js: Removed.
3137         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout_unittests.js: Removed.
3138         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js: Removed.
3139         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/controllers: Removed.
3140         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/controllers.js: Removed.
3141         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/controllers/rebaseline.js: Removed.
3142         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon-green.png: Removed.
3143         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon-red.png: Removed.
3144         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
3145         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/images: Added.
3146         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/images/favicon-green.png: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon-green.png.
3147         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/images/favicon-red.png: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon-red.png.
3148         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html:
3149         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css: Removed.
3150         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js: Removed.
3151         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js: Removed.
3152         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model_unittests.js: Removed.
3153         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net.js: Removed.
3154         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net_unittests.js: Removed.
3155         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/rebaseline.html:
3156         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js: Removed.
3157         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js: Removed.
3158         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
3159         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Bugzilla.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js.
3160         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Buildbot.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js.
3161         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Buildbot_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot_unittests.js.
3162         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Builder.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js.
3163         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Builder_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js.
3164         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/FailingTestsBugForm.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm.js.
3165         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/FailingTestsBugForm_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js.
3166         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/FlakyLayoutTestDetector.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js.
3167         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/FlakyLayoutTestDetector_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector_unittests.js.
3168         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/FlakyTestBugForm.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm.js.
3169         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/FlakyTestBugForm_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm_unittests.js.
3170         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/LayoutTestHistoryAnalyzer.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js.
3171         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/LayoutTestResultsLoader.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js.
3172         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/NRWTResultsParser.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NRWTResultsParser.js.
3173         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/NewBugForm.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm.js.
3174         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/NewBugForm_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm_unittests.js.
3175         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ORWTResultsParser.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ORWTResultsParser.js.
3176         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/PersistentCache.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/PersistentCache.js.
3177         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/TestRelatedBugForm.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm.js.
3178         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/TestRelatedBugForm_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm_unittests.js.
3179         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Trac.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js.
3180         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Utilities.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js.
3181         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Utilities_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities_unittests.js.
3182         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ViewController.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js.
3183         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/WebKitBugzilla.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBugzilla.js.
3184         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/WebKitBuildbot.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBuildbot.js.
3185         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js.
3186         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js.
3187         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/bugzilla_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/bugzilla_unittests.js.
3188         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders.js.
3189         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders_unittests.js.
3190         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js.
3191         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout_unittests.js.
3192         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js.
3193         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/controllers.
3194         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/controllers.js.
3195         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/main.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js.
3196         ():
3197         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js.
3198         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model_unittests.js.
3199         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/net.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net.js.
3200         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/net_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net_unittests.js.
3201         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js.
3202         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js.
3203         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/trac_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/trac_unittests.js.
3204         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js.
3205         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui/results.js.
3206         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui/results_unittests.js.
3207         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js.
3208         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/unexpected-passes.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/unexpected-passes.js.
3209         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/TestFailures.css: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css.
3210         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/main.css: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css.
3211         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html:
3212         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/trac_unittests.js: Removed.
3213         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js: Removed.
3214         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui/results.js: Removed.
3215         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui/results_unittests.js: Removed.
3216         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js: Removed.
3217         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/unexpected-passes.html:
3218         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/unexpected-passes.js: Removed.
3219
3220 2011-08-17  Jochen Eisinger  <jochen@chromium.org>
3221
3222         Make SVNTest.test_svn_apply use the same timezone logic as VCSUtils.pm
3223         https://bugs.webkit.org/show_bug.cgi?id=65877
3224
3225         Reviewed by Eric Seidel.
3226
3227         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
3228
3229 2011-08-17  Dimitri Glazkov  <dglazkov@chromium.org>
3230
3231         Remove base.keys and replace it with Object.keys.
3232         https://bugs.webkit.org/show_bug.cgi?id=66422
3233
3234         Reviewed by Adam Barth.
3235
3236         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js: Removed base.keys.
3237         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js: Removed unit tests for it.
3238         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js: Replaced with Object.keys
3239
3240 2011-08-17  Dimitri Glazkov  <dglazkov@chromium.org>
3241
3242         garden-o-matic Summary view should have items in descending chronological order.
3243         https://bugs.webkit.org/show_bug.cgi?id=66403
3244
3245         Reviewed by Adam Barth.
3246
3247         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js: Renamed "push" to "add" for clarity, cleaned up some stuff.
3248         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Rewrote add to insert DOM elements in order.
3249         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Added unit tests.
3250
3251 2011-08-17  Dimitri Glazkov  <dglazkov@chromium.org>
3252
3253         Fix a few errors in garden-o-matic frontend unit tests.
3254         https://bugs.webkit.org/show_bug.cgi?id=66421
3255
3256         Reviewed by Adam Barth.
3257
3258         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js: Changed my excellent test to view time as variable continuum.
3259         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders_unittests.js: Tweaked the test to accommodate new builder names.
3260         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Hurriedly covered up my prior sins.
3261
3262 2011-08-17  Adam Barth  <abarth@webkit.org>
3263
3264         Update Tools to account for (CG) being added to Chromium builder names
3265         https://bugs.webkit.org/show_bug.cgi?id=66410
3266
3267         Reviewed by Adam Barth.
3268
3269         Some of these names will change back once we turn on Skia bots.
3270
3271         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders_unittests.js:
3272         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
3273         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js:
3274         * Scripts/webkitpy/layout_tests/port/builders.py:
3275         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
3276         * TestResultServer/static-dashboards/builders.js:
3277         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
3278
3279 2011-08-16  Dimitri Glazkov  <dglazkov@chromium.org>
3280
3281         Wire up relative time to garden-o-matic summary view.
3282         https://bugs.webkit.org/show_bug.cgi?id=66343
3283
3284         Reviewed by Adam Barth.
3285
3286         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js: Add time to the commitData object.
3287         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model_unittests.js: Adjusted unit tests.
3288         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Wired up relative date.
3289         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Updated unit tests.
3290         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/trac_unittests.js: Adjusted unit tests.
3291
3292 2011-08-17  Adam Roben  <aroben@apple.com>
3293
3294         Make run-api-tests run tests in alphabetical order
3295
3296         Fixes <http://webkit.org/b/66401> run-api-tests runs tests in a semi-random order
3297
3298         Reviewed by David Kilzer.
3299
3300         * Scripts/run-api-tests:
3301         (runAllTests):
3302         (runAllTestsInSuite):
3303         Sort the lists of tests and suites before iterating over them.
3304
3305 2011-08-17  Adam Barth  <abarth@webkit.org>
3306
3307         garden-o-matic should be able to rebaseline expected failures
3308         https://bugs.webkit.org/show_bug.cgi?id=66204
3309
3310         Reviewed by Dimitri Glazkov.
3311
3312         This patch introduces a new view for displaying expect, actual, and
3313         differences between results.  The new view also has a fledgling
3314         controller that handles some basic operations.  The new view is
3315         exercised by rebaseline.html, which lets you rebaselines expected
3316         failures, but the new view is not fully integrated into
3317         garden-o-matic.html
3318
3319         More patches will be required to make this stuff fully working, but
3320         this patch was already somewhat spiraling out of control.  Hopefully
3321         this patch will serve as a good starting point for further development.
3322
3323         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/controllers: Added.
3324         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/controllers.js: Added.
3325         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/controllers/rebaseline.js: Added.
3326         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
3327         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
3328         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
3329         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/rebaseline.html: Added.
3330         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
3331         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
3332         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
3333         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui/results.js:
3334         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui/results_unittests.js: Added.
3335
3336 2011-08-16  Chang Shu  <cshu@webkit.org>
3337
3338         Support reset in WebCore::Internals
3339         https://bugs.webkit.org/show_bug.cgi?id=66307
3340
3341         Reviewed by Dimitri Glazkov.
3342
3343         Added call to resetInternalsObject in DRT.
3344
3345         * DumpRenderTree/chromium/TestShell.cpp:
3346         (TestShell::resetTestController):
3347         * DumpRenderTree/mac/DumpRenderTree.mm:
3348         (resetWebViewToConsistentStateBeforeTesting):
3349         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
3350         (WebCore::WebPage::resetSettings):
3351
3352 2011-08-17  Alexis Menard  <alexis.menard@openbossa.org>
3353
3354         Add a new build slave for the Qt port on Mac OS SnowLeopard.
3355         https://bugs.webkit.org/show_bug.cgi?id=66392
3356
3357         Modify the config file to include the new build slave.
3358
3359         Reviewed by Adam Roben.
3360
3361         * BuildSlaveSupport/build.webkit.org-config/config.json:
3362
3363 2011-08-17  Sam White  <samuel.white@rochester.edu>
3364
3365         AccessibilityObject levels are inconsistent
3366         https://bugs.webkit.org/show_bug.cgi?id=66180
3367
3368         Added the ability to get numeric attribute values using numberAttributeValue. This
3369         function complements the existing stringAttributeValue and boolAttributeValue functions.
3370         The addition of numberAttributeValue was necessary because the stringAttributeValue
3371         function does a type check and will only return strings. This limitation made it
3372         impossible to get values for attributes that returned an NSNumber.
3373
3374         Reviewed by Chris Fleizach.
3375
3376         * DumpRenderTree/AccessibilityUIElement.cpp:
3377         (numberAttributeValueCallback):
3378         (AccessibilityUIElement::getJSClass):
3379         * DumpRenderTree/AccessibilityUIElement.h:
3380         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3381         (AccessibilityUIElement::numberAttributeValue):
3382         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
3383         (AccessibilityUIElement::numberAttributeValue):
3384         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
3385         (AccessibilityUIElement::numberAttributeValue):
3386
3387 2011-08-17  Adam Roben  <aroben@apple.com>
3388
3389         Tell Trac logs always to display all specified revisions in a single page
3390
3391         Trac imposes a default limit of 100 revisions per log page. We need to tell it to allow more
3392         revisions than that.
3393
3394         Fixes <http://webkit.org/b/66373> Links to Trac from TestFailures page don't always show all
3395         relevant revisions
3396
3397         Reviewed by Daniel Bates.
3398
3399         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
3400         (trac.logURL): Add a limit parameter to the log URL that allows all revisions to be seen in
3401         a single page.
3402
3403         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js:
3404         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/trac_unittests.js:
3405         Updated tests for the above.
3406
3407 2011-08-17  Adam Roben  <aroben@apple.com>
3408
3409         Fix TestFailures's link to file a bug about TestFailures itself
3410
3411         Fixes <http://webkit.org/b/66387> REGRESSION (r92135): TestFailures page's link to file a
3412         bug about itself is broken
3413
3414         Reviewed by Dan Bates.
3415
3416         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
3417         (ViewController.prototype._domForAuxiliaryUIElements): Added a path separator to separate
3418         the host from the path.
3419
3420 2011-08-17  Adam Roben  <aroben@apple.com>
3421
3422         Teach TestFailures to ignore unbelievably short test runs
3423
3424         Fixes <http://webkit.org/b/66385> TestFailures page thinks all tests passed in
3425         http://build.webkit.org/builders/Windows%207%20Release%20(Tests)/builds/14956
3426
3427         Reviewed by Dan Bates.
3428
3429         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
3430         (Builder.prototype.getNumberOfFailingTests): If it looks like all tests passed, but
3431         run-webkit-tests took less than 10 seconds to run, assume that some weird error occurred
3432         that caused it not to run any tests at all (as happened for a while due to
3433         <http://webkit.org/b/64988>). Bumped the cache version to evict old, buggy cached data.
3434
3435         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
3436         Test for the above.
3437
3438         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
3439         (LayoutTestResultsLoader.prototype.start): Bumped the cache version to evict old, buggy cached data.
3440
3441 2011-08-16  Adam Barth  <abarth@webkit.org>
3442
3443         Add a CG qualifier similar to the GPU qualifier
3444         https://bugs.webkit.org/show_bug.cgi?id=66359
3445
3446         Reviewed by Eric Seidel.
3447
3448         Once we have a Skia configuration on Chromium Mac, we'll use this CG
3449         qualifier to note that expectations apply only to the CG configuration.
3450
3451         * Scripts/webkitpy/layout_tests/port/chromium.py:
3452         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
3453
3454 2011-08-16  Eric Seidel  <eric@webkit.org>
3455
3456         REGRESSION(92717): WinPort needs to parse version information from port_name
3457         https://bugs.webkit.org/show_bug.cgi?id=66325
3458
3459         Unreviewed.  Fixing an exception/typo seen on the bots (and testing to make sure it works).
3460
3461         * Scripts/webkitpy/layout_tests/port/mac.py:
3462         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
3463
3464 2011-08-16  Adam Barth  <abarth@webkit.org>
3465
3466         Change webkit-patch analyze-baselines to print in a more unix-friendly way
3467         https://bugs.webkit.org/show_bug.cgi?id=66349
3468
3469         Reviewed by James Robinson.
3470
3471         As requested by jamesr.
3472
3473         * Scripts/webkitpy/tool/commands/rebaseline.py:
3474
3475 2011-08-16  Eric Seidel  <eric@webkit.org>
3476
3477         REGRESSION(92717): WinPort needs to parse version information from port_name
3478         https://bugs.webkit.org/show_bug.cgi?id=66325
3479
3480         Reviewed by Adam Barth.
3481
3482         WinPort didn't have any port_name parsing logic.
3483         Seemed it was best to just share the logic from MacPort.
3484         Sharing the logic from MacPort seemed easiest with a shared superclass.
3485         Once I created ApplePort (the superclass) there was a bunch of logic to share.
3486         I found LeakDetector was just in the way when switching back and forth
3487         between win.py and mac.py so I moved it into its own file.
3488
3489         The only logic change here is that WinPort should now be
3490         able to parse port_name values like MacPort (and other ports)
3491         can.  The fact that ports do this in their constructors is
3492         wrong and a source of many bugs.  But at least now WinPort
3493         matches the status-quo.
3494
3495         * Scripts/webkitpy/layout_tests/port/apple.py: Added.
3496         * Scripts/webkitpy/layout_tests/port/leakdetector.py: Added.
3497         * Scripts/webkitpy/layout_tests/port/leakdetector_unittest.py: Added.
3498         * Scripts/webkitpy/layout_tests/port/mac.py:
3499         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
3500         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
3501         * Scripts/webkitpy/layout_tests/port/win.py:
3502         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
3503
3504 2011-08-16  Adam Barth  <abarth@webkit.org>
3505
3506         Add webkit-patch analyze-baselines, as requested by jamesr.
3507
3508         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
3509         * Scripts/webkitpy/tool/commands/rebaseline.py:
3510
3511 2011-08-16  Adam Barth  <abarth@webkit.org>
3512
3513         Fold bulk-optimize-baselines into optimize-baselines
3514         https://bugs.webkit.org/show_bug.cgi?id=66345
3515
3516         Reviewed by James Robinson.
3517
3518         bulk-optimize-baselines is now a superset of optimize-baselines, so
3519         they can be combined.
3520
3521         * Scripts/webkitpy/tool/commands/rebaseline.py:
3522
3523 2011-08-16  Adam Barth  <abarth@webkit.org>
3524
3525         Move chromium-mac to chromium-cg-mac to prepare for Skia switchover
3526         https://bugs.webkit.org/show_bug.cgi?id=66334
3527
3528         Reviewed by Tony Chang.
3529
3530         This patch moves us closer to switching the Chromium port from
3531         CoreGraphics to Skia.  This patch teaches our tools that the current
3532         Chromium Mac configuration is actually chromium-cg-mac, which makes
3533         room for a future chromium-mac configuration that uses Skia.
3534
3535         Shortly after this patch lands, I'll do a server-side move of the
3536         chromium-mac and chromium-mac-leopard directories to chromium-cg-mac
3537         and chromium-cg-mac-leopard.
3538
3539         * BuildSlaveSupport/build.webkit.org-config/config.json:
3540         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
3541         * Scripts/webkitpy/common/config/build.py:
3542         * Scripts/webkitpy/common/config/build_unittest.py:
3543         * Scripts/webkitpy/layout_tests/port/base.py:
3544         * Scripts/webkitpy/layout_tests/port/builders.py:
3545         * Scripts/webkitpy/layout_tests/port/chromium.py:
3546         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
3547         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
3548         * Scripts/webkitpy/layout_tests/port/dryrun.py:
3549         * Scripts/webkitpy/layout_tests/port/factory.py:
3550         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
3551         * Scripts/webkitpy/layout_tests/port/google_chrome_unittest.py:
3552         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
3553         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
3554         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
3555         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
3556         * TestResultServer/static-dashboards/flakiness_dashboard.html:
3557
3558 2011-08-16  Dimitri Glazkov  <dglazkov@chromium.org>
3559
3560         garden-o-matic frontend needs a friendly relative time description method.
3561         https://bugs.webkit.org/show_bug.cgi?id=66324
3562
3563         Reviewed by Adam Barth.
3564
3565         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js: Added base.relativizeTime method.
3566         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js: Added unit tests.
3567         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Wired up Time widget to use it.
3568
3569 2011-08-16  Dimitri Glazkov  <dglazkov@chromium.org>
3570
3571         Summary view should not add multiple entries for the same test failure.
3572         https://bugs.webkit.org/show_bug.cgi?id=66318
3573
3574         Reviewed by Adam Barth.
3575
3576         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js: Added a simplistic updating button.
3577         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Added ways to check for
3578             equality of ui.notifications.FailingTest, and whether ui.notifications.TestFailures already contains a given failure analysis.
3579         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Added tests.
3580
3581 2011-08-16  Adam Roben  <aroben@apple.com>
3582
3583         Teach TestFailures that tests that have failed or passed many times in a row are not flaky
3584
3585         The basic strategy here is that once a test has failed or passed many times in a row we
3586         never again consider it as a possibly flaky test. It's a simple strategy but seems to result
3587         in many fewer false positives than our current behavior.
3588
3589         Fixes <http://webkit.org/b/66327> TestFailures page considers far too many tests to be
3590         flaky, including tests that failed for a while but then were fixed
3591
3592         Reviewed by Dan Bates.
3593
3594         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
3595         (FlakyLayoutTestDetector): Initialize new _buildCount property, which is used to track the
3596         number of non-too-many-failure builds we've seen.
3597         (FlakyLayoutTestDetector.prototype.incorporateTestResults): Don't track tests which haven't
3598         failed in the _maximumFailOrPassCount most recent builds. For other not-yet-considered-flaky
3599         tests, keep track of how many times they have passed or failed in a row. If they pass or
3600         fail more than _maximumFailOrPassCount times, consider them to be non-flaky. (Once a test is
3601         considered flaky it doesn't matter how many times it passes or fails.)
3602
3603         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector_unittests.js:
3604         Added tests for the above.
3605
3606 2011-08-16  Dimitri Glazkov  <dglazkov@chromium.org>
3607
3608         garden-o-matic needs a summary view with actions for each problem.
3609         https://bugs.webkit.org/show_bug.cgi?id=66144
3610
3611         Reviewed by Adam Barth.
3612
3613         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html: Added notifications tests.
3614         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js: Added.
3615         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Added.
3616         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Added.
3617         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/summary.css: Added.
3618         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html: Added.
3619
3620 2011-08-16  Dean Jackson  <dino@apple.com>
3621
3622         Adding Ted "hober" O'Connor as a non-committing contributor.
3623
3624         * Scripts/webkitpy/common/config/committers.py:
3625
3626 2011-08-16  Adam Roben  <aroben@apple.com>
3627
3628         Teach TestFailures how to interpret unfinished test runs
3629
3630         Fixes <http://webkit.org/b/66309> TestFailures thinks all tests passed in
3631         http://build.webkit.org/builders/Lion%20Intel%20Debug%20(Tests)/builds/136
3632
3633         Reviewed by David Kilzer.
3634
3635         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
3636         (Builder.prototype.getNumberOfFailingTests): Look for "isFinished" in the layout-test step
3637         rather than "isStarted" so that we don't count builds for which the test run never finished.
3638         Bumped the cache version to evict old, buggy cached data.
3639
3640         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
3641         Added a test for the above, and made a somewhat synthetic test more realistic by including
3642         more of the actual JSON data from build.webkit.org.
3643
3644         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
3645         (LayoutTestResultsLoader.prototype.start): Bumped the cache version to evict old, buggy
3646         cached data.
3647
3648 2011-08-16  Adam Roben  <aroben@apple.com>
3649
3650         Make Apple's Windows port fall back to Lion results instead of SnowLeopard
3651
3652         Apple's Windows port now uses Lion-era versions of CoreFoundation, ICU, etc., so in theory
3653         should match Lion's behavior.
3654
3655         Fixes <http://webkit.org/b/66302> Apple's Windows port matches Lion for some test results,
3656         but doesn't use Lion results
3657
3658         Reviewed by David Kilzer.
3659
3660         * Scripts/old-run-webkit-tests:
3661         (expectedDirectoryForTest):
3662         * Scripts/webkitpy/layout_tests/port/win.py:
3663         (WinPort.baseline_search_path):
3664         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
3665         (WinPortTest.test_baseline_search_path):
3666         Replaced mac-snowleopard with mac-lion.
3667
3668         * Scripts/webkitpy/common/config/build.py:
3669         (_should_file_trigger_build): Added an entry for the LayoutTests/platform/mac-lion
3670         directory, which all Apple Mac platforms and Apple Win platforms could potentially use
3671         results from. Removed Apple Win from being affected by mac-snowleopard changes, since it
3672         will no longer look there for results.
3673
3674         * Scripts/webkitpy/common/config/build_unittest.py:
3675         (ShouldBuildTest.test_should_build): Tests for the above changes.
3676
3677 2011-08-16  Dean Jackson  <dino@apple.com>
3678
3679         Move me from contributor to reviewer.
3680
3681         * Scripts/webkitpy/common/config/committers.py:
3682
3683 2011-08-15  Dimitri Glazkov  <dglazkov@chromium.org>
3684
3685         garden-o-matic frontend needs model.commitDataForRevisionRange function.
3686         https://bugs.webkit.org/show_bug.cgi?id=66252
3687
3688         Reviewed by Adam Barth.
3689
3690         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js: Added commitDataForRevisionRange method.
3691         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model_unittests.js: Added tests.
3692
3693 2011-08-15  Dimitri Glazkov  <dglazkov@chromium.org>
3694
3695         garden-o-matic frontend needs a generic way to track updates.
3696         https://bugs.webkit.org/show_bug.cgi?id=66245
3697
3698         Reviewed by Adam Barth.
3699
3700         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js: Added UpdateTracker.
3701         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js: Added tests.
3702
3703 2011-08-15  Adam Roben  <aroben@apple.com>
3704
3705         Teach TestWebKitAPI/gtest how to print JavaScript failures nicely
3706
3707         Failures now give output of the form: foo should be bar but is baz
3708
3709         Fixes <http://webkit.org/b/66240> It's hard to tell what the actual result of a failed JS
3710         test is in TestWebKitAPI's output
3711
3712         Reviewed by David Levin.
3713
3714         * TestWebKitAPI/JavaScriptTest.cpp:
3715         (TestWebKitAPI::JavaScriptCallbackContext::JavaScriptCallbackContext): Removed now-unused
3716         members.
3717         (TestWebKitAPI::javaScriptCallback): Changed to just store the result string in the context
3718         object, rather than doing any testing of it here.
3719         (TestWebKitAPI::runJSTest): Made this function a gtest predicate-formatter. This allows us
3720         to use a pretty error message when the test fails.
3721
3722         * TestWebKitAPI/JavaScriptTest.h: Changed runJSTest to a predicate-formatter, and added nice
3723         gtest-style macros that wrap it.
3724
3725         * TestWebKitAPI/Test.h: Removed now-unused TEST_ASSERT_RETURN.
3726
3727         * TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash.cpp:
3728         (TestWebKitAPI::TEST): Changed to use the new macros.
3729
3730         * TestWebKitAPI/Tests/WebKit2/RestoreSessionStateContainingFormData.cpp:
3731         (TestWebKitAPI::createSessionStateContainingFormData): Ditto. Note that this function no
3732         longer returns 0 when the JS test fails. That shouldn't have any effect on whether or not
3733         the test passes, though. Returning early seems to have been an unnecessary optimization.
3734
3735         * TestWebKitAPI/Tests/WebKit2/SpacebarScrolling.cpp:
3736         (TestWebKitAPI::TEST): Changed to use the new macros.
3737
3738 2011-08-15  Eric Seidel  <eric@webkit.org>
3739
3740         Move the Leaks bot back to ORWT until I can fix LeaksViewer regressions filed by Adam Roben in bug 66227 and 66228.
3741
3742         Unreviewed.
3743
3744         * Scripts/run-webkit-tests:
3745         (usingLeaks):
3746         (useNewRunWebKitTests):
3747
3748 2011-08-14  Adam Barth  <abarth@webkit.org>
3749
3750         Switch results detail view over to new-style object-oriented UI widgets
3751         https://bugs.webkit.org/show_bug.cgi?id=66200
3752
3753         Reviewed by Dimitri Glazkov.
3754
3755         This patch replaces my goofy template-based UI for the results
3756         comparison screen with new object-oriented UI widgets.
3757
3758         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
3759         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
3760         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
3761         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
3762         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
3763         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui/results.js: Added.
3764         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js:
3765
3766 2011-08-13  Dimitri Glazkov  <dglazkov@chromium.org>
3767
3768         garden-o-matic's analyzeUnexpectedFailures needs a completion callback.
3769         https://bugs.webkit.org/show_bug.cgi?id=66166
3770
3771         Also changed base.RequestTracker to:
3772         a) fire callback immediately if requestsInFlight is 0;
3773         b) not barf if callback is not supplied.
3774
3775         Reviewed by Adam Barth.
3776
3777         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js: Started using completion callback.
3778         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js: Added completion callback.
3779         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js: Changed RequestTracker.
3780         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittest.js: Added unit tests.
3781
3782 2011-08-12  Mark Rowe  <mrowe@apple.com>
3783
3784         Be more forward-looking in the choice of compiler.
3785
3786         Rubber-stamped by Jon Honeycutt.
3787
3788         * DumpRenderTree/mac/Configurations/CompilerVersion.xcconfig:
3789         * MiniBrowser/Configurations/CompilerVersion.xcconfig:
3790         * TestWebKitAPI/Configurations/CompilerVersion.xcconfig:
3791         * WebKitTestRunner/Configurations/CompilerVersion.xcconfig:
3792
3793 2011-08-12  Sadrul Habib Chowdhury  <sadrul@chromium.org>
3794
3795         DRT: Add support for gesture events.
3796         https://bugs.webkit.org/show_bug.cgi?id=66105
3797
3798         Reviewed by Darin Fisher.
3799
3800         * DumpRenderTree/chromium/EventSender.cpp:
3801         (EventSender::EventSender):
3802         (EventSender::gestureScrollBegin):
3803         (EventSender::gestureScrollEnd):
3804         (EventSender::gestureTap):
3805         (EventSender::gestureEvent):
3806         * DumpRenderTree/chromium/EventSender.h:
3807
3808 2011-08-12  Andy Estes  <aestes@apple.com>
3809
3810         allowRoundingHacksCallback uses the wrong argument for 'thisObject'.
3811         https://bugs.webkit.org/show_bug.cgi?id=66158
3812
3813         Reviewed by Dan Bernstein.
3814
3815         * DumpRenderTree/LayoutTestController.cpp:
3816         (allowRoundingHacksCallback): The third argument is the 'this'
3817         JSObjectRef, not the second.
3818
3819 2011-08-12  Andy Estes  <aestes@apple.com>
3820
3821         Cancel in onbeforeunload dialog sometime causes a button to stop working.
3822         https://bugs.webkit.org/show_bug.cgi?id=26211
3823
3824         Reviewed by Alexey Proskuryakov.
3825
3826         Implement a new LayoutTestController method that a test can call to set
3827         the return value of DumpRenderTree's beforeunload UI delegate.
3828
3829         * DumpRenderTree/LayoutTestController.cpp:
3830         (LayoutTestController::LayoutTestController): Initialize
3831         m_shouldStayOnPageAfterHandlingBeforeUnload.
3832         (setShouldStayOnPageAfterHandlingBeforeUnloadCallback): Call
3833         LayoutTestController::setShouldStayOnPageAfterHandlingBeforeUnload().
3834         (LayoutTestController::staticFunctions): Register
3835         'setShouldStayOnPageAfterHandlingBeforeUnload' as a static function.
3836         * DumpRenderTree/LayoutTestController.h:
3837         (LayoutTestController::shouldStayOnPageAfterHandlingBeforeUnload):
3838         (LayoutTestController::setShouldStayOnPageAfterHandlingBeforeUnload):
3839         * DumpRenderTree/chromium/LayoutTestController.cpp:
3840         (LayoutTestController::LayoutTestController): Initialize
3841         m_shouldStayOnPageAfterHandlingBeforeUnload and bind
3842         'setShouldStayOnPageAfterHandlingBeforeUnload' to its c++ setter.
3843         (LayoutTestController::reset): Reset
3844         m_shouldStayOnPageAfterHandlingBeforeUnload to false.
3845         (LayoutTestController::setShouldStayOnPageAfterHandlingBeforeUnload):
3846         Set m_shouldStayOnPageAfterHandlingBeforeUnload to the value passed in
3847         from JavaScript.
3848         * DumpRenderTree/chromium/LayoutTestController.h:
3849         (LayoutTestController::shouldStayOnPageAfterHandlingBeforeUnload):
3850         * DumpRenderTree/chromium/WebViewHost.cpp:
3851         (WebViewHost::runModalBeforeUnloadDialog): Add the correct logging and
3852         return the inverse of
3853         LayoutTestController::shouldStayOnPageAfterHandlingBeforeUnload().
3854         * DumpRenderTree/mac/UIDelegate.mm:
3855         (-[UIDelegate webView:runBeforeUnloadConfirmPanelWithMessage:initiatedByFrame:]): Ditto.
3856         * DumpRenderTree/win/UIDelegate.cpp:
3857         (UIDelegate::runBeforeUnloadConfirmPanelWithMessage): Ditto.
3858         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
3859         Define setShouldStayOnPageAfterHandlingBeforeUnload().
3860         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
3861         (WTR::InjectedBundle::postNewBeforeUnloadReturnValue): Post a message
3862         to the Test Controller telling it what value it should return in its
3863         onbeforeunload UI delegate.
3864         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
3865         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
3866         (WTR::LayoutTestController::setShouldStayOnPageAfterHandlingBeforeUnload):
3867         Call InjectedBundle::postNewBeforeUnloadReturnValue().
3868         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
3869         * WebKitTestRunner/TestController.cpp:
3870         (WTR::TestController::TestController): Initialize
3871         m_beforeUnloadReturnValue to true.
3872         (WTR::runBeforeUnloadConfirmPanel): Add the correct logging and return
3873         the value of TestController::beforeUnloadReturnValue().
3874         (WTR::TestController::resetStateToConsistentValues): Reset
3875         m_beforeUnloadReturnValue to true.
3876         * WebKitTestRunner/TestController.h:
3877         (WTR::TestController::beforeUnloadReturnValue):
3878         (WTR::TestController::setBeforeUnloadReturnValue):
3879         * WebKitTestRunner/TestInvocation.cpp:
3880         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Handle the
3881         message posted from the injected bundle by retrieving the message body
3882         as a WKBoolean and calling TestController::setBeforeUnloadReturnValue().
3883
3884 2011-08-12  Sam Weinig  <sam@webkit.org>
3885
3886         Move compiler specific macros to their own header
3887         https://bugs.webkit.org/show_bug.cgi?id=66119
3888
3889         Reviewed by Anders Carlsson.
3890
3891         * DumpRenderTree/ForwardingHeaders/wtf/Compiler.h: Added.
3892
3893 2011-08-11  Dimitri Glazkov  <dglazkov@chromium.org>
3894
3895         Add a way to extend DOM objects in garden-o-matic.
3896         https://bugs.webkit.org/show_bug.cgi?id=66096
3897
3898         Reviewed by Adam Barth.
3899
3900         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js: Added base.extends.
3901         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js: Added tests.
3902
3903 2011-08-11  MORITA Hajime  <morrita@google.com>
3904
3905         [Chromium][DRT] editing/spelling/spelling-backspace-between-lines.html fails
3906         https://bugs.webkit.org/show_bug.cgi?id=45438
3907
3908         Reviewed by Kent Tamura.
3909
3910         MockSpellCheck was assuming the input of spellCheckWord() is a word.
3911         But it can be a series of word. This change made spellCheckWord()
3912         understand simple word boundary which appears in tests.
3913
3914         * DumpRenderTree/chromium/MockSpellCheck.cpp:
3915         (MockSpellCheck::spellCheckWord):
3916
3917 2011-08-11  Mark Rowe  <mrowe@apple.com>
3918
3919         Rework some Makefile logic to remove a double-negative.
3920
3921         Reviewed by Jon Honeycutt.
3922
3923         * DumpRenderTree/mac/PerlSupport/Makefile:
3924
3925 2011-08-11  Anders Carlsson  <andersca@apple.com>
3926
3927         Add a FIXME.
3928
3929         * TestWebKitAPI/TestsController.cpp:
3930         (TestWebKitAPI::TestsController::TestsController):
3931
3932 2011-08-11  Anders Carlsson  <andersca@apple.com>
3933
3934         Initialize threading before running any tests
3935         https://bugs.webkit.org/show_bug.cgi?id=66112
3936
3937         Reviewed by Adam Barth.
3938
3939         * TestWebKitAPI/TestsController.cpp:
3940         (TestWebKitAPI::TestsController::TestsController):
3941
3942 2011-08-11  Adam Barth  <abarth@webkit.org>
3943
3944         Add unexpected-passes.html to TestFailures for marking tests as passing
3945         https://bugs.webkit.org/show_bug.cgi?id=66102
3946
3947         Reviewed by Dimitri Glazkov.
3948
3949         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
3950         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
3951             - Move some code into the library so it can be shared.
3952         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
3953         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
3954         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js:
3955             - Add round-trip unit tests.
3956         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/unexpected-passes.html: Added.
3957         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/unexpected-passes.js: Added.
3958             - I'm not entirely sure this feature should remain a separate HTML
3959               file, but it seemed better than cluttering up the main HTML file
3960               with too much extra stuff.
3961         * Scripts/webkitpy/tool/servers/gardeningserver.py:
3962         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
3963             - The gardening server couldn't handle adding PASS expectations.  Now it can.
3964
3965 2011-08-11  Eric Seidel  <eric@webkit.org>
3966
3967         NRWT has wrong fallback order for Mac now that Lion exists
3968         https://bugs.webkit.org/show_bug.cgi?id=66093
3969
3970         Reviewed by Adam Barth.
3971
3972         ORWT used a different system for fallback orders than NRWT did.
3973         I moved win.py to match ORWT fallback orders in bug 64486.
3974         This bug moves mac.py to match ORWT fallback order
3975         and adds 'lion' as a supported OS version.
3976
3977         Because ChromiumMac shares OS version detection code with AppleMac
3978         this also added support for 'lion' to chromium mac.  Hopefully that's a good thing.
3979
3980         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
3981         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
3982         * Scripts/webkitpy/layout_tests/port/mac.py:
3983         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
3984         * Scripts/webkitpy/layout_tests/port/win.py:
3985
3986 2011-08-11  Adam Barth  <abarth@webkit.org>
3987
3988         Update test results now that we use absolute URLs.
3989
3990         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout_unittests.js:
3991
3992 2011-08-11  Adam Barth  <abarth@webkit.org>
3993
3994         garden-o-matic should be able to determine when tests unexpectedly pass
3995         https://bugs.webkit.org/show_bug.cgi?id=66092
3996
3997         Reviewed by Dimitri Glazkov.
3998
3999         This patch just contains the library methods for extracting this
4000         information.  I'll add some UI in a subsequent patch.
4001
4002         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
4003         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js:
4004
4005 2011-08-11  Dimitri Glazkov  <dglazkov@chromium.org>
4006
4007         New expectation entries should be placed next to existing ones.
4008         https://bugs.webkit.org/show_bug.cgi?id=66066
4009
4010         Reviewed by Adam Barth.
4011
4012         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Added the notion of insertion point,
4013             which is computed as the last found reference to the same test.
4014         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added and adjusted unit tests.
4015
4016 2011-08-11  Dimitri Glazkov  <dglazkov@chromium.org>
4017
4018         Unreviewed, rolling out r92847.
4019         http://trac.webkit.org/changeset/92847
4020         https://bugs.webkit.org/show_bug.cgi?id=65986
4021
4022         Broke check-webkit-style
4023
4024         * Scripts/webkitpy/style/checker.py:
4025         * Scripts/webkitpy/style/checker_unittest.py:
4026
4027 2011-08-11  Dimitri Glazkov  <dglazkov@chromium.org>
4028
4029         Unreviewed, rolling out r92847.
4030         http://trac.webkit.org/changeset/92847
4031         https://bugs.webkit.org/show_bug.cgi?id=65986
4032
4033         Broke check-webkit-style
4034
4035         * Scripts/webkitpy/style/checker.py:
4036         * Scripts/webkitpy/style/checker_unittest.py:
4037
4038 2011-08-11  Pierre Rossi  <pierre.rossi@gmail.com>
4039
4040         [Qt] autotests shouldn't require config.h in the style check
4041         https://bugs.webkit.org/show_bug.cgi?id=65986
4042
4043         Remove the include rules from the style-check for Qt tests.
4044         Since for the most part it's just a matter of using the
4045         API provided by QtWebKit, enforcing the WebCore set of rules
4046         only seems to make sense in the case of MIMESniffing.
4047
4048         Reviewed by Benjamin Poulain.
4049
4050         * Scripts/webkitpy/style/checker.py:
4051         * Scripts/webkitpy/style/checker_unittest.py:
4052
4053 2011-08-11  Zsolt Fehér  <h490267@stud.u-szeged.hu>
4054
4055         [WK2] Implement waitForPolicyDelegate
4056         https://bugs.webkit.org/show_bug.cgi?id=42330
4057
4058         Reviewed by Csaba Osztrogonác.
4059
4060         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
4061         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
4062         (WTR::navigationTypeToString):
4063         (WTR::InjectedBundlePage::decidePolicyForNavigationAction):
4064         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
4065         (WTR::LayoutTestController::LayoutTestController):
4066         (WTR::LayoutTestController::setCustomPolicyDelegate):
4067         (WTR::LayoutTestController::waitForPolicyDelegate):
4068         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
4069         (WTR::LayoutTestController::isPolicyDelegateEnabled):
4070         (WTR::LayoutTestController::isPolicyDelegatePermissive):
4071
4072 2011-08-11  Adam Barth  <abarth@webkit.org>
4073
4074         rebaseline-expectations should use the release bots, not the debug
4075         bots.  The debug bots don't all run all the tests.  Also, delay
4076         updating test_expectations a long as possible to avoid clearing out
4077         lines that we'll need for other ports.
4078
4079         * Scripts/webkitpy/layout_tests/port/builders.py:
4080         * Scripts/webkitpy/tool/commands/rebaseline.py:
4081         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
4082
4083 2011-08-10  Ryosuke Niwa  <rniwa@webkit.org>
4084
4085         r92792 inadvertently removed PlatformSpecificScheduler for all Snow Leopard bots.
4086         Revert it partially so that the remaining Snow Leopard bots can fulfill their duty.
4087
4088         * BuildSlaveSupport/build.webkit.org-config/config.json:
4089
4090 2011-08-10  Eric Seidel  <eric@webkit.org>
4091
4092         new-run-webkit-test's WinPort has no fallback logic
4093         https://bugs.webkit.org/show_bug.cgi?id=64486
4094
4095         Unreviewed.
4096
4097         Turns out that we've long had a subtle bug whereby if you didn't
4098         already have DumpRenderTree built, we might end up running unsupported tests
4099         because the DumpRenderTree --print-supported-features check would silently fail.
4100
4101         When I moved the call from using popen to using Executive.run_command, this silent
4102         failure became noisy.  (Breaking some bots.)
4103
4104         This re-orders our steps so that we always build DRT before collecting expectations,
4105         as well as adds a warning message when DumpRenderTree can't be found.
4106
4107         * Scripts/webkitpy/layout_tests/port/webkit.py:
4108         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
4109
4110 2011-08-10  Eric Seidel  <eric@webkit.org>
4111
4112         new-run-webkit-test's WinPort has no fallback logic
4113         https://bugs.webkit.org/show_bug.cgi?id=64486
4114
4115         Unreviewed.  Fix test-webkitpy under Python 2.5.
4116
4117         * Scripts/webkitpy/layout_tests/port/win.py:
4118
4119 2011-08-10  Benjamin Poulain  <benjamin@webkit.org>
4120
4121         [Qt][WK2] MiniBrowser is firing twice the QDesktopWebView::mousePressEvent method
4122         https://bugs.webkit.org/show_bug.cgi?id=65875
4123
4124         Reviewed by Kenneth Rohde Christiansen.
4125
4126         For some reason, QWidget send fake mouse event on some platform, under certains conditions.
4127
4128         This causes mouse events to be sent twice in some cases. We work around the problem
4129         by defining the first touch point to be a primary touch point (case which hit some
4130         weird condition in QWidget, skipping the generation of fake mouse events).
4131
4132         * MiniBrowser/qt/MiniBrowserApplication.cpp:
4133         (MiniBrowserApplication::notify):
4134
4135 2011-08-10  David Hyatt  <hyatt@apple.com>
4136
4137         https://bugs.webkit.org/show_bug.cgi?id=66004
4138
4139         Remove the regions and exclusions bots and also the flags in build-webkit, since I'm just turning
4140         the code on by default.
4141
4142         Reviewed by Adam Roben.
4143
4144         * BuildSlaveSupport/build.webkit.org-config/config.json:
4145         * Scripts/build-webkit:
4146
4147 2011-08-10  Oliver Hunt  <oliver@apple.com>
4148
4149         JSEvaluteScript does not return the correct object when given JSONP data
4150         https://bugs.webkit.org/show_bug.cgi?id=66003
4151
4152         Reviewed by Gavin Barraclough.
4153
4154         Add a test case.
4155
4156         * DumpRenderTree/mac/DumpRenderTree.mm:
4157         (testStringByEvaluatingJavaScriptFromString):
4158
4159 2011-08-10  Eric Seidel  <eric@webkit.org>
4160
4161         WIN: NRWT runs compositing tests on configurations that don't support compositing
4162         https://bugs.webkit.org/show_bug.cgi?id=64472
4163
4164         Reviewed by Adam Barth.
4165
4166         I think the code was just wrong.  It was never splitting the string
4167         into list pieces before.
4168         I've now tested the supported_features code and theoretically it
4169         should now work with Windows DRT.
4170
4171         I also made the list-lookup functions always return lists, to make
4172         it possible to clean up the list-transform code in the future
4173         (I opted not to do that cleanup in this patch, but at least now
4174         both types of feature lookup functions match return types.)
4175
4176         * Scripts/webkitpy/layout_tests/port/webkit.py:
4177         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
4178
4179 2011-08-10  Eric Seidel  <eric@webkit.org>
4180
4181         Clean up ChromiumDriver a little
4182         https://bugs.webkit.org/show_bug.cgi?id=65995
4183
4184         Unreviewed.
4185
4186         Fix NRWT for chromium.  Silly python.
4187
4188         * Scripts/webkitpy/layout_tests/port/chromium.py:
4189
4190 2011-08-10  Eric Seidel  <eric@webkit.org>
4191
4192         Clean up ChromiumDriver a little
4193         https://bugs.webkit.org/show_bug.cgi?id=65995
4194
4195         Reviewed by Adam Barth.
4196
4197         - We no longer support threading, so remove threading specific code.
4198         - Add FIXMEs about using ServerProcess instead.
4199         - Condense option-mapping if-cascade into a for loop.
4200         - Unindent long if blocks by using early return.
4201         - Unwrap lines which are needlessly wrapped.
4202
4203         There should be no functional changes here, just code cleanup/dead-code removal.
4204
4205         * Scripts/webkitpy/layout_tests/port/chromium.py:
4206
4207 2011-08-10  Dimitri Glazkov  <dglazkov@chromium.org>
4208
4209         Unmuddle construction options for TestConfiguration.
4210         https://bugs.webkit.org/show_bug.cgi?id=65988
4211
4212         The TestConfiguration instance is now constructed only using actual values.
4213         Port-based construction is split into a "from_port" class method.
4214
4215         Reviewed by Eric Seidel.
4216
4217         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Added TestConfiguration.from_port
4218         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Adjusted code to accommodate the change.
4219         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Ditto.
4220         * Scripts/webkitpy/layout_tests/port/base.py: Ditto.
4221         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py: Ditto.
4222
4223 2011-08-10  Eric Seidel  <eric@webkit.org>
4224
4225         Leaks bot shows python logging prefixes as part of summary display
4226         https://bugs.webkit.org/show_bug.cgi?id=65931
4227
4228         Reviewed by Adam Barth.
4229
4230         The big part of this change is adding a unittesting system for
4231         master.cfg, so mere-mortals can make changes to build.webkit.org
4232         with greatly-reduced risk of breaking the world.
4233
4234         Turns out that in typing up my trival logging fix I made 2 typos,
4235         so hopefully the hour spent getting master.cfg to load was worth it. :)
4236
4237         I also tried to make mastercfg_unittest run as part of test-webkitpy
4238         but right now test-webkitpy requires that unittest files be located in
4239         a module-name-compatible directory.  'build.webkit.org-config' is not
4240         a valid module name.
4241
4242         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
4243          - Fixed _parseNewRunWebKitTestsOutput to strip NRWT's python formatter prefix
4244            by using a new _strip_python_logging_prefix function.
4245          - Also fixed this function to file to work with a more modern
4246            simplejson version (which returns unicode objects instead of str objects)
4247            and filed a related bug with buildbot.net due to their handling of unicode builder names.
4248         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py: Added.
4249          - BuildBotConfigLoader is a huge pile of code just to get master.cfg to load.
4250          - The actual tiny unittest of the new _parseNewRunWebKitTestsOutput logic.
4251         * Scripts/webkitpy/thirdparty/__init__.py:
4252          - Add autoinstallation of buildbot (and jinja2) for use by mastercfg_unittest.py
4253
4254 2011-08-09  Dimitri Glazkov  <dglazkov@chromium.org>
4255
4256         garden-o-matic should extrapolate failures given a set of known results
4257         https://bugs.webkit.org/show_bug.cgi?id=65882
4258
4259         Reviewed by Adam Barth.
4260
4261         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Spuriously covered my old sins.
4262         * Scripts/webkitpy/layout_tests/port/builders.py: Augmented builders dictionary to contain build coverage specifiers,
4263             adjusted all call sites.
4264         * Scripts/webkitpy/tool/servers/gardeningserver.py: Added BuildCoverageExtrapolator, a thing that
4265             makes wider net for specific failures.
4266         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Added unit tests.
4267
4268 2011-08-10  Adam Barth  <abarth@webkit.org>
4269
4270         The URL we were computing for retrieving baselines was missing a slash
4271         character because our mocks had an extra slash character.  I've
4272         corrected both the code and the mock.
4273
4274         * Scripts/webkitpy/tool/commands/rebaseline.py:
4275         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
4276         * Scripts/webkitpy/tool/mocktool.py:
4277
4278 2011-08-09  Adam Barth  <abarth@webkit.org>
4279
4280         Move the quote mark to the right place.
4281
4282         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
4283
4284 2011-08-09  Adam Barth  <abarth@webkit.org>
4285
4286         garden-o-matic should use huerstics to narrow regression range
4287         https://bugs.webkit.org/show_bug.cgi?id=65948
4288
4289         Reviewed by Dimitri Glazkov.
4290
4291         When analyzing a regression range, if one of the commit messages
4292         mentions the test in question, we now hueristically assume that
4293         revision caused the failure.  (I'm shamelessly stealing this idea from
4294         TestFailures.)
4295
4296         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
4297
4298 2011-07-13  Eric Seidel  <eric@webkit.org>
4299
4300         new-run-webkit-test's WinPort has no fallback logic
4301         https://bugs.webkit.org/show_bug.cgi?id=64486
4302
4303         Reviewed by Adam Roben.
4304
4305         I've tried to write a patch for bug 64439 twice now, and both times
4306         I've ended up re-writing half the port system.  So I'm breaking
4307         things up into smaller pieces, this being the first.
4308
4309         WinPort still does not have any port_name parsing, so when instantiated
4310         with the name "win-xp" (i.e. by the rebaseline server) it will just behave as the 'win' port.
4311         I'll fix this in a second pass when I standardize port_name parsing for all webkit ports.
4312
4313         Otherwise this should "just work" for windows.  I've not been able to test the
4314         version detection on my mac, but the unit tests show the code behaving as designed.
4315
4316         * Scripts/webkitpy/layout_tests/port/win.py:
4317         * Scripts/webkitpy/layout_tests/port/win_unittest.py: Added.
4318
4319 2011-08-09  Adam Barth  <abarth@webkit.org>
4320
4321         Teach build.webkit.org's garden-o-matic how to talk to the local server
4322         https://bugs.webkit.org/show_bug.cgi?id=65940
4323
4324         Reviewed by Dimitri Glazkov.
4325
4326         After this patch, the version of garden-o-matic on build.webkit.org is
4327         fully functional.  It can interact with the local server via CORS.
4328
4329         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
4330         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
4331         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
4332         * Scripts/webkitpy/tool/commands/gardenomatic.py:
4333         * Scripts/webkitpy/tool/servers/gardeningserver.py:
4334
4335 2011-08-09  Adam Barth  <abarth@webkit.org>
4336
4337         Enable CORS for garden-o-matic
4338         https://bugs.webkit.org/show_bug.cgi?id=65936
4339
4340         Reviewed by Dimitri Glazkov.
4341
4342         This patch will allow the hosted instance of garden-o-matic to
4343         communicate with the local server.
4344
4345         * Scripts/webkitpy/tool/servers/gardeningserver.py:
4346         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
4347
4348 2011-08-09  Adam Barth  <abarth@webkit.org>
4349
4350         Upgrade jquery in TestFailures
4351         https://bugs.webkit.org/show_bug.cgi?id=65935
4352
4353         Reviewed by Dimitri Glazkov.
4354
4355         Upgrade to a newer version of jQuery with better support for CORS and CSP.
4356
4357         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
4358         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html:
4359         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
4360
4361 2011-08-09  Sheriff Bot  <webkit.review.bot@gmail.com>
4362
4363         Unreviewed, rolling out r92683.
4364         http://trac.webkit.org/changeset/92683
4365         https://bugs.webkit.org/show_bug.cgi?id=65921
4366
4367          "Breaks Qt --minimal build. Revert pending update of the Qt
4368         minimal configuration" (Requested by simathur on #webkit).
4369
4370         * QtTestBrowser/cookiejar.cpp:
4371         (TestBrowserCookieJar::TestBrowserCookieJar):
4372         * QtTestBrowser/launcherwindow.cpp:
4373         (LauncherWindow::createChrome):
4374         (LauncherWindow::setDiskCache):
4375         * QtTestBrowser/main.cpp:
4376         (LauncherApplication::handleUserOptions):
4377         * QtTestBrowser/webpage.cpp:
4378         (WebPage::openUrlInDefaultBrowser):
4379
4380 2011-08-09  Siddharth Mathur  <siddharth.mathur@nokia.com>
4381
4382         [Qt] Simplify code by removing QT_NO_DESKTOPSERVICES and QT_NO_NETWORKDISKCACHE
4383         https://bugs.webkit.org/show_bug.cgi?id=65880
4384
4385         Reviewed by Andreas Kling.
4386
4387         QDesktopServices and QNetworkDiskCache have been around for a while now. Kill flags which
4388         check for their feature flags. Assume that any reasonable Qt build provides them.
4389         * QtTestBrowser/cookiejar.cpp:
4390         (TestBrowserCookieJar::TestBrowserCookieJar):
4391         * QtTestBrowser/launcherwindow.cpp:
4392         (LauncherWindow::createChrome):
4393         (LauncherWindow::setDiskCache):
4394         * QtTestBrowser/main.cpp:
4395         (LauncherApplication::handleUserOptions):
4396         * QtTestBrowser/webpage.cpp:
4397         (WebPage::openUrlInDefaultBrowser):
4398
4399 2011-08-09  Adam Barth  <abarth@webkit.org>
4400
4401         BaselineOptimizer created the wrong baseline for fast/js/regexp-overflow.html
4402         https://bugs.webkit.org/show_bug.cgi?id=65891
4403
4404         Reviewed by Eric Seidel.
4405
4406         The problem was that platform/chromium contained a bogus expectation
4407         file that needed to be removed, but by the time we got around to
4408         removing it, we'd already moved the correct baseline into its place.
4409         After this patch, we copy the gold results into memory before
4410         reshuffling things on disk (and we delete bad things before adding good
4411         things).
4412
4413         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
4414         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
4415
4416 2011-08-08  Adam Barth  <abarth@webkit.org>
4417
4418         Chromium Windows bots can't figure out what SVN revision they're running
4419         https://bugs.webkit.org/show_bug.cgi?id=65893
4420
4421         Reviewed by Eric Seidel.
4422
4423         The comment in _engage_awesome_windows_hacks explains why we're making
4424         this change.  It's ugly and rediculous, but this approach seems better
4425         than using shell=True when calling popen.
4426
4427         * Scripts/webkitpy/common/checkout/scm/svn.py:
4428         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
4429
4430 2011-08-08  Adam Barth  <abarth@webkit.org>
4431
4432         Remove deduplicate-tests
4433         https://bugs.webkit.org/show_bug.cgi?id=65886
4434
4435         Reviewed by Tony Chang.
4436
4437         This script is redundant with "webkit-patch bulk-optimize-baselines",
4438         which can find many more duplicates than deduplicate-tests because
4439         bulk-optimize-baselines can move results around whereas
4440         deduplicate-tests can only remove results.
4441
4442         * Scripts/deduplicate-tests: Removed.
4443         * Scripts/webkitpy/to_be_moved/deduplicate_tests.py: Removed.
4444         * Scripts/webkitpy/to_be_moved/deduplicate_tests_unittest.py: Removed.
4445
4446 2011-08-08  Adam Barth  <abarth@webkit.org>
4447
4448         scm.delete should delete empty parent directories as well
4449         https://bugs.webkit.org/show_bug.cgi?id=65878
4450
4451         Reviewed by Eric Seidel.
4452
4453         This behavior makes SVN match GIT (which has no concept of empty
4454         directories).  This bug comes up a lot when optimizing baselines, which
4455         often create empty directories when we're able to optimize everything
4456         out of a given folder.
4457
4458         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
4459         * Scripts/webkitpy/common/checkout/scm/svn.py:
4460
4461 2011-08-08  Lucas Forschler  <lforschler@apple.com>
4462
4463         Add triggers for Lion WK2 test bots.
4464
4465         Reviewed by Stephanie Lewis.
4466
4467         * BuildSlaveSupport/build.webkit.org-config/config.json:
4468
4469 2011-08-08  Adam Barth  <abarth@webkit.org>
4470
4471         bulk-optimize-baselines should take directories on the command line
4472         https://bugs.webkit.org/show_bug.cgi?id=65879
4473
4474         Reviewed by Dimitri Glazkov.
4475
4476         Passing lists of tests via stdin is a pain.  It's easier if the command
4477         takes a list of tests on the command line, like run-webkit-tests.
4478
4479         * Scripts/webkitpy/tool/commands/rebaseline.py:
4480
4481 2011-08-08  Adam Barth  <abarth@webkit.org>
4482
4483         Deprecate rebaseline-chromium-webkit-tests
4484         https://bugs.webkit.org/show_bug.cgi?id=65873
4485
4486         Reviewed by Dimitri Glazkov.
4487
4488         This script never worked correctly and now is redundant with more
4489         modern infrastructure for rebaselining tests.  This patch deprecates
4490         the script in order to flush out any issues with the new tools.  Once
4491         folks are happy with the new tools, we'll delete this script.
4492
4493         * Scripts/rebaseline-chromium-webkit-tests:
4494
4495 2011-08-08  Dimitri Glazkov  <dglazkov@chromium.org>
4496
4497         garden-o-matic should only touch expectations that need updating.
4498         https://bugs.webkit.org/show_bug.cgi?id=65876
4499
4500         Reviewed by Adam Barth.
4501
4502         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Added reconstitute_only_these to TestExpectationSerializer.list_to_string,
4503             so that a caller could influence which lines to serialize from values.
4504         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added tests.
4505         * Scripts/webkitpy/tool/servers/gardeningserver.py: Started using reconstitute_only_these.
4506         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Added tests.
4507
4508 2011-08-08  Dimitri Glazkov  <dglazkov@chromium.org>
4509
4510         TestExpectationsEditor should return a list of updated expectations.
4511         https://bugs.webkit.org/show_bug.cgi?id=65872
4512
4513         Reviewed by Adam Barth.
4514
4515         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Added a list to store lines that are being updated.
4516         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Changed tests to account for updated lines.
4517
4518 2011-08-08  Jochen Eisinger  <jochen@chromium.org>
4519
4520         Fix SCM webkitpy unit test failures
4521         https://bugs.webkit.org/show_bug.cgi?id=65823
4522
4523         Reviewed by Eric Seidel.
4524
4525         * Scripts/webkitpy/common/checkout/scm/git.py:
4526         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
4527
4528 2011-08-08  Eric Seidel  <eric@webkit.org>
4529
4530         Move the --leaks bot back over to NRWT now that I'm back
4531         from vacation and able to debug it.
4532
4533         Unreviewed.
4534
4535         * Scripts/run-webkit-tests:
4536         (usingWebKit2):
4537         (useNewRunWebKitTests):
4538
4539 2011-08-05  Dimitri Glazkov  <dglazkov@chromium.org>
4540
4541         Wire up updating expectations in garden-o-matic.
4542         https://bugs.webkit.org/show_bug.cgi?id=65794
4543
4544         Reviewed by Adam Barth.
4545
4546         * Scripts/webkitpy/layout_tests/port/factory.py: Moved _port_for_builder from rebaseline.py to here, rename to get_from_builder_name.
4547         * Scripts/webkitpy/tool/commands/rebaseline.py: Adapted callsites to new location.
4548         * Scripts/webkitpy/tool/servers/gardeningserver.py: Added GardeningExpectationsUpdater, which knows how to update expectations.
4549         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Added loads of tests.
4550
4551 2011-08-08  Jochen Eisinger  <jochen@chromium.org>
4552
4553         webkit-patch doesn't get along with renamed files
4554         https://bugs.webkit.org/show_bug.cgi?id=48075
4555
4556         Possibly a bit heavy handed - I removed all instances of -C and
4557         changed every instance of -M with '--no-renames' in git.py. This
4558         forces git to not try to tell us about renames at all, which is
4559         ultimately the behaviour we want. The old file is shown deleted,
4560         then the new file is shown added, followed by any changes that
4561         occurred. Also gets rid of the problem where deleting one file
4562         and adding another file which has similar content would
4563         unexpectedly show up as a rename, and fall out of a diff.
4564
4565         Based on a patch by Wyatt Carss.
4566
4567         Reviewed by Eric Seidel.
4568
4569         * Scripts/webkitpy/common/checkout/scm/git.py:
4570
4571 2011-08-07  Sam White  <samuel.white@rochester.edu>
4572
4573         Add the ability to search the AccessibilityObject cache
4574         https://bugs.webkit.org/show_bug.cgi?id=64994
4575
4576         Added testing support for AccessibilityObject cache searching. Currently,
4577         only the mac platform is full supported and has had API exposed. Other
4578         platforms have only have this new functionality stubbed. Full implementation
4579         is a job suited only for an accessibility expert of each respective platform.
4580
4581         Reviewed by Chris Fleizach.
4582
4583         * DumpRenderTree/AccessibilityUIElement.cpp:
4584         (uiElementForSearchPredicateCallback):
4585         (AccessibilityUIElement::getJSClass):
4586         * DumpRenderTree/AccessibilityUIElement.h:
4587         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
4588         (AccessibilityUIElement::uiElementForSearchPredicate):
4589         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
4590         (AccessibilityUIElement::uiElementForSearchPredicate):
4591         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
4592         (AccessibilityUIElement::uiElementForSearchPredicate):
4593
4594 2011-08-06  Adam Barth  <abarth@webkit.org>
4595
4596         Remove dead code in webkitpy
4597         https://bugs.webkit.org/show_bug.cgi?id=65818
4598
4599         Reviewed by Dimitri Glazkov.
4600
4601         This code was added to support the rebaseline2 command six months ago,
4602         but the rebaseline2 command was never finished and no one else has used
4603         this code since it was landed.  This patch removes the dead code.  If
4604         we revive rebaseline2, then we can revert this patch and recover the
4605         code.  Until then, this code is just dead weight.
4606
4607         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
4608         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
4609         * Scripts/webkitpy/common/net/testoutput.py: Removed.
4610         * Scripts/webkitpy/common/net/testoutput_unittest.py: Removed.
4611         * Scripts/webkitpy/common/net/testoutputset.py: Removed.
4612         * Scripts/webkitpy/common/net/testoutputset_unittest.py: Removed.
4613         * Scripts/webkitpy/common/system/directoryfileset.py: Removed.
4614         * Scripts/webkitpy/common/system/directoryfileset_unittest.py: Removed.
4615         * Scripts/webkitpy/layout_tests/port/base.py:
4616         * Scripts/webkitpy/layout_tests/port/chromium.py:
4617         * Scripts/webkitpy/layout_tests/port/webkit.py:
4618
4619 2011-08-06  Adam Barth  <abarth@webkit.org>
4620
4621         webkit-patch optimize-baselines can't handle promotions to base results directory
4622         https://bugs.webkit.org/show_bug.cgi?id=65819
4623
4624         Reviewed by Dimitri Glazkov.
4625
4626         This bug came up when testing the optimizer on some editing baselines.
4627
4628         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
4629         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
4630
4631 2011-08-06  Adam Barth  <abarth@webkit.org>
4632
4633         Add a small command to optimize baselines in bulk.
4634
4635         * Scripts/webkitpy/tool/commands/rebaseline.py:
4636
4637 2011-08-05  Dimitri Glazkov  <dglazkov@chromium.org>
4638
4639         Fix unit test breakage by plumbing convert_404_to_None in MockWeb.
4640
4641         * Scripts/webkitpy/tool/mocktool.py: Added missing parameter to MockWeb.get_binary
4642
4643 2011-08-05  Dimitri Glazkov  <dglazkov@chromium.org>
4644
4645         Garden-o-matic updateexpectations needs a unit test.
4646         https://bugs.webkit.org/show_bug.cgi?id=65780
4647
4648         Reviewed by Adam Barth.
4649
4650         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Added unit test.
4651
4652 2011-08-05  Anders Carlsson  <andersca@apple.com>
4653
4654         Future proof an Objective-C test
4655         https://bugs.webkit.org/show_bug.cgi?id=65788
4656
4657         Reviewed by Dan Bernstein.
4658
4659         Add isObjectInstanceOf to ObjCController.
4660
4661         * DumpRenderTree/mac/ObjCController.m:
4662         (+[ObjCController isSelectorExcludedFromWebScript:]):
4663         (+[ObjCController webScriptNameForSelector:]):
4664         (-[ObjCController isObject:instanceOf:]):
4665
4666 2011-08-05  Adam Barth  <abarth@webkit.org>
4667
4668         trac.js needs unittests
4669         https://bugs.webkit.org/show_bug.cgi?id=65673
4670
4671         Reviewed by Dimitri Glazkov.
4672
4673         This tests also have somewhat large fixtures.  I've trimmed the example
4674         XML down a bunch while still hitting interesting cases in the parsing.
4675
4676         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
4677         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/trac_unittests.js: Added.
4678
4679 2011-08-05  Adam Barth  <abarth@webkit.org>
4680
4681         Rebuild rebaseline-chromium-webkit-tests on top of modern infrastructure
4682         https://bugs.webkit.org/show_bug.cgi?id=65759
4683
4684         Reviewed by Dimitri Glazkov.
4685
4686         rebaseline-chromium-webkit-tests doesn't really fit into webkitpy's
4687         architecture, is poorly tested, and has a bunch of quirks.  This patch
4688         rebuilds the core functionality of rebaseline-chromium-webkit-tests on
4689         top of more modern infrastructure.  In the process, we get more code
4690         re-use and better testing.
4691
4692         Once we're satisfied with this new implementation, we can delete the
4693         old implementation.
4694
4695         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
4696         * Scripts/webkitpy/common/net/web.py:
4697         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
4698         * Scripts/webkitpy/layout_tests/port/builders.py:
4699         * Scripts/webkitpy/layout_tests/port/chromium.py:
4700         * Scripts/webkitpy/layout_tests/port/factory.py:
4701         * Scripts/webkitpy/layout_tests/port/webkit.py:
4702         * Scripts/webkitpy/tool/commands/rebaseline.py:
4703         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
4704
4705 2011-08-05  Anders Carlsson  <andersca@apple.com>
4706
4707         Remove PluginHalter
4708         https://bugs.webkit.org/show_bug.cgi?id=65729
4709
4710         Reviewed by Darin Adler.
4711
4712         Remove call to set the allowed plug-in run time.
4713
4714         * DumpRenderTree/mac/DumpRenderTree.mm:
4715         (resetDefaultsToConsistentValues):
4716
4717 2011-08-05  Ademar de Souza Reis Jr.  <ademar.reis@openbossa.org>
4718
4719         Unreviewed build fix for Qt-Minimal buildbot
4720
4721         Use ~/.QtTestBrowser for the cookiejar when QDesktopServices
4722         is disabled.
4723
4724         Rubber-stamped by Holger Freyther.
4725
4726         * QtTestBrowser/cookiejar.cpp:
4727         (TestBrowserCookieJar::TestBrowserCookieJar):
4728
4729 2011-08-05  Ademar de Souza Reis Jr  <ademar.reis@openbossa.org>
4730
4731         [Qt] QtTestBrowser: Add support for saving cookies on disk
4732         https://bugs.webkit.org/show_bug.cgi?id=65328
4733
4734         Simple cookiejar that saves cookies on disk.
4735
4736         Added menu options to clear cookies (Edit->Clear cookies) and
4737         disable disk cookies (Settings->[ ]Save Cookies on Disk).
4738
4739         Option -no-disk-cookies is also available.
4740
4741         Reviewed by Andreas Kling.
4742
4743         * QtTestBrowser/QtTestBrowser.pro:
4744         * QtTestBrowser/cookiejar.cpp: Added.
4745         * QtTestBrowser/cookiejar.h: Added.
4746         * QtTestBrowser/launcherwindow.cpp:
4747         * QtTestBrowser/launcherwindow.h:
4748         * QtTestBrowser/main.cpp:
4749
4750 2011-08-04  Adam Barth  <abarth@webkit.org>
4751
4752         model.js needs some unit testing
4753         https://bugs.webkit.org/show_bug.cgi?id=65730
4754
4755         Reviewed by Dimitri Glazkov.
4756
4757         These tests do not cover all the functions exported by the model
4758         module.  We'll need to add the remaining ones in the future.
4759
4760         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model_unittests.js: Added.
4761         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
4762
4763 2011-08-04  Adam Barth  <abarth@webkit.org>
4764
4765         Another speculative fix for the "svn revision on Windows" problem.
4766
4767         * Scripts/webkitpy/common/checkout/scm/git.py:
4768
4769 2011-08-04  Mark Rowe  <mrowe@apple.com>
4770
4771         Future-proof Xcode configuration settings.
4772
4773         * DumpRenderTree/mac/Configurations/Base.xcconfig:
4774         * DumpRenderTree/mac/Configurations/CompilerVersion.xcconfig:
4775         * DumpRenderTree/mac/Configurations/DebugRelease.xcconfig:
4776         * MiniBrowser/Configurations/Base.xcconfig:
4777         * MiniBrowser/Configurations/CompilerVersion.xcconfig:
4778         * MiniBrowser/Configurations/DebugRelease.xcconfig:
4779         * TestWebKitAPI/Configurations/Base.xcconfig:
4780         * TestWebKitAPI/Configurations/CompilerVersion.xcconfig:
4781         * TestWebKitAPI/Configurations/DebugRelease.xcconfig:
4782         * WebKitTestRunner/Configurations/Base.xcconfig:
4783         * WebKitTestRunner/Configurations/CompilerVersion.xcconfig:
4784         * WebKitTestRunner/Configurations/DebugRelease.xcconfig:
4785
4786 2011-08-04  Adam Barth  <abarth@webkit.org>
4787
4788         Fancier debug logging.
4789
4790         * Scripts/webkitpy/layout_tests/controllers/manager.py:
4791
4792 2011-08-04  David Levin  <levin@chromium.org>
4793
4794         [chromium] Fix assertion failure in ThreadIdentifierData::identifier in TestShell.
4795         https://bugs.webkit.org/show_bug.cgi?id=65685
4796
4797         Reviewed by Dmitry Titov.
4798
4799         * DumpRenderTree/chromium/WebViewHost.cpp:
4800         (WebViewHost::WebViewHost): Intialize WTF threading before
4801         anything in WTF is used. Usually this is done by WebViewImpl::WebViewImpl
4802         but in some cases code may use WTF without creating WebViewImpl and
4803         WTF code functionality may call WTF::currentThread, which requires
4804         this initialization now.
4805
4806 2011-08-04  Adam Barth  <abarth@webkit.org>
4807
4808         An educated guess at causing the Windows bots on build.webkit.org to
4809         find the SVN revision number.
4810
4811         * Scripts/webkitpy/layout_tests/controllers/manager.py:
4812         * Scripts/webkitpy/layout_tests/port/base.py:
4813
4814 2011-08-04  Adam Barth  <abarth@webkit.org>
4815
4816         More logging to help debug the "no SVN rev on Chromium Windows bots"
4817         bug.  My current theory is that we're trying to detect the SCM system
4818         in the wrong directory.
4819
4820         * Scripts/webkitpy/layout_tests/port/base.py:
4821
4822 2011-08-04  Dimitri Glazkov  <dglazkov@chromium.org>
4823
4824         Speed up webkit-patch optimize-expectations significantly by not collecting test files.
4825         https://bugs.webkit.org/show_bug.cgi?id=65697
4826
4827         I neglected to realize that collecting test names is not necessary for this operation.
4828
4829         Reviewed by Adam Barth.
4830
4831         * Scripts/webkitpy/tool/commands/expectations.py: Changed to pass an empty list of test files.
4832
4833 2011-08-04  Adam Barth  <abarth@webkit.org>
4834
4835         checkout_unittest.js should have better coverage of checkout.js
4836         https://bugs.webkit.org/show_bug.cgi?id=65671
4837
4838         Reviewed by Dimitri Glazkov.
4839
4840         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
4841             - Remove unused functions.
4842         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout_unittests.js:
4843             - Test remaining public functions in the checkout module.
4844
4845 2011-08-04  Dmitry Lomov  <dslomov@google.com>
4846
4847         https://bugs.webkit.org/show_bug.cgi?id=65706
4848         Run run-unit-tests on release mode test bots.
4849         Reenabling after 61812 is fixed.
4850
4851         Reviewed by Adam Roben.
4852
4853         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
4854
4855 2011-08-04  Adam Barth  <abarth@webkit.org>
4856
4857         builders.js needs unit tests
4858         https://bugs.webkit.org/show_bug.cgi?id=65670
4859
4860         Reviewed by Dimitri Glazkov.
4861
4862         This test requires somewhat large test fixtures, but using real
4863         examples seemed better than trimming them down too much.
4864
4865         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders.js:
4866         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders_unittests.js: Added.
4867         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
4868             - Note: trac_unittests.js doesn't exist (yet!).
4869
4870 2011-08-04  Alexis Menard  <alexis.menard@openbossa.org>
4871
4872         [Qt] Make navigation actions properly usable in QML.
4873         https://bugs.webkit.org/show_bug.cgi?id=65624
4874
4875         Fix compilation after the introduction of QWebNavigationController.
4876
4877         Reviewed by Benjamin Poulain.
4878
4879         * MiniBrowser/qt/BrowserView.cpp:
4880         (BrowserView::navigationAction):
4881
4882 2011-08-04  Adam Barth  <abarth@webkit.org>
4883
4884         Another subtle bug due to lack of testing in main.js.  :(
4885
4886         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
4887
4888 2011-08-04  Adam Barth  <abarth@webkit.org>
4889
4890         Fix types.  Sadly, main.js has no test coverage.  (I need to think
4891         about how to test this part of the code.)
4892
4893         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
4894
4895 2011-08-04  Adam Barth  <abarth@webkit.org>
4896
4897         bugzilla.js is missing unit tests
4898         https://bugs.webkit.org/show_bug.cgi?id=65656
4899
4900         Reviewed by Adam Roben.
4901
4902         Testing makes perfect.
4903
4904         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/bugzilla_unittests.js: Added.
4905         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net_unittests.js:
4906         (NetworkSimulator.prototype.runTest.get if):
4907         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
4908
4909 2011-08-04  Adam Barth  <abarth@webkit.org>
4910
4911         Add missing license blocks to garden-o-matic
4912         https://bugs.webkit.org/show_bug.cgi?id=65654
4913
4914         Reviewed by Dimitri Glazkov.
4915
4916         This patch is entirely boilerplate.
4917
4918         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
4919         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js:
4920         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js:
4921         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders.js:
4922         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
4923         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout_unittests.js:
4924         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
4925         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
4926         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
4927         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
4928         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
4929         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net.js:
4930         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net_unittests.js:
4931         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
4932         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js:
4933         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
4934         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
4935         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js:
4936
4937 2011-08-04  Adam Barth  <abarth@webkit.org>
4938
4939         garden-o-matic needs a way to mock out the network
4940         https://bugs.webkit.org/show_bug.cgi?id=65653
4941
4942         Reviewed by Dimitri Glazkov.
4943
4944         This patch introduces the net module, which is a useful mock point for
4945         abstracting away the network.  Future patches will use this
4946         infrastructure to test some currently untested code.
4947
4948         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js:
4949         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
4950         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js:
4951         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders.js:
4952         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
4953         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
4954         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html:
4955         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net.js: Added.
4956         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net_unittests.js: Added.
4957         (NetworkSimulator):
4958         (NetworkSimulator.prototype.scheduleCallback):
4959         (NetworkSimulator.prototype.runTest):
4960         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
4961         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js:
4962         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
4963         * Scripts/webkitpy/tool/servers/gardeningserver.py:
4964
4965 2011-08-03  Adam Barth  <abarth@webkit.org>
4966
4967         Use failureInfo more pervasively in garden-o-matic
4968         https://bugs.webkit.org/show_bug.cgi?id=65650
4969
4970         Reviewed by Dimitri Glazkov.
4971
4972         These three variable appears in concert together many places.  This
4973         patch unifies the remaining uses into being failureInfo objects, which
4974         saves some marshalling.
4975
4976         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
4977         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
4978         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js:
4979         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
4980         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js:
4981
4982 2011-08-03  Adam Barth  <abarth@webkit.org>
4983
4984         garden-o-matic should be able to update expecations from the details pane
4985         https://bugs.webkit.org/show_bug.cgi?id=65648
4986
4987         Reviewed by Dimitri Glazkov.
4988
4989         The new button queues up the updates, which get executed all at once.
4990
4991         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
4992         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
4993         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
4994
4995 2011-08-03  Adam Barth  <abarth@webkit.org>
4996
4997         Add UI to garden-o-matic for updating expectations
4998         https://bugs.webkit.org/show_bug.cgi?id=65644
4999
5000         Reviewed by Dimitri Glazkov.
5001
5002         The patch plumbs the UI back to the gardening server, but the gardening
5003         server endpoint is just a stub at this point.
5004
5005         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
5006         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
5007         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
5008         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
5009         * Scripts/webkitpy/tool/servers/gardeningserver.py:
5010         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
5011
5012 2011-08-03  Adam Barth  <abarth@webkit.org>
5013
5014         garden-o-matic details view should queue rebaselines for later execution
5015         https://bugs.webkit.org/show_bug.cgi?id=65636
5016
5017         Reviewed by Dimitri Glazkov.
5018
5019         This patch doesn't contain any confirmation UI, but we'll probably want
5020         that at some point.
5021
5022         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
5023         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
5024         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
5025         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
5026
5027 2011-08-03  Kentaro Hara  <haraken@google.com>
5028
5029         Implement EventSender.scalePageBy()
5030         https://bugs.webkit.org/show_bug.cgi?id=58013
5031
5032         Reviewed by Darin Fisher.
5033
5034         Implemented EventSender.scalePageBy(f, x, y), which scales a page by a factor of f
5035         and then sets a scroll position to (x, y). Enabled the tests that had been waiting
5036         for the implementation of EventSender.scalePageBy(f, x, y).
5037
5038         Tests: compositing/scaling/tiled-layer-recursion.html
5039                fast/repaint/scale-page-shrink.html
5040                fast/dom/Element/scale-page-client-rects.html
5041                fast/dom/Range/scale-page-client-rects.html
5042                fast/events/scroll-in-scaled-page-with-overflow-hidden.html
5043                fast/dom/Element/scale-page-bounding-client-rect.html
5044                fast/dom/Range/scale-page-bounding-client-rect.html
5045
5046         * DumpRenderTree/chromium/EventSender.cpp:
5047         (EventSender::EventSender): Added bindings for scalePageBy().
5048         (EventSender::scalePageBy): A wrapper method for scalePage() in WebView.
5049         * DumpRenderTree/chromium/EventSender.h:
5050         * DumpRenderTree/chromium/TestShell.cpp:
5051         (TestShell::resetTestController): Resets the scale factor to 1.
5052
5053 2011-08-03  David Levin  <levin@chromium.org>
5054
5055         Rename WEBKIT_API to WEBKIT_EXPORT in check-webkit-style.
5056         https://bugs.webkit.org/show_bug.cgi?id=65652
5057
5058         Reviewed by Adam Barth.
5059
5060         * Scripts/webkitpy/style/checkers/cpp.py:
5061         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
5062
5063 2011-08-03  Filip Pizlo  <fpizlo@apple.com>
5064
5065         Adding Filip Pizlo to committer list.
5066
5067         Unreviewed.
5068
5069         * Scripts/webkitpy/common/config/committers.py:
5070
5071 2011-08-03  Dimitri Glazkov  <dglazkov@chromium.org>
5072
5073         Add "optimize-expectations" command to webkit-patch.
5074         https://bugs.webkit.org/show_bug.cgi?id=65633
5075
5076         For now, only operates on Chromium test_expectations.txt and does not report errors/warnings.
5077
5078         Reviewed by Adam Barth.
5079
5080         * Scripts/webkitpy/tool/commands/__init__.py: Added command.
5081         * Scripts/webkitpy/tool/commands/expectations.py: Added.
5082
5083 2011-08-03  Dimitri Glazkov  <dglazkov@chromium.org>
5084
5085         Allow multiple bug identifiers in test expectations.
5086         https://bugs.webkit.org/show_bug.cgi?id=65642
5087
5088         Reviewed by Adam Barth.
5089
5090         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Tweaked to allow multiple bug ids.
5091         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added tests.
5092
5093 2011-08-03  Dimitri Glazkov  <dglazkov@chromium.org>
5094
5095         Teach BuilderToPort to distinguish between Release and Debug builders.
5096         https://bugs.webkit.org/show_bug.cgi?id=65628
5097
5098         Reviewed by Adam Barth.
5099
5100         * Scripts/webkitpy/tool/commands/rebaseline.py: Added a smart options object and passed it to port factory.
5101         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py: Added tests.
5102
5103 2011-08-03  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
5104
5105         [Qt] Fix running qmltests for WK2 in the bot
5106         https://bugs.webkit.org/show_bug.cgi?id=65621
5107
5108         Reviewed by Csaba Osztrogonác.
5109
5110         The path for QML_IMPORT_PATH was assuming we were running the WK1 tests,
5111         but now there's the case for WK2 UIProcess tests as well.
5112
5113         * Scripts/run-qtwebkit-tests: add both possible paths to the QML_IMPORT_PATH.
5114
5115 2011-08-02  Dimitri Glazkov  <dglazkov@chromium.org>
5116
5117         Add a way to edit test expectations.
5118         https://bugs.webkit.org/show_bug.cgi?id=64922
5119
5120         Implements a TestExpectationsEditor, which provides a standard way to
5121         update and remove test expectations.
5122
5123         Reviewed by Adam Barth.
5124
5125         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Changed TestExpectationSerializer to skip lines that match no test configurations,
5126             added TestExpectationLine.is_flaky, added BugManager, a simple interface to be fleshed out later, and a TestExpectationsEditor, the big
5127             enchilada.
5128         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Loads of tests to capture desired behavior.
5129
5130 2011-08-03  Andreas Kling  <kling@webkit.org>
5131
5132         [Qt] MiniBrowser: Unbreak load progress indication.
5133         https://bugs.webkit.org/show_bug.cgi?id=65614
5134
5135         Reviewed by Kenneth Rohde Christiansen.
5136
5137         loadProgress() was renamed to loadProgressChanged().
5138
5139         * MiniBrowser/qt/BrowserWindow.cpp:
5140         (BrowserWindow::BrowserWindow):
5141         (BrowserWindow::onLoadProgressChanged):
5142         * MiniBrowser/qt/BrowserWindow.h:
5143
5144 2011-08-03  Martin Robinson  <mrobinson@igalia.com>
5145
5146         Build fix for WebKit2GTK+.
5147
5148         * WebKitTestRunner/GNUmakefile.am: Include libInternals as one of
5149         of the library dependencies of the injected bundle.
5150
5151 2011-08-03  Carlos Garcia Campos  <cgarcia@igalia.com>
5152
5153         [GTK] Install web and plugin processes in libexecdir instead of bindir
5154         https://bugs.webkit.org/show_bug.cgi?id=65600
5155
5156         Reviewed by Martin Robinson.
5157
5158         Set WEBKIT_EXEC_PATH environment variable to use web and plugin
5159         process from current directory instead of the intalled ones.
5160
5161         * MiniBrowser/gtk/GNUmakefile.am:
5162         * MiniBrowser/gtk/main.c:
5163         (main):
5164         * WebKitTestRunner/GNUmakefile.am:
5165         * WebKitTestRunner/gtk/main.cpp:
5166         (main):
5167
5168 2011-08-02  Anders Carlsson  <andersca@apple.com>
5169
5170         Add mac-lion platform to old-run-webkit-tests
5171         https://bugs.webkit.org/show_bug.cgi?id=65577
5172
5173         * Scripts/old-run-webkit-tests:
5174         * Scripts/webkitdirs.pm:
5175         (isLion):
5176
5177 2011-08-02  Alexandru Chiculita  <achicu@adobe.com>
5178
5179         Buildbot: The Properties object has no "hasProperty" method in buildbot 8.3
5180         https://bugs.webkit.org/show_bug.cgi?id=65554
5181
5182         Using "has_key" instead of "hasProperty" in order to support buildbot 8.3.
5183
5184         Reviewed by Adam Roben.
5185
5186         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
5187
5188 2011-08-02  Jochen Eisinger  <jochen@chromium.org>
5189
5190         Implement WebPermissionClient::allowPlugins for layout tests
5191         https://bugs.webkit.org/show_bug.cgi?id=33991
5192
5193         Reviewed by Adam Barth.
5194
5195         * DumpRenderTree/chromium/LayoutTestController.cpp:
5196         (LayoutTestController::LayoutTestController):
5197         (LayoutTestController::setPluginsAllowed):
5198         * DumpRenderTree/chromium/LayoutTestController.h:
5199         * DumpRenderTree/chromium/WebPermissions.cpp:
5200         (WebPermissions::allowPlugins):
5201         (WebPermissions::setPluginsAllowed):
5202         (WebPermissions::reset):
5203         * DumpRenderTree/chromium/WebPermissions.h:
5204
5205 2011-08-02  Adam Barth  <abarth@webkit.org>
5206
5207         TestFailures unit tests leak a global variable
5208         https://bugs.webkit.org/show_bug.cgi?id=65500
5209
5210         Reviewed by Adam Roben.
5211
5212         qunit detects the leak if you check the "noglobals" box on run-unittests.html.
5213
5214         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
5215
5216 2011-08-02  Alexandru Chiculita  <achicu@adobe.com>
5217
5218         CSS Regions build bot should archive and upload output files
5219         https://bugs.webkit.org/show_bug.cgi?id=65200
5220
5221         Until now the buildbots would only upload the result archive if the 'trigger' parameters was specified. The
5222         upload was needed, because the testbots needed to download and use the archive. For CSS Regions we don't have
5223         a testbot yet, but we want to save the archive on the server for manual testing. Added new parameter
5224         in config.json, called "upload" that when set to "true" will force the upload to the buildmaster server.
5225
5226         The "features" argument is now taken into account when determining the name of the archive on the server. That
5227         was needed in order to avoid the CSS Regions Mac buildbot overwrite the SnowLeopard Mac build.
5228
5229         Reviewed by Adam Roben.
5230
5231         * BuildSlaveSupport/build.webkit.org-config/config.json:
5232         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
5233
5234 2011-08-02  Patrick Gansterer  <paroga@webkit.org>
5235
5236         Fix _win32_check_running_pid on 64bit windows python
5237         https://bugs.webkit.org/show_bug.cgi?id=64950
5238
5239         Reviewed by Adam Roben.
5240
5241         Use the correct types for the PROCESSENTRY32 struct.
5242         th32DefaultHeapID is declared as ULONG_PTR and not as DWORD.
5243         Using 32bit for pointers instead of 64bit results in an ERROR_BAD_LENGTH
5244         failure, when calling the Process32First function.
5245         This is covered by the executive unittest.
5246
5247         * Scripts/webkitpy/common/system/executive.py:
5248
5249 2011-08-01  Adam Barth  <abarth@webkit.org>
5250
5251         garden-o-matic should call optimize-baselines when rebaselining tests
5252         https://bugs.webkit.org/show_bug.cgi?id=65499
5253
5254         Reviewed by Dimitri Glazkov.
5255
5256         I took the opportunity to modernize this code to use some of our more
5257         powerful primitives from base.
5258
5259         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js:
5260         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js:
5261         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
5262         * Scripts/webkitpy/tool/servers/gardeningserver.py:
5263         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
5264
5265 2011-08-01  Stephanie Lewis  <slewis@apple.com>
5266
5267         Finish reconfiguration started by Lucas Forschler
5268         Reconfigured slaves:
5269         Removed Leopard Intel Release (Build)
5270         Removed Leopard Intel Release (Tests)
5271         Remove apple-pixel-1
5272         Added Lion slaves from disabled Leopard slaves
5273
5274         Added builders:
5275         Lion Intel Release (Build)
5276         Lion Intel Debug (Build)
5277         Lion Intel Release (Tests)
5278         Lion Intel Release WK2 (Tests)
5279         Lion Intel Debug (Tests)
5280         Lion Intel Debug WK2 (Tests)
5281
5282         Added triggers:
5283         lion-intel-debug-tests
5284         lion-intel-debug-test-wk2
5285         lion-intel-release-tests
5286         lion-intel-release-tests-wk2
5287
5288         Removed trigger:
5289         leopard-intel-release-tests
5290
5291         Added PlatformSpecificScheduler:
5292         mac-lion
5293
5294         Added "mac-lion" to build.py and build_unitttest.py
5295
5296         Reviewed by Mark Rowe.
5297
5298         * BuildSlaveSupport/build.webkit.org-config/config.json:
5299         * Scripts/webkitpy/common/config/build.py:
5300         * Scripts/webkitpy/common/config/build_unittest.py:
5301
5302
5303 2011-08-01  Adam Barth  <abarth@webkit.org>
5304
5305         webkit-patch needs to be able to "optimize" the storage of baselines on disk
5306         https://bugs.webkit.org/show_bug.cgi?id=65418
5307
5308         Reviewed by Dimitri Glazkov.
5309
5310         If we're not careful when rebaselining tests, we can end up with lots
5311         of duplicate expected results files in the tree.  This patch adds a
5312         webkit-patch command that optimizes the storage of expected results on
5313         disk.
5314
5315         This command is similar to deduplicate-tests, except that it can move
5316         test results around rather than just remove duplicate results.
5317
5318         Unfortunately, this problem is very tricky because the baseline search
5319         structure is a hypergraph.  This patch include a huerstic optimizer
5320         that appears to work on a bunch of examples I've tried.  We'll likely
5321         need to refine it as gain experience.
5322
5323         * Scripts/webkitpy/common/system/filesystem.py:
5324         * Scripts/webkitpy/tool/commands/rebaseline.py:
5325
5326 2011-08-01  Dimitri Glazkov  <dglazkov@chromium.org>
5327
5328         Teach TestExpectationSerializer about parsed expectations.
5329         https://bugs.webkit.org/show_bug.cgi?id=65444
5330
5331         This is a bit largish in size, but the essentials are:
5332         * TestExpectationSerializer, given a parsed TestExpectationLine, can now spit out a pretty string.
5333         * We now store original content of the expectation line on TestExpectationLine.
5334         * Lots of tests.
5335
5336         Reviewed by Adam Barth.
5337
5338         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Fixed an issue where we didn't
5339             populate macros correctly, also fixed an issue where we incorrectly returned an empty list
5340             for all configurations.
5341         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added tests for those.
5342         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Removed vestiges of code long gone,
5343             refactored TestExpectationsSerializer to recognize and correctly serialize parsed TestExpectationLine instances,
5344             changed to store and use TestExpectationLine.original_string.
5345         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added tests.
5346
5347 2011-08-01  Adam Barth  <abarth@webkit.org>
5348
5349         Remove unused functionality from gardeningserver
5350         https://bugs.webkit.org/show_bug.cgi?id=65479
5351
5352         Reviewed by Dimitri Glazkov.
5353
5354         We now handle these operations directly from the front-end.  That means
5355         they work even without the gardening server running!
5356
5357         Also, add some missing files to the list of static files.
5358
5359         * Scripts/webkitpy/tool/servers/gardeningserver.py:
5360         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
5361
5362 2011-08-01  Dimitri Glazkov  <dglazkov@chromium.org>
5363
5364         Introduce SpecifierSorter, a thing that knows how specifiers should be ordered.
5365         https://bugs.webkit.org/show_bug.cgi?id=65413
5366
5367         Reviewed by Adam Barth.
5368
5369         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Added SpecifierSorter and converted
5370             existing code to use it.
5371         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added unit tests and cleaned up a bit.
5372
5373 2011-08-01  Adam Barth  <abarth@webkit.org>
5374
5375         Refactor bugzilla.js for use by garden-o-matic
5376         https://bugs.webkit.org/show_bug.cgi?id=65450
5377
5378         Reviewed by Dimitri Glazkov.
5379
5380         This patch refactors bugzilla.js to use the AsynchronousCache and
5381         updates the style to use a module instead of an object.  This patch
5382         then fixes all the existing code that uses this class to use the new
5383         API style.
5384
5385         This main benefit of this patch is we remove the tricky manual caching
5386         and this code is now available to use in garden-o-matic (since the
5387         dependency on Utilities.js is now gone).
5388
5389         I ran all the unit tests and poked around in TestFailures a bit to see
5390         that everything seemed to be working properly.
5391
5392         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js:
5393         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm.js:
5394         (FailingTestsBugForm):
5395         (FailingTestsBugForm.prototype._createBugTitle):
5396         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js:
5397         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm.js:
5398         (FlakyTestBugForm):
5399         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm_unittests.js:
5400         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm.js:
5401         (NewBugForm):
5402         (NewBugForm.prototype.domElement):
5403         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm_unittests.js:
5404         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm.js:
5405         (TestRelatedBugForm):
5406         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm_unittests.js:
5407         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
5408         (ViewController.prototype._displayBuilder.start):
5409         (ViewController.prototype._displayBuilder):
5410         (ViewController.prototype._domForAuxiliaryUIElements):
5411         (ViewController.prototype._domForNewAndExistingBugs.bugzilla.quickSearch):
5412         (ViewController.prototype._domForPossiblyFlakyTests.flakyList.appendChildren):
5413         (ViewController.prototype._domForPossiblyFlakyTests):
5414         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
5415         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
5416         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html:
5417
5418 2011-08-01  Adam Barth  <abarth@webkit.org>
5419
5420         garden-o-matic shouldn't require local server to determine if compile failed
5421         https://bugs.webkit.org/show_bug.cgi?id=65446
5422
5423         Reviewed by Dimitri Glazkov.
5424
5425         This patch adds some basic bindings for the buildbot JSON API.  I
5426         wanted to use Buildbot.js an Builder.js from the original TestFailures,
5427         but they make a number of WebKit-specific assumptions.  We'll
5428         eventually need to reconcile these pieces of code.
5429
5430         I should really add tests for this code, but that will require me to
5431         build a better mock network.  I'll be happy to do that in a follow-up
5432         patch.
5433
5434         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders.js: Added.
5435             - It's very unclear to me whether didFail is correct.  Supposedly
5436               the first entry here is the exit code of the step, which is 0 on
5437               success.  I suspect we'll need to improve this function over time.
5438         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
5439         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
5440         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
5441
5442 2011-08-01  Adam Barth  <abarth@webkit.org>
5443
5444         garden-o-matic should show all failures, even ones that it can't find a regression range for
5445         https://bugs.webkit.org/show_bug.cgi?id=65447
5446
5447         Reviewed by Dimitri Glazkov.
5448
5449         Previously, we just wouldn't display test failures that we couldn't
5450         identify a regression range for (or whose regression range had fallen
5451         off the page).  This patch just displays them at the end of the list,
5452         which is slightly (but not much) better.
5453
5454         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
5455
5456 2011-08-01  Jeff Miller  <jeffm@apple.com>
5457
5458         Windows build fix after r92096.
5459
5460         * TestWebKitAPI/win/InjectedBundleControllerWin.cpp: Use #include, not #import, and don't include Foundation.h.
5461         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj: Add InjectedBundleControllerWin.cpp.
5462
5463 2011-07-31  Anders Carlsson  <andersca@apple.com>
5464
5465         DownloadDecideDestinationCrash WebKit2 API tests crashes
5466         https://bugs.webkit.org/show_bug.cgi?id=65442
5467
5468         Reviewed by Sam Weinig.
5469
5470         Copy the 18-characters.html file to the destination directory so we won't
5471         crash trying to make an URL for it.
5472
5473         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
5474
5475 2011-07-31  Anders Carlsson  <andersca@apple.com>
5476
5477         Fix SpacebarScrolling WebKit2 API test
5478         https://bugs.webkit.org/show_bug.cgi?id=65431
5479
5480         Reviewed by Sam Weinig.
5481
5482         Add a way to do platform-specific bundle initialization, and use it to disable smooth
5483         scrolling. Also make it possible for a test to load the injected bundle without actually running
5484         a test from it.
5485
5486         * TestWebKitAPI/InjectedBundleController.cpp:
5487         (TestWebKitAPI::InjectedBundleController::initialize):
5488         * TestWebKitAPI/InjectedBundleController.h:
5489         * TestWebKitAPI/PlatformUtilities.cpp:
5490         (TestWebKitAPI::Util::createContextWithInjectedBundle):
5491         * TestWebKitAPI/PlatformUtilities.h:
5492         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
5493         * TestWebKitAPI/Tests/WebKit2/SpacebarScrolling.cpp:
5494         (TestWebKitAPI::TEST):
5495         * TestWebKitAPI/mac/InjectedBundleControllerMac.mm: Added.
5496         (TestWebKitAPI::InjectedBundleController::platformInitialize):
5497         * TestWebKitAPI/win/InjectedBundleControllerWin.cpp: Added.
5498         (TestWebKitAPI::InjectedBundleController::platformInitialize):
5499         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
5500
5501 2011-07-31  Dimitri Glazkov  <dglazkov@chromium.org>
5502
5503         Use set comparison to find duplicate or overlapping specifiers in test expectations.
5504         https://bugs.webkit.org/show_bug.cgi?id=65430
5505
5506         Reviewed by Adam Barth.
5507
5508         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Changed to use set comparisons, because length
5509             of a set is wrong.
5510         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Added some unit tests to actually catch
5511             regressions.
5512
5513 2011-07-31  Xan Lopez  <xlopez@igalia.com>
5514
5515         Group dependencies CFLAGS together.
5516
5517         * GNUmakefile.am: ditto.
5518
5519 2011-07-31  Daniel Bates  <dbates@webkit.org>
5520
5521         Attempt to fix the GTK build after <http://trac.webkit.org/changeset/92059>
5522         (https://bugs.webkit.org/show_bug.cgi?id=65419).
5523
5524         * GNUmakefile.am: Add $(CAIRO_CFLAGS) to libWebCoreInternals_la_CPPFLAGS.
5525
5526 2011-07-30  Patrick Gansterer  <paroga@webkit.org>
5527
5528         Fix ports_unittest.py on native windows
5529         https://bugs.webkit.org/show_bug.cgi?id=64959
5530
5531         Reviewed by Adam Barth.
5532
5533         Replace WebKitPort.script_path with WebKitPort.script_shell_command.
5534         The latter includes the interpreter executable required on native windows.
5535
5536         * Scripts/webkitpy/common/config/ports_unittest.py:
5537
5538 2011-07-30  Adam Barth  <abarth@webkit.org>
5539
5540         Properly escape regular expressions.  (Maybe these shouldn't be regular
5541         expressions?)
5542
5543         * Scripts/webkitpy/tool/commands/rebaseline.py:
5544
5545 2011-07-30  Adam Barth  <abarth@webkit.org>
5546
5547         Add support for Win7 to garden-o-matic.
5548
5549         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
5550         * Scripts/webkitpy/tool/commands/rebaseline.py:
5551
5552 2011-07-30  Adam Barth  <abarth@webkit.org>
5553
5554         garden-o-matic regression ranges expand after fast bots start passing
5555         https://bugs.webkit.org/show_bug.cgi?id=65406
5556
5557         Reviewed by Dimitri Glazkov.
5558
5559         Once the gardener has fixed a problem, the fast bots pick up the fix,
5560         which means we end up using the slow bots to compute the regression
5561         range.  Unfortunately, the slow bots give us large regression ranges,
5562         leading to this strange effect where the regression range quality
5563         degrades as fixes roll in.
5564
5565         This patch causes us to remember the previous regression ranges we've
5566         seen and to use that information in the future, as long as it's
5567         consistent with the new information we receive.
5568
5569         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
5570
5571 2011-07-30  Adam Barth  <abarth@webkit.org>
5572
5573         garden-o-matic should have a Refresh button to update its display
5574         https://bugs.webkit.org/show_bug.cgi?id=65403
5575
5576         Reviewed by Dimitri Glazkov.
5577
5578         The refresh button should be faster and less disruptive than the browser's reload button.
5579
5580         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
5581         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
5582
5583 2011-07-30  Adam Barth  <abarth@webkit.org>
5584
5585         garden-o-matic should look for the "patch by" line in revisions ostensibly authored by the commit-queue
5586         https://bugs.webkit.org/show_bug.cgi?id=65231
5587
5588         Reviewed by Dimitri Glazkov.
5589
5590         The human readable name is easier to understand, especially when the patch is landed by the commit-queue.
5591
5592         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
5593
5594 2011-07-30  Adam Barth  <abarth@webkit.org>
5595
5596         garden-o-matic should strikeout revisions that have been rolled out
5597         https://bugs.webkit.org/show_bug.cgi?id=65227
5598
5599         Reviewed by Dimitri Glazkov.
5600
5601         This helps gardeners keep track of which problems have been solved.
5602
5603         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
5604         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
5605         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
5606         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
5607
5608 2011-07-30  Adam Barth  <abarth@webkit.org>
5609
5610         Add missing files to the server.
5611
5612         * Scripts/webkitpy/tool/servers/gardeningserver.py:
5613
5614 2011-07-29  Adam Barth  <abarth@webkit.org>
5615
5616         Actually import the logging module!
5617
5618         * Scripts/webkitpy/common/checkout/scm/git.py:
5619
5620 2011-07-29  Adam Barth  <abarth@webkit.org>
5621
5622         More debug logging.  Maybe we're running the git version of this
5623         function somehow?  (That would explain why it doesn't work!)
5624
5625         * Scripts/webkitpy/common/checkout/scm/git.py:
5626
5627 2011-07-29  Adam Barth  <abarth@webkit.org>
5628
5629         Add some temporary debug logging messages to try to understand why the
5630         windows bots can't figure out their current SVN revision.
5631
5632         * Scripts/webkitpy/common/checkout/scm/svn.py:
5633
5634 2011-07-29  Adam Barth  <abarth@webkit.org>
5635
5636         Chromium try servers should use downstream test_expectations
5637         https://bugs.webkit.org/show_bug.cgi?id=65390
5638
5639         Reviewed by Dirk Pranke.
5640
5641         Otherwise the try servers will report bogus failures.
5642
5643         * Scripts/webkitpy/layout_tests/port/chromium.py:
5644
5645 2011-07-29  Dimitri Glazkov  <dglazkov@chromium.org>
5646
5647         Fix an error in macro collapsing, refactor code to be pretty.
5648         https://bugs.webkit.org/show_bug.cgi?id=65387
5649
5650         My set math was wrong.
5651
5652         Reviewed by Adam Barth.
5653
5654         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Moved collapsing into a classmethod, fixed the math to be correct.
5655         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added unit tests.
5656
5657 2011-07-27  Ryuan Choi  <ryuan.choi@samsung.com>
5658
5659         Style bot complains about EFL-specific forward declaration naming.
5660         https://bugs.webkit.org/show_bug.cgi?id=64680
5661
5662         Exclude EFL-specific forward declarations from indentifiers with underscores checks.
5663
5664         Reviewed by Adam Barth.
5665
5666         * Scripts/webkitpy/style/checkers/cpp.py:
5667         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
5668
5669 2011-07-27  Adam Roben  <aroben@apple.com>
5670
5671         Use wacky !svn/bc URLs consistently in Leaks Viewer
5672
5673         This ensures that any resources loaded via relative URLs will still get pulled from the same
5674         revision of Subversion.
5675
5676         Fixes <http://webkit.org/b/65267> Focus/exclude button glyphs don't load on Leaks Viewer
5677
5678         Reviewed by David Kilzer.
5679
5680         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html: Use the
5681         !svn/bc-style URL for inspector.css, too, just like we do for all other resources we pull
5682         from Subversion.
5683
5684 2011-07-27  Raphael Kubo da Costa  <kubo@profusion.mobi>
5685
5686         [EFL] Drop WorkQueueItemEfl.h in favour of the general WorkQueueItem.h
5687         https://bugs.webkit.org/show_bug.cgi?id=65269
5688
5689         Reviewed by Kent Tamura.
5690
5691         WorkQueueItemEfl.h was being included by efl's
5692         LayoutTestControllerEfl.cpp, however the top-level
5693         LayoutTestController still included WorkQueueItem.h, causing a lot of
5694         type mismatches when running the WorkQueueItem code.
5695
5696         We can stick to the original WorkQueueItem header by making a few
5697         changes to EFL's LayoutTestController.
5698
5699         This should make tests such as
5700         fast/loader/non-deferred-substitute-load.html stop crashing.
5701
5702         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
5703         (LayoutTestController::queueLoad):
5704         * DumpRenderTree/efl/WorkQueueItemEfl.cpp:
5705         (LoadItem::invoke):
5706         (LoadHTMLStringItem::invoke):
5707         (ScriptItem::invoke):
5708         * DumpRenderTree/efl/WorkQueueItemEfl.h: Removed.
5709
5710 2011-07-27  Dimitri Glazkov  <dglazkov@chromium.org>
5711
5712         Re-add unit tests I accidentally zapped in r91878.
5713
5714         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Re-added unit tests.
5715
5716 2011-07-27  Dimitri Glazkov  <dglazkov@chromium.org>
5717
5718         Replace SpecificityCalculator with TestConfiguration-driven logic.
5719         https://bugs.webkit.org/show_bug.cgi?id=65206
5720
5721         Use our newly-acquired ability to expand modifiers into a set of
5722         matching TestConfiguration instances to calculate specificity and
5723         determine whether expectation applies to a given test configuration.
5724
5725         Also, store bug modifier on TestExpectationsLine.
5726
5727         Reviewed by Adam Barth.
5728
5729         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Replaced the logic of calculating specificity with
5730             test configuration matching, removed a bunch of code.
5731         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Removed a bunch of tests that aren't useful anymore.
5732
5733
5734 2011-07-27  Dimitri Glazkov  <dglazkov@chromium.org>
5735
5736         Allow ports to specify their own test expectation specifier macros.
5737         https://bugs.webkit.org/show_bug.cgi?id=65291
5738
5739         Reviewed by Adam Barth.
5740
5741         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Changed SpecificityCalculator to use port macros.
5742         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Ditto.
5743         * Scripts/webkitpy/layout_tests/port/base.py: Added configuration_specifier_macros to retrieve the macros.
5744         * Scripts/webkitpy/layout_tests/port/chromium.py: Overrode it to return Chromium's macros.
5745         * Scripts/webkitpy/layout_tests/port/test.py: Added test macros.
5746
5747 2011-07-27  Dimitri Glazkov  <dglazkov@chromium.org>
5748
5749         Allow TestConfigurationConverter report conversion errors.
5750         https://bugs.webkit.org/show_bug.cgi?id=65287
5751
5752         Also, fix a few miscellaneous testing issues.
5753
5754         Reviewed by Adam Barth.
5755
5756         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Added a way to report unknown modifier errors.
5757         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added a test for reporting errors.
5758         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py: Fixed an error where an invalid TestConfiguration instance
5759             was created. 'default' is not a valid build type.
5760         * Scripts/webkitpy/layout_tests/port/test.py: Ditto. '' is not a valid version.
5761
5762 2011-07-27  Dimitri Glazkov  <dglazkov@chromium.org>
5763
5764         Teach all ports about their test configurations.
5765         https://bugs.webkit.org/show_bug.cgi?id=65278
5766
5767         Each Port object should return a valid set of all TestConfiguration instances,
5768         available for this port.
5769
5770         Reviewed by Philippe Normand.
5771
5772         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Removed TestConfiguration.all_test_configurations,
5773             since it is different for each port.
5774         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Removed related test.
5775         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Changed SpecificityCalculator to take a list of
5776             all test configurations as a param.
5777         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Adjusted arguments accordingly.
5778         * Scripts/webkitpy/layout_tests/port/base.py: Forced all ports to override generation of all test configurations.
5779         * Scripts/webkitpy/layout_tests/port/chromium.py: Implemented _generate_all_test_configurations for the port.
5780         * Scripts/webkitpy/layout_tests/port/gtk.py: Ditto.
5781         * Scripts/webkitpy/layout_tests/port/mac.py: Ditto.
5782         * Scripts/webkitpy/layout_tests/port/qt.py: Ditto.
5783         * Scripts/webkitpy/layout_tests/port/win.py:
5784         * Scripts/webkitpy/layout_tests/port/port_testcase.py: Added an extra test to ensure we're generating a good set.
5785         * Scripts/webkitpy/layout_tests/port/test.py: Implemented a fixed-in-time all_test_configurations for testing.
5786         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py: Added a mock all_test_configurations implementation.
5787         * Scripts/webkitpy/layout_tests/port/win_unittest.py: Added a unit test for the port.
5788
5789 2011-07-27  Benjamin Poulain  <benjamin@webkit.org>
5790
5791         [WK2][Qt] Move from QGraphicsView to Qt Scene Graph
5792         https://bugs.webkit.org/show_bug.cgi?id=62969
5793
5794         Reviewed by Andreas Kling.
5795
5796         Co-authored with Simon Hausmann  <simon.hausmann@nokia.com>
5797
5798         With the scene graph, the interface to have a window on screen is
5799         QSGCanvas. QGraphicsView is not used, and openGL is an enforced
5800         dependency.
5801
5802         Both the BrowserView and the WebKitTest runner are changed from
5803         QGraphicsView to QSGCanvas.
5804
5805         * MiniBrowser/qt/BrowserView.cpp:
5806         (BrowserView::BrowserView):
5807         (BrowserView::resizeEvent):
5808         (BrowserView::view):
5809         * MiniBrowser/qt/BrowserView.h:
5810         * MiniBrowser/qt/BrowserWindow.cpp:
5811         (BrowserWindow::BrowserWindow):
5812         (BrowserWindow::urlChanged):
5813         (BrowserWindow::~BrowserWindow):
5814         * MiniBrowser/qt/BrowserWindow.h:
5815         * MiniBrowser/qt/MiniBrowser.pro:
5816         * MiniBrowser/qt/main.cpp:
5817         (main):
5818         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
5819         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
5820         (WTR::WebView::WebView):
5821         (WTR::PlatformWebView::PlatformWebView):
5822         * WebKitTestRunner/qt/WebKitTestRunner.pro:
5823
5824 2011-07-27  Sheriff Bot  <webkit.review.bot@gmail.com>
5825
5826         Unreviewed, rolling out r91792.
5827         http://trac.webkit.org/changeset/91792
5828         https://bugs.webkit.org/show_bug.cgi?id=65242
5829
5830         It broke NRWT on Qt and GTK (Requested by Ossy_night on
5831         #webkit).
5832
5833         * Scripts/webkitpy/layout_tests/models/test_configuration.py:
5834         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py:
5835         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
5836         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
5837         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
5838         * Scripts/webkitpy/layout_tests/port/test.py:
5839
5840 2011-07-26  Adam Barth  <abarth@webkit.org>
5841
5842         Switch to an older idiom that works on Python 2.5.
5843
5844         * Scripts/webkitpy/layout_tests/models/test_configuration.py:
5845
5846 2011-07-26  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
5847
5848         Add an exception to coding style check for EFL port.
5849         https://bugs.webkit.org/show_bug.cgi?id=65162
5850
5851         Add an exception of whitespace/declaration for EFL port. Because,
5852         '*' operator is next to variable in EFL coding style.
5853
5854         Reviewed by Kent Tamura.
5855
5856         * Scripts/webkitpy/style/checker.py:
5857
5858 2011-07-26  Keith Kyzivat  <keith.kyzivat@nokia.com>
5859
5860         [Qt] Add option to turn on disk caching in QtTestBrowser
5861         https://bugs.webkit.org/show_bug.cgi?id=65007
5862
5863         Reviewed by Noam Rosenthal.
5864
5865         Add menu item and command line option to turn on disk caching in
5866         QtTestBrowser. Simple QNetworkDiskCache added to the page's
5867         QNetworkAccssManager.
5868
5869         * QtTestBrowser/launcherwindow.cpp:
5870         (LauncherWindow::initializeView):
5871         (LauncherWindow::createChrome):
5872         (LauncherWindow::setDiskCache):
5873         * QtTestBrowser/launcherwindow.h:
5874         (WindowOptions::WindowOptions):
5875         * QtTestBrowser/main.cpp:
5876         (LauncherApplication::handleUserOptions):
5877
5878 2011-07-26  Sadrul Habib Chowdhury  <sadrul@chromium.org>
5879
5880         Add support for download='filename' attribute in anchors.
5881         https://bugs.webkit.org/show_bug.cgi?id=64580
5882
5883         Reviewed by Adam Barth.
5884
5885         * DumpRenderTree/chromium/WebViewHost.cpp:
5886         (WebViewHost::loadURLExternally):
5887         * DumpRenderTree/chromium/WebViewHost.h:
5888
5889 2011-07-26  Dan Bernstein  <mitz@apple.com>
5890
5891         Rubber-stamped by Anders Carlsson.
5892
5893         Update DumpRenderTree and API tests for the pictograph font family preference.
5894
5895         * DumpRenderTree/mac/DumpRenderTree.mm:
5896         (resetDefaultsToConsistentValues):
5897         * DumpRenderTree/win/DumpRenderTree.cpp:
5898         (resetDefaultsToConsistentValues):
5899         * TestWebKitAPI/Tests/WebKit2/WKPreferences.cpp:
5900         (TestWebKitAPI::TEST):
5901         * WebKitTestRunner/TestController.cpp:
5902         (WTR::TestController::resetStateToConsistentValues):
5903
5904 2011-07-26  Dimitri Glazkov  <dglazkov@chromium.org>
5905
5906         Replace SpecificityCalculator with TestConfiguration-driven logic.
5907         https://bugs.webkit.org/show_bug.cgi?id=65206
5908
5909         Use our newly-acquired ability to expand modifiers into a set of
5910         matching TestConfiguration instances to calculate specificity and
5911         determine whether expectation applies to a given test configuration.
5912
5913         Also, store bug modifier on TestExpectationsLine.
5914
5915         Reviewed by Adam Barth.
5916
5917         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Added a way to report unknown modifier errors.
5918         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added a test for reporting errors.
5919         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Replaced the logic of calculating specificity with
5920             test configuration matching, removed a bunch of code.
5921         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Removed a bunch of tests that aren't useful anymore.
5922         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py: Fixed an error where an invalid TestConfiguration instance
5923             was created. 'default' is not a valid build type.
5924         * Scripts/webkitpy/layout_tests/port/test.py: Ditto. '' is not a valid version.
5925
5926 2011-07-26  Adam Barth  <abarth@webkit.org>
5927
5928         Increase the information density in garden-o-matic
5929         https://bugs.webkit.org/show_bug.cgi?id=65141
5930
5931         Reviewed by Dimitri Glazkov.
5932
5933         This patch changes lets us show each revision on a single line,
5934         allowing use to show an entire day's worth of commits at once.
5935
5936         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
5937         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
5938         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
5939
5940 2011-07-26  Dimitri Glazkov  <dglazkov@chromium.org>
5941
5942         TestConfigurationConverter should not barf on invalid specifiers.
5943         https://bugs.webkit.org/show_bug.cgi?id=65194
5944
5945         Reviewed by Adam Barth.
5946
5947         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Refactored TestConfigurationConverter.to_config_set a bit,
5948             made it return empty set whenever an invalid specifier is encountered.
5949         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added corresponding tests.
5950
5951 2011-07-26  Dimitri Glazkov  <dglazkov@chromium.org>
5952
5953         Introduce a way to convert between specifier lists and TestConfiguration sets.
5954         https://bugs.webkit.org/show_bug.cgi?id=65094
5955
5956         Add TestConfigurationConverter, a facility to convert:
5957         1) a set of specifiers into a set of TestConfiguration instances;
5958         2) a set of TestConfiguration instances into a list of sets of specifiers.
5959
5960         Reviewed by Adam Barth.
5961
5962         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Added TestConfigurationConverter.
5963         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added tests for it.
5964
5965 2011-07-26  Alexandru Chiculita  <achicu@adobe.com>
5966
5967         CSS Regions build bot should build with "--css-regions" and "--css-exclusion" flags
5968         https://bugs.webkit.org/show_bug.cgi?id=65171
5969
5970         Added "features" parameter in config.json for "builder" objects. The parameter is optional and must be an array. All the items
5971         of the array will be passed to the build-webkit script. The items will automatically be prefixed with "--".
5972
5973         Reviewed by Adam Roben.
5974
5975         * BuildSlaveSupport/build.webkit.org-config/config.json:
5976         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
5977
5978 2011-07-26  Chris Rogers  <crogers@google.com>
5979
5980         Avoid outputting #EOF twice for web audio tests in chromium DRT
5981         https://bugs.webkit.org/show_bug.cgi?id=65201
5982
5983         Reviewed by Dirk Pranke.
5984
5985         * DumpRenderTree/chromium/TestShell.cpp:
5986         (TestShell::dump):
5987
5988 2011-07-26  Adam Barth  <abarth@webkit.org>
5989
5990         Windows bots on build.chromium.org can't figure out the head_svn_revision
5991         https://bugs.webkit.org/show_bug.cgi?id=65156
5992
5993         Reviewed by Tony Chang.
5994
5995         This patch is a bit of a guess as to the issue based on the
5996         investigation in https://bugs.webkit.org/show_bug.cgi?id=65148.  I'm
5997         not sure whether it will fix the issue, but it seems worth a try.
5998
5999         * Scripts/webkitpy/common/checkout/scm/svn.py:
6000
6001 2011-07-26  Carlos Garcia Campos  <cgarcia@igalia.com>
6002
6003         Unreviewed. Fix WebKitGtk build with GTK+ 3.x after r91707.
6004
6005         * DumpRenderTree/gtk/SelfScrollingWebKitWebView.cpp:
6006         (getPreferredSizeMethod):
6007
6008 2011-07-26  MORITA Hajime  <morrita@google.com>
6009
6010         [Chromium][DRT] MockSpellCheck.cpp should be clang clean
6011         https://bugs.webkit.org/show_bug.cgi?id=65166
6012
6013         Reviewed by Kent Tamura.
6014
6015         * DumpRenderTree/chromium/MockSpellCheck.cpp:
6016         (MockSpellCheck::spellCheckWord): -Wsign-compare warning.
6017
6018 2011-07-25  Patrick Gansterer  <paroga@webkit.org>
6019
6020         Ignore line endings in ChangeLogTests
6021         https://bugs.webkit.org/show_bug.cgi?id=64956
6022
6023         Reviewed by Adam Barth.
6024
6025         Split the lines of the ChangeLog text before comparing them.
6026         This fixes 3 tests on windows, since only the content importent.
6027
6028         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
6029         * Scripts/webkitpy/tool/steps/preparechangelog_unittest.py:
6030
6031 2011-07-25  Adam Barth  <abarth@webkit.org>
6032
6033         Add some debug logging to NRWT to diagnose why the windows bots can't determine their SVN revision
6034         https://bugs.webkit.org/show_bug.cgi?id=65148
6035
6036         Reviewed by Tony Chang.
6037
6038         * Scripts/webkitpy/layout_tests/controllers/manager.py:
6039
6040 2011-07-25  Ojan Vafai  <ojan@chromium.org>
6041
6042         Tweak whitespace wrapping to improve readability.
6043         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6044
6045 2011-07-25  Ojan Vafai  <ojan@chromium.org>
6046
6047         Minor tweak to wrap label elements as atomic units.
6048
6049         * TestResultServer/static-dashboards/dashboard_base.js:
6050         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6051
6052 2011-07-25  Ojan Vafai  <ojan@chromium.org>
6053
6054         always show flakiness details on the builder page
6055         https://bugs.webkit.org/show_bug.cgi?id=65139
6056
6057         Reviewed by Tony Chang.
6058
6059         In order to simplify the UI, we hide the flakiness details by default,
6060         but there's no actual use-case for this page without the flakiness details.
6061         So get rid of that option and always show the flakiness details.
6062
6063         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6064
6065 2011-07-25  Ojan Vafai  <ojan@chromium.org>
6066
6067         simplify the UI of the header above the test list table
6068         https://bugs.webkit.org/show_bug.cgi?id=65136
6069
6070         Reviewed by Adam Barth.
6071
6072         Use checkboxes instead of links since this is really about which tests to include.
6073         Ends up using less text and put the checkboxes to the right of the header instead of below it.
6074
6075         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6076         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
6077
6078 2011-07-25  Chris Rogers  <crogers@google.com>
6079
6080         Update new-run-webkit-tests to accomodate binary audio data
6081         https://bugs.webkit.org/show_bug.cgi?id=65135
6082
6083         Reviewed by Dirk Pranke.
6084
6085         * Scripts/webkitpy/layout_tests/port/chromium.py:
6086
6087 2011-07-25  Chris Rogers  <crogers@google.com>
6088
6089         Update Chromium DRT to output binary (instead of base64-encoded) data for web audio testing
6090         https://bugs.webkit.org/show_bug.cgi?id=65039
6091
6092         Reviewed by Tony Chang.
6093
6094         * DumpRenderTree/chromium/LayoutTestController.cpp:
6095         (LayoutTestController::LayoutTestController):
6096         (LayoutTestController::setAudioData):
6097         * DumpRenderTree/chromium/LayoutTestController.h:
6098         (LayoutTestController::audioData):
6099         * DumpRenderTree/chromium/TestEventPrinter.cpp:
6100         (DRTPrinter::handleAudioHeader):
6101         (TestShellPrinter::handleAudioHeader):
6102         * DumpRenderTree/chromium/TestShell.cpp:
6103         (TestShell::dump):
6104
6105 2011-06-07  Martin Robinson  <mrobinson@igalia.com>
6106
6107         Reviewed by Gustavo Noronha Silva.
6108
6109         [GTK] Use WebCore scrollbars for the DRT instead of GtkScrolledWindow
6110         https://bugs.webkit.org/show_bug.cgi?id=62252
6111
6112         Add a new subclass of WebKitWebView that overrides the size request to
6113         return 1 pixel by 1 pixel. This ensure that it operates properly when
6114         not packed into a GtkScrolledWindow. Also construct the WebKitWebView
6115         with self-scrolling set to true.
6116
6117         * DumpRenderTree/gtk/DumpRenderTree.cpp:
6118         (webInspectorInspectWebView): Use the new SelfScrollingWebKitWebview class here.
6119         (createWebView): Ditto.
6120         (main): Pack the widget into a GtkBox instead of a GtkScrolledWindow.
6121         * DumpRenderTree/gtk/SelfScrollingWebKitWebView.cpp: Added.
6122         (self_scrolling_webkit_web_view_class_init):
6123         (self_scrolling_webkit_web_view_init):
6124         (self_scrolling_webkit_web_view_new):
6125         (sizeRequestMethod):
6126         (getPreferredSizeMethod):
6127         * DumpRenderTree/gtk/SelfScrollingWebKitWebView.h: Added.
6128         * GNUmakefile.am: Added SelfScrollingWebKitWebView files to the source list.
6129
6130 2011-07-25  Ojan Vafai  <ojan@chromium.org>
6131
6132         by default hide tests that have passes for all recorded runs
6133         https://bugs.webkit.org/show_bug.cgi?id=65127
6134
6135         Reviewed by Adam Barth.
6136
6137         This is part of simplifying the default views of the dashboard to
6138         make it show less information for the common use cases.
6139
6140         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6141         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
6142
6143 2011-07-25  Ojan Vafai  <ojan@chromium.org>
6144
6145         stop generating expectations.json now that it's unused
6146         https://bugs.webkit.org/show_bug.cgi?id=65130
6147
6148         Reviewed by Adam Barth.
6149
6150         * Scripts/webkitpy/layout_tests/controllers/manager.py:
6151         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
6152         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
6153         * TestResultServer/handlers/menu.py:
6154         * TestResultServer/static-dashboards/dashboard_base.js:
6155         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6156
6157 2011-07-25  Ojan Vafai  <ojan@chromium.org>
6158
6159         don't show builders in the flakiness dashboard that have stopped running a given test suite
6160         https://bugs.webkit.org/show_bug.cgi?id=65126
6161
6162         Reviewed by Mihai Parparita.
6163
6164         * TestResultServer/static-dashboards/dashboard_base.js:
6165
6166 2011-07-25  Adam Barth  <abarth@webkit.org>
6167
6168         Overhaul garden-o-matic frontend to integrate revision history and failures
6169         https://bugs.webkit.org/show_bug.cgi?id=65089
6170
6171         Reviewed by Dimitri Glazkov.
6172
6173         This is a major overhaul of the garden-o-matic frontend.  The new UI
6174         integrates the revision history information with failures.  Failures
6175         appear in context adjacent to the revisions that caused them (which are
6176         highlighted).  Each buildbot receives a collumn to the left, which
6177         indicates which revisions that bot has tested and whether that bot
6178         showed the indicated failures.
6179
6180         Failures are selectable with checkboxes, and can then either be
6181         examined or rebaselined.
6182
6183         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
6184         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js:
6185         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
6186         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
6187         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
6188         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
6189         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js:
6190         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
6191         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js:
6192
6193 2011-07-25  Holger Hans Peter Freyther  <zecke@selfish.org>
6194
6195         [buildbot] Add Qt Linux MIPSEL Configuration
6196         https://bugs.webkit.org/show_bug.cgi?id=65116
6197
6198         Reviewed by Adam Roben.
6199
6200         Introduce a Qt Linux MIPS EL build slave. The configuration
6201         is a build only configuration.
6202
6203         * BuildSlaveSupport/build.webkit.org-config/config.json: Add MIPS buildbot.
6204
6205 2011-07-25  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
6206
6207         [Qt][WK2] Add multi-point touch mocking to MiniBrowser
6208         https://bugs.webkit.org/show_bug.cgi?id=64374
6209
6210         Reviewed by Benjamin Poulain.
6211
6212         Each mouse button generate a touch point. The touch point stay
6213         virtually on screen if the mouse release happened while the Ctrl
6214         modifier was pressed.
6215
6216         * MiniBrowser/qt/MiniBrowserApplication.cpp:
6217         (MiniBrowserApplication::notify):
6218         * MiniBrowser/qt/MiniBrowserApplication.h:
6219
6220 2011-07-25  Balazs Kelemen  <kbalazs@webkit.org>
6221
6222         [Qt][WK2] Use NRWT for Qt-WK2
6223         https://bugs.webkit.org/show_bug.cgi?id=65097
6224
6225         Reviewed by Andreas Kling.
6226
6227         * Scripts/run-webkit-tests:
6228         (useNewRunWebKitTests):
6229
6230 2011-07-24  Sheriff Bot  <webkit.review.bot@gmail.com>
6231
6232         Unreviewed, rolling out r91635.
6233         http://trac.webkit.org/changeset/91635
6234         https://bugs.webkit.org/show_bug.cgi?id=65083
6235
6236         It broke the minimal build (Requested by Ossy___OFFLINE on
6237         #webkit).
6238
6239         * QtTestBrowser/launcherwindow.cpp:
6240         (LauncherWindow::initializeView):
6241         (LauncherWindow::createChrome):
6242         * QtTestBrowser/launcherwindow.h:
6243         (WindowOptions::WindowOptions):
6244         * QtTestBrowser/main.cpp:
6245         (LauncherApplication::handleUserOptions):
6246
6247 2011-07-24  Adam Barth  <abarth@webkit.org>
6248
6249         Rename garden-o-matic "alert" to "infobar" and increase the width to 100%.
6250
6251         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
6252         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
6253         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
6254         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
6255         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
6256         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js:
6257
6258 2011-07-23  Dimitri Glazkov  <dglazkov@chromium.org>
6259
6260         TestConfiguration should be hashable and comparable for equality.
6261         https://bugs.webkit.org/show_bug.cgi?id=65070
6262
6263         Reviewed by Adam Barth.
6264
6265         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Added __hash__ and __eq__ methods.
6266         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added unit tests.
6267
6268 2011-07-23  Dimitri Glazkov  <dglazkov@chromium.org>
6269
6270         Extract TestConfiguration from base and write a few unit tests to capture current behavior.
6271         https://bugs.webkit.org/show_bug.cgi?id=65066
6272
6273         Reviewed by Adam Barth.
6274
6275         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Added.
6276         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added.
6277         * Scripts/webkitpy/layout_tests/port/base.py: Moved code to test_configuration.py
6278
6279 2011-07-23  Alok Priyadarshi  <alokp@chromium.org>
6280
6281         Switching off acceleration for small canvas broke gpu tests
6282         https://bugs.webkit.org/show_bug.cgi?id=65053
6283
6284         Reviewed by Stephen White.
6285
6286         * DumpRenderTree/chromium/WebPreferences.cpp:
6287         (WebPreferences::reset):
6288         (WebPreferences::applyTo):
6289         * DumpRenderTree/chromium/WebPreferences.h:
6290
6291 2011-07-23  Keith Kyzivat  <keith.kyzivat@nokia.com>
6292
6293         [Qt] Add option to turn on disk caching in QtTestBrowser
6294         https://bugs.webkit.org/show_bug.cgi?id=65007
6295
6296         Reviewed by Noam Rosenthal.
6297
6298         Add menu item and command line option to turn on disk caching in
6299         QtTestBrowser. Simple QNetworkDiskCache added to the page's
6300         QNetworkAccssManager.
6301
6302         * QtTestBrowser/launcherwindow.cpp:
6303         (LauncherWindow::initializeView):
6304         (LauncherWindow::createChrome):
6305         (LauncherWindow::setDiskCache):
6306         * QtTestBrowser/launcherwindow.h:
6307         (WindowOptions::WindowOptions):
6308         * QtTestBrowser/main.cpp:
6309         (LauncherApplication::handleUserOptions):
6310
6311 2011-07-22  Ojan Vafai  <ojan@chromium.org>
6312
6313         Don't match 'DEBUG' when trying to filter BUG modifiers.
6314         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6315         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
6316
6317 2011-07-22  Ojan Vafai  <ojan@chromium.org>
6318
6319         Fix issue loading gtests since there's no g_expectations.
6320         Also, make filing a bug file a webkit.org bug, not a crbug since
6321         the link is only shown for layout tests.
6322         * TestResultServer/static-dashboards/dashboard_base.js:
6323         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6324
6325 2011-07-22  Ojan Vafai  <ojan@chromium.org>
6326
6327         pull test expectations into the dashboard from wvn.webkit.org
6328         https://bugs.webkit.org/show_bug.cgi?id=65054
6329
6330         Reviewed by Adam Barth.
6331
6332         Now that svn.webkit.org supports cors headers, we can just grab the txt
6333         file from there instead of generating a JSON file that we upload
6334         to appengine. We'll stop generating the JSON file in a followup patch.
6335
6336         * TestResultServer/static-dashboards/dashboard_base.js:
6337         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6338         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
6339
6340 2011-07-22  Adam Barth  <abarth@webkit.org>
6341
6342         Fix typo so the tool works on a case-sensitive file system.
6343
6344         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
6345
6346 2011-07-22  Adam Barth  <abarth@webkit.org>
6347
6348         garden-o-matic should know what patches are landing
6349         https://bugs.webkit.org/show_bug.cgi?id=64978
6350
6351         Reviewed by Dimitri Glazkov.
6352
6353         This patch adds a display of the recently committed patches to
6354         garden-o-matic.  The UI needs iteration, but this patch establishes the
6355         plumbing.
6356
6357         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
6358             - Teach trac how to extract more information from the commit log.
6359         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
6360             - Add a data attribute for storing the revision.
6361         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
6362             - Add a container for the recent commit history.
6363         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
6364             - CSS for displaying non-ugly commit history.
6365         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
6366             - Wire up trac to the UI container.
6367         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
6368             - A basic, compact display for a list of commits.
6369
6370 2011-07-22  Ojan Vafai  <ojan@chromium.org>
6371
6372         Fix bug when switching from layout-tests to another test-type if
6373         ToT webkit.org is the selected group.
6374         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6375
6376 2011-07-22  Ojan Vafai  <ojan@chromium.org>
6377
6378         Add back in line accidentally deleted in http://trac.webkit.org/changeset/91612.
6379         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6380
6381 2011-07-22  Ojan Vafai  <ojan@chromium.org>
6382
6383         simplify the flakiness dashboard to only show modifiers/expectations for the current builder
6384         https://bugs.webkit.org/show_bug.cgi?id=65049
6385
6386         Reviewed by Tony Chang.
6387
6388         This communicates less information, but probably makes the dashboard more usable.
6389         Also make it show that bug lines that don't point to bug numbers still
6390         show up in the bugs column (e.g. BUG_TONY) and stop showing the extra/missing
6391         columns in any view. The data is still used, but only in the view that updates
6392         the test_expectations.txt file.
6393
6394         * TestResultServer/static-dashboards/dashboard_base.js:
6395         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6396         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
6397
6398 2011-07-21  Ojan Vafai  <ojan@chromium.org>
6399
6400         hook in the rest of the bots to the dashboard now they run NRWT
6401         https://bugs.webkit.org/show_bug.cgi?id=64985
6402
6403         Reviewed by Adam Barth.
6404
6405         * TestResultServer/static-dashboards/builders.js:
6406
6407 2011-07-22  Alexey Proskuryakov  <ap@apple.com>
6408
6409         Adding Tim Horton to contributor list for Bugzilla auto-completion. He is CC'ed quite frequently.
6410
6411         Unreviewed.
6412
6413         * Scripts/webkitpy/common/config/committers.py:
6414
6415 2011-07-22  Michal Pakula vel Rutka  <m.pakula@samsung.com>
6416
6417         [EFL] ewk_frame_hit_test_new enchancement
6418         https://bugs.webkit.org/show_bug.cgi?id=64260
6419
6420         Apply changes done in ewk_frame_hit_test_new in EWebLauncher:
6421         Replacing Ewk_Hit_Test_Result_Context structure 'flags'
6422         in hit test output to by enum 'context'.
6423
6424         Reviewed by Antonio Gomes.
6425
6426         * EWebLauncher/main.c:
6427         (on_key_down):
6428
6429 2011-07-21  Adam Roben  <aroben@apple.com>
6430
6431         Fix typo in TestFailures
6432
6433         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
6434
6435 2011-07-21  Ojan Vafai  <ojan@chromium.org>
6436
6437         simplify gtest display now that we strip modifiers from the JSON
6438         https://bugs.webkit.org/show_bug.cgi?id=64990
6439
6440         Reviewed by Adam Barth.
6441
6442         Now that names are normalized, we can remove all the code that
6443         handles gtest name changes (e.g. for adding modifiers like DISABLED_).
6444         Instead, if you try to list a test with a modifier in it, we need
6445         to strip the modifier so we get the normalized value.
6446
6447         We also get rid of the concept of extra/missing expectations for gtests.
6448         In a patch soon, we'll stop showing extra/missing expectations from the UI
6449         entirely and only leave it for the special updating test_expectations.txt
6450         view of the dashboard, which doesn't apply to gtests.
6451
6452         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6453         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
6454
6455 2011-07-21  Adam Barth  <abarth@webkit.org>
6456
6457         Refactor Trac.js for use in garden-o-matic
6458         https://bugs.webkit.org/show_bug.cgi?id=64998
6459
6460         Reviewed by Adam Roben.
6461
6462         This patch refactors Trac.js to match the architecture and style of
6463         garden-o-matic.  I've also factored AsynchronousCache out of Trac (and
6464         moved it to base) because it was redundant with some other code we
6465         already had.
6466
6467         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm.js:
6468         (FailingTestsBugForm):
6469         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js:
6470         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
6471         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
6472         (ViewController.prototype._displayBuilder.start):
6473         (ViewController.prototype._displayBuilder):
6474         (ViewController.prototype._domForRegressionRange.trac.getCommitDataForRevisionRange):
6475         (ViewController.prototype._domForRegressionRange):
6476         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js:
6477         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
6478         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html:
6479         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
6480         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
6481
6482 2011-07-21  Ojan Vafai  <ojan@chromium.org>
6483
6484         Fix issue throwing a 500 to the bots. We were modifying
6485         the dictionary during an iteration over it's keys and would get
6486         RuntimeError: dictionary changed size during iteration
6487
6488         * TestResultServer/model/jsonresults.py:
6489
6490 2011-07-21  Adam Barth  <abarth@webkit.org>
6491
6492         ORWT fails to consume some NRWT arguments
6493         https://bugs.webkit.org/show_bug.cgi?id=64988
6494
6495         Reviewed by Adam Roben.
6496
6497         * Scripts/old-run-webkit-tests:
6498
6499 2011-07-21  Adam Barth  <abarth@webkit.org>
6500
6501         Merge TestFailures and garden-o-matic unit tests harness
6502         https://bugs.webkit.org/show_bug.cgi?id=64984
6503
6504         Reviewed by Adam Roben.
6505
6506         One unit testing framework is better than two.
6507
6508         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-more-unittests.html: Removed.
6509         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
6510
6511 2011-07-21  Adam Barth  <abarth@webkit.org>
6512
6513         Rubber-stamped by Adam Roben.
6514
6515         Move garden-o-matic files into the same directory as TestFailures to
6516         make it easier to integrate the codebases.
6517
6518         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/base.js.
6519         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js.
6520         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js.
6521         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout_unittests.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/checkout_unittests.js.
6522         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/config.js.
6523         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon-green.png: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-green.png.
6524         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon-red.png: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-red.png.
6525         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/index.html.
6526         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/main.css.
6527         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/main.js.
6528         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/partytime.gif: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/partytime.gif.
6529         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/results.js.
6530         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js.
6531         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-more-unittests.html: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html.
6532         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js.
6533         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js.
6534         * Scripts/webkitpy/tool/commands/abstractlocalservercommand.py:
6535         * Scripts/webkitpy/tool/commands/gardenomatic.py:
6536         * Scripts/webkitpy/tool/servers/data/gardeningserver: Removed.
6537         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js: Removed.
6538         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js: Removed.
6539         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js: Removed.
6540         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout_unittests.js: Removed.
6541         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js: Removed.
6542         * Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-green.png: Removed.
6543         * Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-red.png: Removed.
6544         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html: Removed.
6545         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css: Removed.
6546         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js: Removed.
6547         * Scripts/webkitpy/tool/servers/data/gardeningserver/partytime.gif: Removed.
6548         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js: Removed.
6549         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js: Removed.
6550         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html: Removed.
6551         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js: Removed.
6552         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js: Removed.
6553         * Scripts/webkitpy/tool/servers/gardeningserver.py:
6554
6555 2011-07-19  Ojan Vafai  <ojan@chromium.org>
6556
6557         fix links on the flakiness dashboard
6558         https://bugs.webkit.org/show_bug.cgi?id=64832
6559
6560         Reviewed by Tony Chang.
6561
6562         Make them actual anchor tags and linkify the test name instead of
6563         adding an extra link. Also simplify the update view since we now
6564         show the linkified test name.
6565
6566         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6567
6568 2011-07-21  Adam Barth  <abarth@webkit.org>
6569
6570         webkit.org builders should upload results to test-results AppEngine instance
6571         https://bugs.webkit.org/show_bug.cgi?id=64979
6572
6573         Reviewed by Adam Roben.
6574
6575         The master just needs to pass some flags to NRWT.  This patch also lets
6576         ORWT consume the flags (ignoring them) so we can pass the flags
6577         unconditionally.
6578
6579         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
6580         * Scripts/old-run-webkit-tests:
6581
6582 2011-07-21  Ojan Vafai  <ojan@chromium.org>
6583
6584         normalize gtest names in the result json
6585         https://bugs.webkit.org/show_bug.cgi?id=64971
6586
6587         Reviewed by Eric Seidel.
6588
6589         This will make the resulting JSON smaller and make the
6590         flakiness dashboard much more sane (e.g. we won't need to
6591         show you the FLAKY_ version and the normal version).
6592
6593         * TestResultServer/model/jsonresults.py:
6594         * TestResultServer/model/jsonresults_unittest.py:
6595
6596 2011-07-21  Eric Seidel  <eric@webkit.org>
6597
6598         Unreviewed.  Reenable ORWT for the --leaks bot.
6599         Something is still wrong with the leaks summarizing
6600         but I'm about to head out on vacation for 2 weeks,
6601         so it's best to just revert back to ORWT until I'm back.
6602
6603         * Scripts/run-webkit-tests:
6604         (usingLeaks):
6605         (useNewRunWebKitTests):
6606
6607 2011-07-21  Ojan Vafai  <ojan@chromium.org>
6608
6609         cleanup jsonresults.py style in preparation for making some changes
6610         https://bugs.webkit.org/show_bug.cgi?id=64968
6611
6612         Reviewed by Adam Barth.
6613
6614         No code changes. Unittest still passes. Mostly just removing
6615         useless comments and 80 character wrapping.
6616
6617         * TestResultServer/model/jsonresults.py:
6618         * TestResultServer/model/jsonresults_unittest.py:
6619
6620 2011-07-21  Adam Barth  <abarth@webkit.org>
6621
6622         Add missing column header.
6623
6624         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
6625
6626 2011-07-21  Eric Seidel  <eric@webkit.org>
6627
6628         Unreviewed.  Just fixing (and unittesting) a previous typo.
6629
6630         Fix typo in print_leaks_summary regexp which was causing
6631         leaks summarizing to fail on the --leaks bot.
6632
6633         * Scripts/webkitpy/layout_tests/port/mac.py:
6634         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
6635
6636 2011-07-21  Adam Barth  <abarth@webkit.org>
6637
6638         garden-o-matic should be able to rebaseline many tests at once
6639         https://bugs.webkit.org/show_bug.cgi?id=64963
6640
6641         Reviewed by Dimitri Glazkov.
6642
6643         This patch contains some UI for rebaselining tests in bulk.  I don't
6644         expect this UI to be the final UI, but the underlying infrastructure
6645         will be useful.
6646
6647         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js:
6648         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
6649         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
6650         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
6651         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
6652
6653 2011-07-21  Adam Barth  <abarth@webkit.org>
6654
6655         garden-o-matic has trouble when the same test fails in different ways on diffrent bots
6656         https://bugs.webkit.org/show_bug.cgi?id=64830
6657
6658         Unreviewed.  (Missing the patch is blocking me from fixing the tree.)
6659
6660         This patch is just some minor bug fixes.  One of the bugs results from
6661         the sadness of having two representations (a list and a space-separated
6662         string) for the set of failure types.
6663
6664         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
6665         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
6666         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
6667         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
6668
6669 2011-07-20  Ryosuke Niwa  <rniwa@webkit.org>
6670
6671         Buildbot marks a nrwt bot red when tests are missing results
6672         https://bugs.webkit.org/show_bug.cgi?id=64812
6673
6674         Reviewed by Adam Barth.
6675
6676         The bug was caused by multiple expressions matching on the single output.
6677         Fixed it by exiting the loop as soon as one expression matches.
6678
6679         Because the regular expression for 'failures' is most general,
6680         moved it to the end of the list to avoid it catching other cases.
6681
6682         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
6683
6684 2011-07-20  Chang Shu  <cshu@webkit.org>
6685
6686         Adding myself to the reviewers list; No review needed.
6687
6688         * Scripts/webkitpy/common/config/committers.py:
6689
6690 2011-07-20  Eric Seidel  <eric@webkit.org>
6691
6692         new-run-webkit-tests --leaks does not restart DumpRenderTree often enough
6693         https://bugs.webkit.org/show_bug.cgi?id=64912
6694
6695         Reviewed by Adam Barth.
6696
6697         The bug was that set_default_option only works if the option's value
6698         is "None".  So I fixed --batch-size to default to None instead of 0
6699         (which kinda makes sense).
6700
6701         * Scripts/webkitpy/layout_tests/controllers/worker.py:
6702         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
6703
6704 2011-07-20  Eric Seidel  <eric@webkit.org>
6705
6706         Make kill-old-processes kill httpd on mac
6707         https://bugs.webkit.org/show_bug.cgi?id=64898
6708
6709         Reviewed by Adam Barth.
6710
6711         * BuildSlaveSupport/kill-old-processes:
6712          - The line endings seem to have flipped again,
6713            Thus making this diff gigantic.  The only line
6714            changed is the "httpd" line.  I suspect we'll need
6715            to set the line ending style from an svn checkout.
6716
6717 2011-07-20  Ojan Vafai  <ojan@chromium.org>
6718
6719         show a list of average test times in the treemap
6720         https://bugs.webkit.org/show_bug.cgi?id=64899
6721
6722         Reviewed by Adam Roben.
6723
6724         For now, you click a link and it replaces the treemap
6725         with a reverse sorted list of directories based on
6726         average test runtime within that directory.
6727
6728         To cut some of the noise, we don't show directories
6729         that only have one test in them or where the average
6730         runtime is less than 100ms.
6731
6732         * TestResultServer/static-dashboards/treemap.html:
6733
6734 2011-07-20  Eric Seidel  <eric@webkit.org>
6735
6736         Teach build.webkit.org how to identify leaks in NRWT output
6737         https://bugs.webkit.org/show_bug.cgi?id=64891
6738
6739         Reviewed by Adam Barth.
6740
6741         This was the last remaining piece blocking turning on
6742         NRWT for the leaks bot, so I did that in this change too.
6743         If for some reason we see trouble on the bots, I'll
6744         revert that part of this change.
6745
6746         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
6747         * Scripts/run-webkit-tests:
6748
6749 2011-07-20  Ojan Vafai  <ojan@chromium.org>
6750
6751         make the treemap display more human friendly
6752         https://bugs.webkit.org/show_bug.cgi?id=64896
6753
6754         Reviewed by Tony Chang.
6755
6756         * TestResultServer/static-dashboards/treemap.html:
6757
6758 2011-07-20  Eric Seidel  <eric@webkit.org>
6759
6760         Move HttpLock to using a FileSystem object
6761         https://bugs.webkit.org/show_bug.cgi?id=64885
6762
6763         Reviewed by Adam Barth.
6764
6765         I made a typo in my previous change, taking the value
6766         of read_text_file(pid_file) and setting it to lock_pid_file
6767         instead of current_pid.  Fixed now. :)
6768
6769         In order to test my new change I had to overhaul the unittests
6770         for this class to create a separate set of tests which work off
6771         of Mock objects instead of the real filesystem.
6772
6773         Since Executive doesn't yet wrap os.getpid() I added a FIXME
6774         in several places where we're currently calling os.getpid().  I
6775         felt adding Executive.getpid was outside of the scope of this change
6776         but once it exists some of this code will be much simpler to mock.
6777
6778         * Scripts/webkitpy/layout_tests/port/http_lock.py:
6779         * Scripts/webkitpy/layout_tests/port/http_lock_unittest.py:
6780         * Scripts/webkitpy/tool/mocktool.py:
6781
6782 2011-07-20  Adam Roben  <aroben@apple.com>
6783
6784         Fix typo in TestFailures's Bugzilla constants
6785
6786         This was preventing the "Version" field from being pre-filled when filing new bugs.
6787
6788         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBugzilla.js:
6789
6790 2011-07-20  Eric Seidel  <eric@webkit.org>
6791
6792         Move HttpLock to using a FileSystem object
6793         https://bugs.webkit.org/show_bug.cgi?id=64885
6794
6795         Reviewed by Adam Barth.
6796
6797         There is no functional change here.  Just moving off
6798         of direct os.foo calls to their filesystem equivalent.
6799
6800         This makes it possible to now unittest HttpLock using a
6801         MockFileSystem instead of needing to talk to the filesystem directly.
6802
6803         This also will have changed any port test cases which were using a
6804         mock file system to now have an HttpLock based on a mock filesystem
6805         instead.  I saw no change in the unittests so it's possible
6806         we're not testing that configuration (or more likely that it "just works").
6807
6808         * Scripts/webkitpy/layout_tests/port/base.py:
6809         * Scripts/webkitpy/layout_tests/port/http_lock.py:
6810         * Scripts/webkitpy/layout_tests/port/http_lock_unittest.py:
6811
6812 2011-07-20  Adam Barth  <abarth@webkit.org>
6813
6814         NRWT results uses too much space per run
6815         https://bugs.webkit.org/show_bug.cgi?id=64845
6816
6817         Reviewed by Ojan Vafai.
6818
6819         Removing these files saves about 750K per build.
6820
6821         * Scripts/webkitpy/layout_tests/controllers/manager.py:
6822
6823 2011-07-20  Zan Dobersek  <zandobersek@gmail.com>
6824
6825         [Gtk] [NRWT] Xvfb produces a lot of stderr output
6826         https://bugs.webkit.org/show_bug.cgi?id=64669
6827
6828         Reviewed by Philippe Normand.
6829
6830         Redirect the stderr output of the driver's Xvfb process
6831         to /dev/null to suppress unneeded output.
6832
6833         * Scripts/webkitpy/layout_tests/port/gtk.py:
6834
6835 2011-07-19  Eric Seidel  <eric@webkit.org>
6836
6837         parse-malloc-history always exits 1, causing NRWT to fail
6838         https://bugs.webkit.org/show_bug.cgi?id=64835
6839
6840         Reviewed by Adam Barth.
6841
6842         ORWT just never checked the return code. :)
6843
6844         * Scripts/parse-malloc-history:
6845         (main):
6846         * Scripts/webkitpy/layout_tests/port/mac.py:
6847
6848 2011-07-19  Kevin Ollivier  <kevino@theolliviers.com>
6849
6850         [wx] Unreviewed build fix, don't add the debug prefix as 2.9 no longer uses one.
6851
6852         * waf/build/wxpresets.py:
6853
6854 2011-07-19  Eric Seidel  <eric@webkit.org>
6855
6856         new-run-webkit-tests should only enable MallocStackLogging for DRT
6857         https://bugs.webkit.org/show_bug.cgi?id=64792
6858
6859         Reviewed by Adam Barth.
6860
6861         The previous code would enable it for all servers launched
6862         by the port, which included Apache, the python websocket server
6863         as well as ImageDiff.  Now only DumpRenderTree will have
6864         MallocStackLogging enabled or the GuardMalloc library injected.
6865
6866         I also cleaned up the websocket_server code to use filesystem
6867         while I was in it.
6868
6869         I also made DRT restart every 1000 tests when running with
6870         --leaks enabled.  I believe this made the --leaks run slightly
6871         faster, but it still takes over an hour on my machine. :(
6872
6873         * Scripts/webkitpy/layout_tests/controllers/worker.py:
6874         * Scripts/webkitpy/layout_tests/port/base.py:
6875         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
6876         * Scripts/webkitpy/layout_tests/port/gtk.py:
6877         * Scripts/webkitpy/layout_tests/port/mac.py:
6878         * Scripts/webkitpy/layout_tests/port/qt.py:
6879         * Scripts/webkitpy/layout_tests/port/server_process.py:
6880         * Scripts/webkitpy/layout_tests/port/webkit.py:
6881         * Scripts/webkitpy/layout_tests/servers/http_server.py:
6882         * Scripts/webkitpy/layout_tests/servers/websocket_server.py:
6883
6884 2011-07-19  Ojan Vafai  <ojan@chromium.org>
6885
6886         remove the concept of platform fallbacks
6887         https://bugs.webkit.org/show_bug.cgi?id=64829
6888
6889         Reviewed by Adam Barth.
6890
6891         In the process, wrote tests for this code and fixed a pretty major
6892         bug (now covered by the test).
6893
6894         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6895         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
6896
6897 2011-07-19  Ojan Vafai  <ojan@chromium.org>
6898
6899         cleanup some of the PLATFORM logic in the flakiness dashboard
6900         https://bugs.webkit.org/show_bug.cgi?id=64821
6901
6902         Reviewed by Adam Barth.
6903
6904         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6905         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
6906
6907 2011-07-19  Ojan Vafai  <ojan@chromium.org>
6908
6909         cleanup some of the PLATFORM logic in the flakiness dashboard
6910         https://bugs.webkit.org/show_bug.cgi?id=64821
6911
6912         Reviewed by Adam Barth.
6913
6914         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6915         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
6916
6917 2011-07-19  Dimitri Glazkov  <dglazkov@chromium.org>
6918
6919         Store line number on TestExpectationLine.
6920         https://bugs.webkit.org/show_bug.cgi?id=64800
6921
6922         Reviewed by Adam Barth.
6923
6924         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Refactored to store line number on TestExpectationLine instances.
6925
6926 2011-07-19  Ojan Vafai  <ojan@chromium.org>
6927
6928         fix flakiness dashboard for XP bots with XP in the name instead of WIN
6929         https://bugs.webkit.org/show_bug.cgi?id=64826
6930
6931         Reviewed by Mihai Parparita.
6932
6933         * TestResultServer/static-dashboards/flakiness_dashboard.html:
6934         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
6935
6936 2011-07-19  Cary Clark  <caryclark@google.com>
6937
6938         add myself to committers.py
6939         https://bugs.webkit.org/show_bug.cgi?id=64816
6940
6941         No review needed.
6942
6943         * Scripts/webkitpy/common/config/committers.py:
6944
6945 2011-07-19  Daniel Bates  <dbates@webkit.org>
6946
6947         Rename variable isEmptyTextPatch in svn-{apply, unapply}
6948         https://bugs.webkit.org/show_bug.cgi?id=64648
6949
6950         Reviewed by Adam Roben.
6951
6952         Rename the variable isEmptyTextPatch to hasTextChunks, which is
6953         more descriptive and makes its usage read well in control statements.
6954
6955         * Scripts/svn-apply:
6956         (patch):
6957         * Scripts/svn-unapply:
6958         (patch):
6959
6960 2011-07-19  Adam Roben  <aroben@apple.com>
6961
6962         Make TestFailures show existing bugs and a new bug link for flaky tests
6963
6964         Fixes <http://webkit.org/b/63728> TestFailures page should make it easy to file bugs about
6965         flaky tests
6966         and
6967         <http://webkit.org/b/63830> TestFailures page doesn't show related bugs for possibly-flaky
6968         tests, but should
6969
6970         Reviewed by Sam Weinig.
6971
6972         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm.js:
6973         Added. This is what's used to file new bugs about flaky tests.
6974         (FlakyTestBugForm): Calls up to the base class, stores the arguments, and sets our
6975         title, description, and URL.
6976
6977         (FlakyTestBugForm.prototype._createBugDescription):
6978         (FlakyTestBugForm.prototype._createBugTitle):
6979         These do what they say.
6980
6981         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm_unittests.js:
6982         Added. Tests of the above.
6983
6984         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
6985         (#failure-history > li):
6986         (#possibly-flaky-tests > li):
6987         (#failure-history > li, #possibly-flaky-tests > li):
6988         (#possibly-flaky-tests > li > :first-child):
6989         (.flakiness-examples-list):
6990         Styling to account for the new elements.
6991
6992         (.expandable):
6993         (.expanded > .expandable):
6994         Generalized the .flakiness-example-list code to a generic .expandable class. Removed the
6995         transition properties because we can't transition to/from 'auto' (but really would like
6996         to!).
6997
6998         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
6999         (ViewController.prototype._displayBuilder): Moved code to create the FailingTestsBugForm
7000         here from _domForNewAndExistingBugs. Updated for changes to _domForPossiblyFlakyTests.
7001         (ViewController.prototype._domForNewAndExistingBugs): Changed to take a NewBugForm as an
7002         argument instead of creating one ourselves.
7003         (ViewController.prototype._domForPossiblyFlakyTests): Changed to take all analyzed builds as
7004         an argument instead of just the number of analyzed builds. Put the examples list inside a
7005         container <div>, which is also used to hold the new/existing bugs UI. Removed code to deal
7006         with animating the height of the examples list; we don't do this anymore because it's hard
7007         to make it work correctly with the asynchronous loading of existing bugs. When we populate
7008         the examples list, also set up the new/existing bug UI.
7009
7010         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
7011         FlakyTestBugForm.
7012
7013         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
7014         Pull in FlakyTestBugForm and its tests.
7015
7016 2011-07-19  Ryosuke Niwa  <rniwa@webkit.org>
7017
7018         Buildbot marks a nrwt bot red when tests are missing results
7019         https://bugs.webkit.org/show_bug.cgi?id=64812
7020
7021         Reviewed by Tony Chang.
7022
7023         Added "missing results" label for tests that are missing results.
7024
7025         buildbot Now reports shows a label "X missing results" on a run when X tests were missing
7026         expected results on that run.
7027
7028         Also rephrased "X failed" and "X were flaky" to "X failures" and "X flakes".
7029
7030         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
7031
7032 2011-07-19  Adam Roben  <aroben@apple.com>
7033
7034         Extract some of FailingTestsBugForm's code into a base class
7035
7036         Prep work for fixing <http://webkit.org/b/63728> TestFailures page should make it easy to
7037         file bugs about flaky tests
7038
7039         Reviewed by Sam Weinig.
7040
7041         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm.js:
7042         Moved BugzillaConstants to new WebKitBugzilla file. Moved a bunch of other code from here to
7043         TestRelatedBugForm.js.
7044
7045         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js:
7046         Moved some tests to TestRelatedBugForm_unittests.js.
7047
7048         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm.js: Added.
7049         (TestRelatedBugForm):
7050         (TestRelatedBugForm.prototype.domElement):
7051         (TestRelatedBugForm.prototype._computeOperatingSystem):
7052         (TestRelatedBugForm.prototype._computePlatform):
7053         Code came from FailingTestsBugForm.
7054
7055         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm_unittests.js:
7056         Added. Tests came from FailingTestsBugForm_unittests.js
7057
7058         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBugzilla.js:
7059         Added. Code came from FailingTestsBugForm.js.
7060
7061         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
7062         TestRelatedBugForm.js and WebKitBugzilla.js. Moved Bugzilla.js out of the list of files that
7063         need to be pulled in early for parsing reasons.
7064
7065         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
7066         Pull in TestRelatedBugForm and tests and WebKitBugzilla.
7067
7068 2011-07-19  Sam Weinig  <sam@webkit.org>
7069
7070         Remove obsolete Mac NRWT test bots
7071         https://bugs.webkit.org/show_bug.cgi?id=64806
7072
7073         Reviewed by Adam Roben.
7074
7075         Remove "Leopard Intel Release (NRWT)" and "SnowLeopard Intel Release (NRWT)"
7076         which have been disconnected for some time and have been obsoleted by
7077         the main testers using NRWT.
7078
7079         * BuildSlaveSupport/build.webkit.org-config/config.json:
7080
7081 2011-07-19  Sam Weinig  <sam@webkit.org>
7082
7083         webkit-patch rebaseline crashes on use
7084         https://bugs.webkit.org/show_bug.cgi?id=64775
7085
7086         Reviewed by Adam Roben.
7087
7088         Fix incorrectly spelled attribute name.
7089
7090         * Scripts/webkitpy/common/net/layouttestresults.py:
7091         The attribute is test_name, not filename.
7092
7093         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
7094         Add test for tests_matching_failure_types.
7095
7096 2011-07-18  Eric Seidel  <eric@webkit.org>
7097
7098         new-run-webkit-tests should support --leaks
7099         https://bugs.webkit.org/show_bug.cgi?id=63832
7100
7101         Reviewed by Dirk Pranke.
7102
7103         This may not be sufficient to actually transition over the leaks bot,
7104         but this is a huge step in the right direction.
7105
7106         I had to make parse-malloc-history understand being passed more than
7107         one file (to avoid the silly cat | nonsense in old-run-webkit-tests).
7108
7109         I removed some dead code relating to previous iterations of our crash detection.
7110
7111         I created a new class "LeakDetector" to encapsulate all this logic.
7112         Eventually we should consider pulling that class out of mac.py and
7113         sharing with other ports.  However given that ORWT has had
7114         --leaks support on Mac for almost 7 years and no other port has added
7115         it, leaves me to believe we're in no rush to move LeakDetector.
7116
7117         I've tested --leaks locally.  I suspect there are more bugs to shake out
7118         but it seems to work well enough to start.
7119
7120         I also added support for --guard-malloc, but have not tested it much.  It
7121         should be viewed as experimental at this time.
7122
7123         I also fixed various os.path uses to self._filesystem as I was reading
7124         through the various files to understand how best to fix this bug.
7125
7126         * Scripts/old-run-webkit-tests:
7127         (parseLeaksandPrintUniqueLeaks):
7128         * Scripts/parse-malloc-history:
7129         (main):
7130         * Scripts/webkitpy/common/system/crashlogs.py:
7131         * Scripts/webkitpy/layout_tests/controllers/manager.py:
7132         * Scripts/webkitpy/layout_tests/controllers/worker.py:
7133         * Scripts/webkitpy/layout_tests/port/base.py:
7134         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
7135         * Scripts/webkitpy/layout_tests/port/gtk.py:
7136         * Scripts/webkitpy/layout_tests/port/mac.py:
7137         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
7138         * Scripts/webkitpy/layout_tests/port/server_process.py:
7139         * Scripts/webkitpy/layout_tests/port/webkit.py:
7140         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
7141
7142 2011-07-19  Adam Barth  <abarth@webkit.org>
7143
7144         garden-o-matic should have a "Triage Failures" button for iterating over failures
7145         https://bugs.webkit.org/show_bug.cgi?id=64769
7146
7147         Reviewed by Eric Seidel.
7148
7149         This patch adds some global static state and refactors things a bit so
7150         that we can iterate over all the failures in the details pane.
7151
7152         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
7153             - Add a generic callback iterator to iterate through a series of
7154               callbacks. We use this to iterate through the failures we want to
7155               display the details of.
7156         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
7157             - Add a toolbar for the results summary and change the buttons to
7158               be real buttons.
7159         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
7160             - Make the detail pane 75% of the window.
7161             - Make the buttons pretty.
7162             - Allow the failure type badges to apply to the details pane as well.
7163         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
7164             - Restructure how we bring up the details pane so we can iterate
7165               through a bunch of failures.
7166             - Remove the transition between results details because it's
7167               annoying when you want to click through a bunch of failures.
7168         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
7169             - Add some more structure to the title bar for the results details
7170               so it's clearer which results we're currently displaying.
7171
7172 2011-07-19  Adam Barth  <abarth@webkit.org>
7173
7174         garden-o-matic should live update as conditions change on the buildbot
7175         https://bugs.webkit.org/show_bug.cgi?id=64646
7176
7177         Reviewed by Eric Seidel.
7178
7179         This patch causes the garden-o-matic display to update every 10 minutes
7180         to show the latest failures.  This patch required a moderately large
7181         refactoring of main.js to make the display incremental instead of
7182         all-at-once.
7183
7184         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
7185         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
7186         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
7187         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
7188
7189 2011-07-18  Tab Atkins  <jackalmage@gmail.com>
7190
7191         Tab Atkins is not in committers.py (as a non-committer)
7192         https://bugs.webkit.org/show_bug.cgi?id=64770
7193
7194         Reviewed by Eric Seidel.
7195
7196         * Scripts/webkitpy/common/config/committers.py:
7197
7198 2011-07-18  Ojan Vafai  <ojan@chromium.org>
7199
7200         Load ToT chromium bots by default in the flakiness dashboard
7201         https://bugs.webkit.org/show_bug.cgi?id=64756
7202
7203         Reviewed by Adam Barth.
7204
7205         For now, these are the most useful bots to look at. Eventually,
7206         we may want to load the build.webkit.org bots by default once
7207         most of the bots there upload results.
7208
7209         * TestResultServer/static-dashboards/dashboard_base.js:
7210
7211 2011-07-13  Jon Honeycutt  <jhoneycutt@apple.com>
7212
7213         Focus and selection events are not fired when a <select>'s selection
7214         changes
7215         https://bugs.webkit.org/show_bug.cgi?id=64504
7216         <rdar://problem/9319881>
7217
7218         Reviewed by Alice Liu.
7219
7220         * DumpRenderTree/AccessibilityController.h:
7221         Added m_notificationsEventHook for addNotificationListener().
7222         m_allEventsHook will now be used for setLogAccessibilityEvents().
7223
7224         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
7225         (AccessibilityController::AccessibilityController):
7226         Initialize m_notificationsEventHook.
7227         (AccessibilityController::~AccessibilityController):
7228         Turn off logging of all accessibility events. If
7229         m_notificationsEventHook is non-null, unhook it.
7230         (logEventProc):
7231         Add handling of EVENT_OBJECT_SELECTION.
7232         (AccessibilityController::setLogAccessibilityEvents):
7233         If the state of logging is not changing, return early. If we're turning
7234         off logging, unhook m_allEventsHook, and zero it out. Otherwise, add a
7235         hook for all events.
7236         (AccessibilityController::addNotificationListener):
7237         Use m_notificationsEventHook rather than m_allEventsHook.
7238
7239 2011-07-18  Eric Seidel  <eric@webkit.org>
7240
7241         webkit-patch apply-attachment does not work when not called from the root of the checkout
7242         https://bugs.webkit.org/show_bug.cgi?id=64751
7243
7244         Unreviewed.  Just updating unit test results after bug 64751.
7245
7246         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
7247         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
7248
7249 2011-07-18  Adam Barth  <abarth@webkit.org>
7250
7251         simplejson has trouble on chromium-linux
7252         https://bugs.webkit.org/show_bug.cgi?id=64757
7253
7254         Reviewed by Eric Seidel.
7255
7256         Use the native JSON, if available.
7257
7258         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
7259
7260 2011-07-18  Ojan Vafai  <ojan@chromium.org>
7261
7262         update the flakiness dashboard to understand the new platforms/formats in test_expectations
7263         https://bugs.webkit.org/show_bug.cgi?id=64743
7264
7265         Reviewed by Adam Barth.
7266
7267         The test expectations format changed a long time ago and the flakiness dashboard
7268         was never updated to match.
7269
7270         * TestResultServer/static-dashboards/dashboard_base.js:
7271         * TestResultServer/static-dashboards/flakiness_dashboard.html:
7272         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
7273         * TestResultServer/static-dashboards/timeline_explorer.html:
7274
7275 2011-07-18  Eric Seidel  <eric@webkit.org>
7276
7277         webkit-patch apply-attachment does not work when not called from the root of the checkout
7278         https://bugs.webkit.org/show_bug.cgi?id=64751
7279
7280         Reviewed by Adam Barth.
7281
7282         Last week I removed a os.chdir in EnsureWorkingDirectoryClean, which
7283         is a step that we run for most commands (as one of the earliest steps).
7284         EnsureWorkingDirectoryClean was incorrectly calling os.chdir to change
7285         the CWD to the checkout root.  This magically made a bunch of otherwise
7286         wrong code work.
7287
7288         When I realized that apply-attachment no longer worked as expected today,
7289         I went and fixed the bug, then realizing that we were not testing
7290         what the cwd was when running various commands.  I fixed our MockExecutive
7291         to always log what the cwd is and fixed a whole bunch of places
7292         where we needed to be setting the cwd.
7293
7294         Hopefully this will solve our cwd problems once and for-all, and webkit-patch
7295         will again correctly work when called from any directory (including outside
7296         of a webkit checkout).
7297
7298         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
7299         * Scripts/webkitpy/common/system/executive.py:
7300         * Scripts/webkitpy/common/system/workspace_unittest.py:
7301         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
7302         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
7303         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
7304         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
7305         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
7306         * Scripts/webkitpy/tool/bot/irc_command.py:
7307         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
7308         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
7309         * Scripts/webkitpy/tool/commands/download_unittest.py:
7310         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
7311         * Scripts/webkitpy/tool/commands/queues.py:
7312         * Scripts/webkitpy/tool/commands/queues_unittest.py:
7313         * Scripts/webkitpy/tool/commands/queuestest.py:
7314         * Scripts/webkitpy/tool/mocktool.py:
7315         * Scripts/webkitpy/tool/servers/gardeningserver.py:
7316         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
7317         * Scripts/webkitpy/tool/steps/build.py:
7318         * Scripts/webkitpy/tool/steps/checkstyle.py:
7319         * Scripts/webkitpy/tool/steps/editchangelog.py:
7320         * Scripts/webkitpy/tool/steps/preparechangelog.py:
7321         * Scripts/webkitpy/tool/steps/preparechangelogforrevert.py:
7322         * Scripts/webkitpy/tool/steps/runtests.py:
7323         * Scripts/webkitpy/tool/steps/steps_unittest.py:
7324         * Scripts/webkitpy/tool/steps/update.py:
7325
7326 2011-07-18  Adam Barth  <abarth@webkit.org>
7327
7328         Increase information garden-o-matic information density by switching to a table
7329         https://bugs.webkit.org/show_bug.cgi?id=64642
7330
7331         Reviewed by Dimitri Glazkov.
7332
7333         The old layout couldn't handle the large number of failures we had
7334         today.  This patch switches garden-o-matic to a table-based layout,
7335         which can handle many more failures gracefully.  (I expect we'll need
7336         more UI iterations.)
7337
7338         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
7339         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
7340         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
7341         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
7342
7343 2011-07-18  Adam Barth  <abarth@webkit.org>
7344
7345         garden-o-matic should be able to determine when compile breaks
7346         https://bugs.webkit.org/show_bug.cgi?id=64190
7347
7348         Reviewed by Dimitri Glazkov.
7349
7350         This patch adds a red-ish box to the top of the page whenever there is
7351         a compile error on the bots.  The box automatically opens and closes as
7352         appropriate and links to the waterfall display.  In the future, we
7353         might want to compute a regression range.
7354
7355         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js:
7356             - Add the build-only bots to the config.  We use these to check
7357               whether the build failed, which is faster than waiting for the
7358               tester bots to cycle.
7359         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
7360             - Add DOM for the alert bar.
7361         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
7362             - CSS to support the alert bar.
7363         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
7364             - Wiring up events to poll the buildbot to see whether compile has
7365               failed.
7366         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
7367             - Infrastructure for fetching and parsing the buildbot status JSON
7368               blob.  This code could be better factored for testability.  :(
7369         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
7370             - UI for displaying compile errors.
7371         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
7372
7373 2011-07-18  Sheriff Bot  <webkit.review.bot@gmail.com>
7374
7375         Unreviewed, rolling out r91132 and r91135.
7376         http://trac.webkit.org/changeset/91132
7377         http://trac.webkit.org/changeset/91135
7378         https://bugs.webkit.org/show_bug.cgi?id=64681
7379
7380         Broke GTK and Chromium debug bots (Requested by rniwa on
7381         #webkit).
7382
7383         * DumpRenderTree/AccessibilityController.h:
7384         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
7385         (AccessibilityController::AccessibilityController):
7386         (AccessibilityController::~AccessibilityController):
7387         (logEventProc):
7388         (AccessibilityController::setLogAccessibilityEvents):
7389         (AccessibilityController::addNotificationListener):
7390
7391 2011-07-17  Philippe Normand  <pnormand@igalia.com>
7392
7393         test-webkitpy failing with Python 2.5
7394         https://bugs.webkit.org/show_bug.cgi?id=64594
7395
7396         Reviewed by Eric Seidel.
7397
7398         Updated simplejson to version 2.1.6 that fixes an issue when
7399         dumping slashes. Also use cgi.parse_qs instead of the urlparse
7400         version to be compatible with Python2.5. Verified the fix with
7401         test-webkitpy on python 2.5 and python 2.6.
7402
7403         * Scripts/webkitpy/thirdparty/simplejson/README.txt:
7404         * Scripts/webkitpy/thirdparty/simplejson/__init__.py:
7405         * Scripts/webkitpy/thirdparty/simplejson/_speedups.c:
7406         (json_PyOS_string_to_double):
7407         (_convertPyInt_AsSsize_t):
7408         (_convertPyInt_FromSsize_t):
7409         (ascii_escape_char):
7410         (ascii_escape_unicode):
7411         (ascii_escape_str):
7412         (raise_errmsg):
7413         (join_list_unicode):
7414         (join_list_string):
7415         (_build_rval_index_tuple):
7416         (scanstring_str):
7417         (scanstring_unicode):
7418         (py_scanstring):
7419         (py_encode_basestring_ascii):
7420         (scanner_dealloc):
7421         (scanner_traverse):
7422         (scanner_clear):
7423         (_parse_object_str):
7424         (_parse_object_unicode):
7425         (_parse_array_str):
7426         (_parse_array_unicode):
7427         (_parse_constant):
7428         (_match_number_str):
7429         (_match_number_unicode):
7430         (scan_once_str):
7431         (scan_once_unicode):
7432         (scanner_call):
7433         (scanner_new):
7434         (scanner_init):
7435         (encoder_new):
7436         (encoder_init):
7437         (encoder_call):
7438         (_encoded_const):
7439         (encoder_encode_float):
7440         (encoder_encode_string):
7441         (_steal_list_append):
7442         (encoder_listencode_obj):
7443         (encoder_listencode_dict):
7444         (encoder_listencode_list):
7445         (encoder_dealloc):
7446         (encoder_traverse):
7447         (encoder_clear):
7448         (init_speedups):
7449         * Scripts/webkitpy/thirdparty/simplejson/decoder.py:
7450         * Scripts/webkitpy/thirdparty/simplejson/encoder.py:
7451         * Scripts/webkitpy/thirdparty/simplejson/ordered_dict.py: Added.
7452         * Scripts/webkitpy/thirdparty/simplejson/scanner.py:
7453         * Scripts/webkitpy/thirdparty/simplejson/tool.py: Added.
7454         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
7455
7456 2011-07-17  Dimitri Glazkov  <dglazkov@chromium.org>
7457
7458         Rename ModifierMatcher to SpecificityCalculator.
7459         https://bugs.webkit.org/show_bug.cgi?id=64660
7460
7461         It's a little longer than before, but it is much clearer.
7462
7463         Reviewed by Adam Barth.
7464
7465         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Renamed and adjusted
7466             all callsites, also rewrote the comments.
7467         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Adjusted callsites
7468             and renamed tests.
7469
7470 2011-07-15  Adam Barth  <abarth@webkit.org>
7471
7472         gardening server should proxy buildbot status for garden-o-matic
7473         https://bugs.webkit.org/show_bug.cgi?id=64588
7474
7475         Reviewed by Eric Seidel.
7476
7477         This information will be used to detect build breaks.  I took this
7478         opportunity to make ChromiumBuildBot non-static.
7479
7480         * Scripts/webkitpy/common/host.py:
7481         * Scripts/webkitpy/tool/commands/rebaseline.py:
7482         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
7483         * Scripts/webkitpy/tool/mocktool.py:
7484         * Scripts/webkitpy/tool/servers/gardeningserver.py:
7485         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
7486
7487 2011-07-15  Julien Chaffraix  <jchaffraix@webkit.org>
7488
7489         [NRWT] Add support for --no-http
7490         https://bugs.webkit.org/show_bug.cgi?id=64564
7491
7492         Reviewed by Dirk Pranke.
7493
7494         Added support for --no-http, which disables both HTTP and websockets tests.
7495         It also matches the old-run-webkit-tests behavior if --force is used.
7496
7497         * Scripts/webkitpy/layout_tests/controllers/manager.py:
7498         Fixed HTTP_SUBDIR and WEBSOCKET_SUBDIR as tests do not start with a leading separator.
7499         We check if --no-http is set and add the HTTP / websockets tests to the skipped list prior to looking
7500         at the expectation file. Fixed the  _test_requires_lock function to use the same code path to determine
7501         what is worth have an HTTP lock as --no-http to avoid badness.
7502
7503         * Scripts/webkitpy/layout_tests/port/test.py: Added 2 new tests to our mock filesystem to validate that
7504         we do skip properly HTTP / websocket tests inside platform/.
7505
7506         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
7507         Added tests that we properly skip all the tests.
7508
7509         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
7510         Added checks for the command line arguments.
7511
7512 2011-07-13  Jon Honeycutt  <jhoneycutt@apple.com>
7513
7514         Focus and selection events are not fired when a <select>'s selection
7515         changes
7516         https://bugs.webkit.org/show_bug.cgi?id=64504
7517         <rdar://problem/9319881>
7518
7519         Reviewed by Alice Liu.
7520
7521         * DumpRenderTree/AccessibilityController.h:
7522         Added m_notificationsEventHook for addNotificationListener().
7523         m_allEventsHook will now be used for setLogAccessibilityEvents().
7524
7525         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
7526         (AccessibilityController::AccessibilityController):
7527         Initialize m_notificationsEventHook.
7528         (AccessibilityController::~AccessibilityController):
7529         Turn off logging of all accessibility events. If
7530         m_notificationsEventHook is non-null, unhook it.
7531         (logEventProc):
7532         Add handling of EVENT_OBJECT_SELECTION.
7533         (AccessibilityController::setLogAccessibilityEvents):
7534         If the state of logging is not changing, return early. If we're turning
7535         off logging, unhook m_allEventsHook, and zero it out. Otherwise, add a
7536         hook for all events.
7537         (AccessibilityController::addNotificationListener):
7538         Use m_notificationsEventHook rather than m_allEventsHook.
7539
7540 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
7541
7542         Refactor TestExpectationModel to use TestExpectationLine as data item.
7543         https://bugs.webkit.org/show_bug.cgi?id=64635
7544
7545         This is a bit largish in scope. Does the following things:
7546
7547         1) Adds "path" member to TestExpectationLine to hold normalized path to test, computed at parsing,
7548            and changes code that used Port.normalize_test_name to rely on TestExpectationLine.path. As a result, TestExpectationModel no longer
7549            needs to have any port knowledge.
7550
7551         2) Adds "create_passing_expectation" class method to TestExpectationLine to generate a pristine passing expectation out of a test name,
7552            and changes TestExpectations._process_tests_without_expectations to use it, thus eliminating the need for a special API entry point.
7553            Now all expectations are added to the model in the same way!
7554
7555         3) Changes TestExpectationModel's main test index to store a tuple consisting of line number and TestExpectationLine instance.
7556
7557         Reviewed by Adam Barth.
7558
7559         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Refactored code.
7560
7561 2011-07-15  Adam Roben  <aroben@apple.com>
7562
7563         Teach TestFailures how to detect interrupted build steps
7564
7565         Fixes <http://webkit.org/b/64619> TestFailures page thinks all tests passed in
7566         http://build.webkit.org/builders/SnowLeopard%20Intel%20Release%20(WebKit2%20Tests)/builds/13401
7567
7568         Reviewed by Daniel Bates.
7569
7570         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
7571         (Builder.prototype.getNumberOfFailingTests): If the build step has a result code of 4, the
7572         build step was interrupted. Treat it as an error (by returning a failureCount of -1).
7573
7574         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
7575         Added a test for the above.
7576
7577         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
7578         (LayoutTestResultsLoader.prototype.start): Bumped the cache version to evict old, buggy
7579         cache data that was afflicted by the above bug.
7580
7581 2011-07-15  Eric Seidel  <eric@webkit.org>
7582
7583         new-run-webkit-tests crashes on Apple's Windows port when trying to launch Apache
7584         https://bugs.webkit.org/show_bug.cgi?id=64533
7585
7586         Reviewed by Adam Roben.
7587
7588         Blind removal of Chromium-specific code (now that Chromium does not use this file).
7589
7590         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
7591
7592 2011-07-15  Adam Roben  <aroben@apple.com>
7593
7594         Make TestFailures's list of possibly-flaky tests not so tall
7595
7596         Fixes <http://webkit.org/b/64618> TestFailures page's list of flaky tests takes up way too
7597         much room
7598
7599         Reviewed by Daniel Bates.
7600
7601         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
7602         Removed a no-longer-used .flakiness-example-separator selector.
7603
7604         (#failure-history > li):
7605         (#possibly-flaky-tests > li):
7606         (#failure-history > li, #possibly-flaky-tests > li):
7607         Use a lot less padding for the possibly-flaky tests.
7608
7609 2011-07-15  Tony Chang  <tony@chromium.org>
7610
7611         [chromium] land linux selection color layout test
7612         https://bugs.webkit.org/show_bug.cgi?id=64631
7613
7614         Reviewed by Ojan Vafai.
7615
7616         I had removed this code in r76620 because I thought it wasn't used.
7617         Turns out there was a layout test in the chromium tree that never
7618         got migrated that used it.  Re-add the code and land the layout
7619         test (I will remove it from the chromium tree soon).
7620
7621         * DumpRenderTree/chromium/LayoutTestController.cpp:
7622         (LayoutTestController::LayoutTestController):
7623         (LayoutTestController::forceRedSelectionColors):
7624         * DumpRenderTree/chromium/LayoutTestController.h:
7625
7626 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
7627
7628         Revert http://trac.webkit.org/changeset/91091, since Python 2.5
7629         does not support enumerate() start argument.
7630
7631         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Reverted.
7632
7633 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
7634
7635         Clean up test_expectations.py after refactorings.
7636         https://bugs.webkit.org/show_bug.cgi?id=64620
7637
7638         * Renamed all variables holding TestExpectationLine to expectation_line to avoid confusion with
7639           its sub-part, the actual expectation.
7640         * Renamed all references to options to modifier to eliminate dual terminology.
7641         * Made a bunch of parser constants and changed all callsites to use them.
7642         * Various other minor clean-ups.
7643
7644         Reviewed by Adam Barth.
7645
7646         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Cleaned up stuff.
7647         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Changed callsites after cleanup.
7648         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Ditto.
7649
7650 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
7651
7652         Move expectation parsing code to TestExpectationParser.
7653         https://bugs.webkit.org/show_bug.cgi?id=64605
7654
7655         This is a somewhat mechanical move, with two interesting bits:
7656
7657         1) TestExpectationParser.parse methods renamed to tokenize, to better
7658         reflect what they do
7659
7660         2) TestExpectationLine now carries all of its info, from tokens to parsed data,
7661         and even the list of tests that it matches (a line may refer to more than one test).
7662
7663         Reviewed by Adam Barth.
7664
7665         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Moved parsing-related TestExpectaions methods
7666             to TestExpectationParser, added more members to TestExpectationLine to carry parsed info, renamed existing parse methods
7667             to "tokenize", changed callsites to use new code.
7668         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Changed unit tests to reflect renames.
7669
7670 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
7671
7672         Switch to use Python enumerate function to enumerate line numbers.
7673         https://bugs.webkit.org/show_bug.cgi?id=64602
7674
7675         Reviewed by Adam Barth.
7676
7677         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Changed to use enumerate.
7678
7679 2011-07-15  Martin Robinson  <mrobinson@igalia.com>
7680
7681         Build fixes for WebKit2. Ensure that all generated sources are
7682         on nodist primaries, that they are on forward declared variables
7683         so that BUILT_SOURCES is calculated properly and that zlib is
7684         included during linking (for WOFF support).
7685
7686         * WebKitTestRunner/GNUmakefile.am:
7687
7688 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
7689
7690         Store error and warning information on TestExpectationLine.
7691         https://bugs.webkit.org/show_bug.cgi?id=64565
7692
7693         Keeping errors and warnings on the TestExpectationLine instance allows us to decouple storing errors
7694         from various parsing and validation mechanisms and have more flexibility in reporting and understanding the origin of the errors.
7695
7696         Reviewed by Adam Barth.
7697
7698         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Added TestExpectationLine.warnings list to keep track of non-fatal errors,
7699             converted the code to add errors and warnings to corresponding TestExpectationLine instances, removed the code that used to store
7700             this info on TestExpectations. In the process, had to refactor ModifierMatcher a bit to take in a TestExpdectationLine instance.
7701         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Changed ModifierMatcher callsite.
7702
7703 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
7704
7705         Plumb the use of TestExpectationLine deeper, clean up.
7706         https://bugs.webkit.org/show_bug.cgi?id=64559
7707
7708         Instead of carrying various bits of TestExpectationLine, plumb it down to its consumers,
7709         also cleaning up names and remove an unused TestExpectations._get_options_list member.
7710
7711         Reviewed by Adam Barth.
7712
7713         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
7714
7715 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
7716
7717         Remove the notion of TestExpectationLine.valid, start storing parsing errors in expectations themselves.
7718         https://bugs.webkit.org/show_bug.cgi?id=64554
7719
7720         This moves us toward the world where errors are collected on the expectations, which allows us to
7721         easily enumerate them, keep association with the point of origin, and freely pass TestExpectationLine instances around.
7722
7723         Also eliminate the validator idea, since validation is a context-dependent concept and has to be decoupled from initial
7724         parsing.
7725
7726         Reviewed by Adam Barth.
7727
7728         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Removed TestExpectation.valid, validator,
7729             changed TestExpectationParser to collect errors in TestExpectationLine, refactored surrounding code.
7730         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Changed tests to accommodate changes.
7731         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Ditto.
7732
7733 2011-07-15  Adam Roben  <aroben@apple.com>
7734
7735         Rename TestFailureBugForm to FailingTestsBugForm
7736
7737         The new name will match better with a forthcoming FlakyTestBugForm class.
7738
7739         Fixes <http://webkit.org/b/64598> TestFailures page's TestFailureBugForm class has a bad
7740         name
7741
7742         Reviewed by Daniel Bates.
7743
7744         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm.js: Renamed from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm.js.
7745         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js: Renamed from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm_unittests.js.
7746
7747         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
7748         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
7749         Updated for renames.
7750
7751         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Ditto, and
7752         reordered <script>s to put files which others depend on for parsing first.
7753
7754 2011-07-15  Adam Barth  <abarth@webkit.org>
7755
7756         NRWT stores the Chromium revision number in full_results.json
7757         https://bugs.webkit.org/show_bug.cgi?id=64586
7758
7759         I have no earthly idea how to test this change.
7760
7761         * Scripts/webkitpy/layout_tests/controllers/manager.py:
7762         * Scripts/webkitpy/layout_tests/port/base.py:
7763
7764 2011-07-14  Ojan Vafai  <ojan@chromium.org>
7765
7766         fix remaining style issues in the static-dashboards directory
7767         https://bugs.webkit.org/show_bug.cgi?id=64561
7768
7769         Reviewed by Adam Barth.
7770
7771         * TestResultServer/static-dashboards/aggregate_results.html:
7772         * TestResultServer/static-dashboards/builders.js:
7773         * TestResultServer/static-dashboards/dashboard_base.js:
7774         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
7775         * TestResultServer/static-dashboards/timeline_explorer.html:
7776         * TestResultServer/static-dashboards/treemap.html:
7777
7778 2011-07-14  Yuta Kitamura  <yutak@chromium.org>
7779
7780         WebSocket: Introduce pywebsocket-0.6b2
7781         https://bugs.webkit.org/show_bug.cgi?id=64534
7782
7783         Reviewed by Kent Tamura.
7784
7785         * Scripts/webkitpy/thirdparty/__init__.py:
7786
7787 2011-07-14  Dimitri Glazkov  <dglazkov@chromium.org>
7788
7789         Introduce TestExpectationsModel, split out of TestExpectations.
7790         https://bugs.webkit.org/show_bug.cgi?id=64531
7791
7792         This is a simple split-and-make-work refactoring, a first step among many.
7793
7794         Reviewed by Adam Barth.
7795
7796         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Moved all model-related members
7797             out of TestExpectations and into TestExpectationsModel.
7798
7799 2011-07-14  Eric Seidel  <eric@webkit.org>
7800
7801         Move webkitpy off of loose mocks
7802         https://bugs.webkit.org/show_bug.cgi?id=64508
7803
7804         Unreviewed.  Fixing a test which fails under test-webkitpy --all (but not in a normal run).
7805
7806         * Scripts/webkitpy/common/checkout/scm/git.py:
7807
7808 2011-07-14  Adam Barth  <abarth@webkit.org>
7809
7810         garden-o-matic should have a "rebaseline" button
7811         https://bugs.webkit.org/show_bug.cgi?id=64446
7812
7813         Reviewed by Ojan Vafai.
7814
7815         This patch adds a basic Rebaseline button that copies the baselines
7816         displayed in the results pane into the appropriate directory in your
7817         working copy.
7818
7819         There are two main limitations:
7820
7821         1) There is no UI for actually committing the baselines.
7822
7823         2) The baselines are not optimized for redundancy (meaning you can have
7824            identical baselines in both chromium-mac and chromium-win).
7825
7826         * Scripts/webkitpy/tool/commands/rebaseline.py:
7827             - Turns out we need to create the directory for the baseline if it doesn't exist yet.
7828         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js:
7829             - Add an programatic API to call the server.
7830         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
7831             - Add the rebaseline button itself.
7832         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
7833             - Change the CSS so that the Rebaseline and Close buttons can
7834               appear in the normal order in the DOM.
7835         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
7836             - Bind the event and translate the parameters.
7837             - Hide/show the rebaseline button, as appropriate.
7838         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
7839             - Add some helpful utility functions for manipulating failure types.
7840         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
7841
7842 2011-07-14  Ryosuke Niwa  <rniwa@webkit.org>
7843
7844         REGRESSION: webkit-patch roll-chromium-deps is broken
7845         https://bugs.webkit.org/show_bug.cgi?id=64568
7846
7847         Reviewed by Dirk Pranke
7848
7849         Use _filesystem.join.
7850
7851         * Scripts/webkitpy/common/checkout/checkout.py:
7852         * Scripts/webkitpy/common/checkout/checkout_unittests.py:
7853
7854 2011-07-14  Noam Rosenthal  <noam.rosenthal@nokia.com>
7855
7856         Adding myself to the reviewers list; No review needed.
7857
7858         * Scripts/webkitpy/common/config/committers.py:
7859
7860 2011-07-14  Adam Roben  <aroben@apple.com>
7861
7862         Don't use Element.prototype.classList in TestFailures
7863
7864         Safari 5 doesn't support it.
7865
7866         Fixes <http://webkit.org/b/64550> Can't expand flaky tests on TestFailures page in Safari 5
7867
7868         Reviewed by Daniel Bates.
7869
7870         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
7871         (Element.prototype.hasStyleClass):
7872         (Element.prototype.addStyleClass):
7873         (Element.prototype.removeStyleClass):
7874         (Element.prototype.toggleStyleClass):
7875         Added these helper functions which simulate classList functionality.
7876
7877         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities_unittests.js:
7878         Added. Tests for the above.
7879
7880         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
7881         (ViewController.prototype._domForPossiblyFlakyTests): Changed to use
7882         toggleStyleClass/hasStyleClass instead of classList.
7883
7884         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
7885         Added Utilities_unittests.js, and reordered the tested files to be in
7886         roughly dependency order (i.e., the lowest-level files are imported
7887         and tested first).
7888
7889 2011-07-14  Eric Seidel  <eric@webkit.org>
7890
7891         NRWT doesn't store the svn revision in full_results.json on chromium-win
7892         https://bugs.webkit.org/show_bug.cgi?id=64492
7893
7894         Unreviewed.  Just fixing my test-webkitpy regression.
7895
7896         Fix the unit tests.  Unfortunately scm does not use a filesystem
7897         object so we can't control the result of detect_scm_system.
7898         When detect_scm_system would fail, we would log, which would
7899         cause all passing_run integration tests to fail.
7900
7901         * Scripts/webkitpy/common/checkout/scm/scm.py:
7902         * Scripts/webkitpy/layout_tests/controllers/manager.py:
7903         * Scripts/webkitpy/layout_tests/port/base.py:
7904         * Scripts/webkitpy/layout_tests/port/test.py:
7905         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
7906
7907 2011-07-14  Eric Seidel  <eric@webkit.org>
7908
7909         NRWT doesn't store the svn revision in full_results.json on chromium-win
7910         https://bugs.webkit.org/show_bug.cgi?id=64492
7911
7912         Reviewed by Ojan Vafai.
7913
7914         This should fix the bug.  Unfortunately this code is currently
7915         impossible to test since it's impossible to mock detect_scm_system at the
7916         moment.  I started re-writing scm.detection.py to be mockable, but decided that
7917         was best left for another day.
7918
7919         * Scripts/webkitpy/layout_tests/controllers/manager.py:
7920         * Scripts/webkitpy/layout_tests/port/base.py:
7921
7922 2011-07-14  Ojan Vafai  <ojan@chromium.org>
7923
7924         fix coding style of dashboard_base.js
7925         https://bugs.webkit.org/show_bug.cgi?id=64545
7926
7927         Reviewed by Adam Roben.
7928
7929         The code changes to the files other than dashboard_base.js
7930         are just fallout from renaming globals in dashboard_base.js.
7931
7932         * TestResultServer/static-dashboards/aggregate_results.html:
7933         * TestResultServer/static-dashboards/builders.js:
7934         * TestResultServer/static-dashboards/dashboard_base.js:
7935         * TestResultServer/static-dashboards/flakiness_dashboard.html:
7936         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
7937         * TestResultServer/static-dashboards/timeline_explorer.html:
7938         * TestResultServer/static-dashboards/treemap.html:
7939
7940 2011-07-14  Carlos Garcia Campos  <cgarcia@igalia.com>
7941
7942         Unreviewed. Fix WebKit2 GTK build after r90953.
7943
7944         * WebKitTestRunner/GNUmakefile.am:
7945
7946 2011-07-14  Adam Barth  <abarth@webkit.org>
7947
7948         garden-o-matic results pane should be more discoverable
7949         https://bugs.webkit.org/show_bug.cgi?id=64513
7950
7951         Reviewed by Eric Seidel.
7952
7953         This patch causes us to trigger the results pane on mouse clicks
7954         instead of mousein.  Also, handle the case of no results details more
7955         elegantly.
7956
7957         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
7958         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
7959             - Change the binding to "click".
7960         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
7961             - Handle the "no results URLs" case explicitly instead of never
7962               calling the callback.
7963         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
7964         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
7965             - Show a message when there are no result details to show.
7966         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
7967
7968 2011-07-14  Adam Barth  <abarth@webkit.org>
7969
7970         garden-o-matic should badge TIMEOUT failures
7971         https://bugs.webkit.org/show_bug.cgi?id=64435
7972
7973         Reviewed by Eric Seidel.
7974
7975         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
7976         (.regression .what a[draggable].TIMEOUT::before):
7977             - Add TIMEOUT.
7978
7979 2011-07-14  Adam Barth  <abarth@webkit.org>
7980
7981         Fix Python exception blocking the commit-queue.
7982
7983         * Scripts/webkitpy/common/checkout/scm/git.py:
7984
7985 2011-07-13  Eric Seidel  <eric@webkit.org>
7986
7987         Move webkitpy off of loose mocks
7988         https://bugs.webkit.org/show_bug.cgi?id=64508
7989
7990         Reviewed by Adam Barth.
7991
7992         Using Mock has caused us more pain than help.
7993         It's possible that there was a cleaner way to use it
7994         (maybe Mock(class) instead of inheriting from it?).
7995         But for now, I've removed all uses of Mock from mocktool.py.
7996
7997         I also moved run_command into the only 3 files which call it
7998         instead of leaving the deprecated method in executive.py.
7999
8000         I changed various direct calls to os.* to use filesystem instead.
8001
8002         * Scripts/webkitpy/common/checkout/checkout.py:
8003         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
8004         * Scripts/webkitpy/common/checkout/scm/git.py:
8005         * Scripts/webkitpy/common/checkout/scm/scm.py:
8006         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
8007         * Scripts/webkitpy/common/checkout/scm/svn.py:
8008         * Scripts/webkitpy/common/system/executive.py:
8009         * Scripts/webkitpy/common/system/executive_unittest.py:
8010         * Scripts/webkitpy/tool/mocktool.py:
8011         * Scripts/webkitpy/tool/steps/cleanworkingdirectory_unittest.py:
8012         * Scripts/webkitpy/tool/steps/ensurelocalcommitifneeded.py:
8013
8014 2011-07-13  Eric Seidel  <eric@webkit.org>
8015
8016         NRWT doesn't store the svn revision in full_results.json on chromium-win
8017         https://bugs.webkit.org/show_bug.cgi?id=64492
8018
8019         Reviewed by Adam Barth.
8020
8021         Add a warning message in the revision='' case to help us diagnose
8022         what's going wrong on the bot.
8023
8024         * Scripts/webkitpy/layout_tests/controllers/manager.py:
8025
8026 2011-07-13  Dirk Pranke  <dpranke@chromium.org>
8027
8028         new-run-webkit-tests: print baseline search path as part of config output
8029         https://bugs.webkit.org/show_bug.cgi?id=64499
8030
8031         Reviewed by Eric Seidel.
8032
8033         * Scripts/webkitpy/layout_tests/controllers/manager.py:
8034         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
8035
8036 2011-07-13  Adam Barth  <abarth@webkit.org>
8037
8038         gardening server should have an API for parsing changelogs
8039         https://bugs.webkit.org/show_bug.cgi?id=64495
8040
8041         Reviewed by Eric Seidel.
8042
8043         This patch exposes much of the same information from CommitInfo in a
8044         dictionary form, which is easier to send over-the-wire as JSON to the
8045         frontend.
8046
8047         * Scripts/webkitpy/common/checkout/checkout.py:
8048         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
8049         * Scripts/webkitpy/tool/mocktool.py:
8050         * Scripts/webkitpy/tool/servers/gardeningserver.py:
8051         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
8052         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
8053
8054 2011-07-13  Eric Seidel  <eric@webkit.org>
8055
8056         REGRESSION: GitTestWithMock.test_create_patch fails
8057         https://bugs.webkit.org/show_bug.cgi?id=62945
8058
8059         Reviewed by Daniel Bates.
8060
8061         I was not able to reproduce the exact failure seen in the bug,
8062         however this test was failing on my machine for other reasons.
8063
8064         I went through and did an audit of our run_command usage, it's
8065         entirely in scm classes after this change.  (Not surprising given
8066         that scm.py was the second file ever created in webkit.py.)
8067
8068         The real bug I'm fixing here is that we were setting executive.should_log
8069         when the value had changed to executive._should_log.  Now we set the right one
8070         and the test works again.
8071
8072         * Scripts/webkitpy/common/checkout/checkout.py:
8073         * Scripts/webkitpy/common/checkout/scm/git.py:
8074         * Scripts/webkitpy/common/checkout/scm/scm.py:
8075         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
8076         * Scripts/webkitpy/common/checkout/scm/svn.py:
8077
8078 2011-07-13  Ilya Sherman  <isherman@chromium.org>
8079
8080         Fix WTF header guard style check
8081         https://bugs.webkit.org/show_bug.cgi?id=64488
8082
8083         Reviewed by David Levin.
8084
8085         * Scripts/webkitpy/style/checkers/cpp.py:
8086         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Verify that we don't *always* suggest "WTF_" as a prefix
8087
8088 2011-07-13  Ojan Vafai  <ojan@chromium.org>
8089
8090         bring flakiness_dashboard.html closer to webkit style
8091         https://bugs.webkit.org/show_bug.cgi?id=64477
8092
8093         Reviewed by Adam Barth.
8094
8095         * TestResultServer/static-dashboards/dashboard_base.js:
8096         Fix bad variable name that would hit only when running the tests.
8097         * TestResultServer/static-dashboards/flakiness_dashboard.html:
8098         No code/logic changes. All moving/removing brackets, indents and moving things
8099         to one line.
8100
8101 2011-07-13  Xan Lopez  <xlopez@igalia.com>
8102
8103         [GTK] Fix distcheck
8104
8105         Reviewed by Martin Robinson.
8106
8107         * WebKitTestRunner/GNUmakefile.am: mark built sources as nodist.
8108
8109 2011-07-13  Joseph Pecoraro  <joepeck@webkit.org>
8110
8111         Unreviewed fix. Add a "\n" to a printf that somehow got lost.
8112
8113         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
8114         (WebCore::DumpRenderTree::dumpApplicationCacheQuota):
8115
8116 2011-07-13  Joseph Pecoraro  <joepeck@webkit.org>
8117
8118         ApplicationCache Quota Output is Flakey
8119         https://bugs.webkit.org/show_bug.cgi?id=64410
8120
8121         Reviewed by Alexey Proskuryakov.
8122
8123         Unify the delegate logging for reaching application cache
8124         quotas. Truncate the space needed to the nearest 10000
8125         for less flakey test output. This also better supports
8126         printing NSUInteger on non-64 bit systems by casting to
8127         an unsigned long.
8128
8129         * DumpRenderTree/mac/UIDelegate.mm:
8130         (-[UIDelegate webView:exceededApplicationCacheOriginQuotaForSecurityOrigin:totalSpaceNeeded:]):
8131         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
8132         (WebCore::DumpRenderTree::dumpApplicationCacheQuota):
8133
8134 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
8135
8136         Remove unused TestExpectations._overrides.
8137         https://bugs.webkit.org/show_bug.cgi?id=64470
8138
8139         Reviewed by Adam Barth.
8140
8141         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Removed unused member.
8142
8143 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
8144
8145         Introduce TestExpectationSerializer.list_to_string.
8146         https://bugs.webkit.org/show_bug.cgi?id=64462
8147
8148         Reviewed by Adam Barth.
8149
8150         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Added list_to_string and change the relevant callsite to use it.
8151         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added tests for it.
8152
8153 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
8154
8155         Consolidate expectations parsing code.
8156         https://bugs.webkit.org/show_bug.cgi?id=64460
8157
8158         Reviewed by Adam Barth.
8159
8160         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Folded TestExpectationParser._split_expectation_string into TestExpectationParser.parse.
8161
8162 2011-07-12  Brent Fulgham  <bfulgham@webkit.org>
8163
8164         Standardize WinCairo conditionalized code under PLATFORM macro.
8165         https://bugs.webkit.org/show_bug.cgi?id=64377
8166
8167         Reviewed by Maciej Stachowiak.
8168
8169         Update compile-time conditions that had previously used #ifdef
8170         of WIN_CAIRO to consistently use PLATFORM(WIN_CAIRO).
8171
8172         * DumpRenderTree/config.h: Switch to PLATFORM(WIN_CAIRO)
8173         * DumpRenderTree/win/DumpRenderTreeCairo.vsprops: Remove redundant
8174           define of WIN_CAIRO=1.
8175         * DumpRenderTree/win/DumpRenderTreeDebugCairoCFLite.vsprops: Add
8176           WinCairo.vsprops to set (to match Release build).
8177
8178 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
8179
8180         Eliminate TestExpectationsFile.
8181         https://bugs.webkit.org/show_bug.cgi?id=64458
8182
8183         Turns out, we can just use a Python list.
8184
8185         Reviewed by Adam Barth.
8186
8187         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Folded TestExpectationsFile.append into TestExpectationParser.parse_list,
8188             removed TestExpectationsFile.
8189         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Moved tests to better reflect new names, removed iterator test,
8190             since there's no more custom iterator machinery.
8191
8192 2011-07-13  Philippe Normand  <pnormand@igalia.com>
8193
8194         Unreviewed, added my other email addresses.
8195
8196         * Scripts/webkitpy/common/config/committers.py:
8197
8198 2011-07-13  Adam Barth  <abarth@webkit.org>
8199
8200         Folks should only be listed once.
8201
8202         * Scripts/webkitpy/common/config/committers.py:
8203
8204 2011-07-13  Philippe Normand  <pnormand@igalia.com>
8205
8206         Unreviewed, adding myself as Reviewer.
8207
8208         * Scripts/webkitpy/common/config/committers.py:
8209
8210 2011-07-13  Adam Barth  <abarth@webkit.org>
8211
8212         Fix flickering bug introduced by my previous patch.  I forgot to change
8213         the name of the class everywhere.
8214
8215         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
8216
8217 2011-07-12  Philippe Normand  <pnormand@igalia.com>
8218
8219         [GTK] media/media-blocked-by-willsendrequest.html fails
8220         https://bugs.webkit.org/show_bug.cgi?id=63699
8221
8222         Reviewed by Martin Robinson.
8223
8224         * DumpRenderTree/gtk/DumpRenderTree.cpp:
8225         (willSendRequestCallback): Abort the request if explicitely asked
8226         by the LayoutTestController's willSendRequestReturnsNull() function.
8227
8228 2011-07-13  Adam Barth  <abarth@webkit.org>
8229
8230         Remove "Dismiss" button from garden-o-matic butterbar
8231         https://bugs.webkit.org/show_bug.cgi?id=64443
8232
8233         Reviewed by Dimitri Glazkov.
8234
8235         We don't have any persistent butter bar messages yet, so the dismiss
8236         button is premature (and fairly heavy, visually).
8237
8238         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
8239         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
8240
8241 2011-07-13  Adam Barth  <abarth@webkit.org>
8242
8243         Refine garden-o-matic status pane
8244         https://bugs.webkit.org/show_bug.cgi?id=64442
8245
8246         Reviewed by Dimitri Glazkov.
8247
8248         This patch makes two improvements to the status pane:
8249
8250         1) We only query the server for failure types that we're expecting.
8251            This dramatically reduces the number of HTTP requests, making
8252            loading the status pane faster.
8253
8254         2) The status pane now displays which test and which builder it is
8255            showing results for.
8256
8257         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js:
8258         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
8259         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
8260         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
8261         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
8262         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
8263         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
8264
8265 2011-07-13  Adam Barth  <abarth@webkit.org>
8266
8267         garden-o-matic should show test results
8268         https://bugs.webkit.org/show_bug.cgi?id=64440
8269
8270         Reviewed by Adam Roben.
8271
8272         This patch introduces the results detail pane, which appears at the
8273         bottom of the window and contains test results from the bots.
8274         Currently, you can activate the pane by mousing over one of the builder
8275         names associated with a failing test.
8276
8277         This is just a first iteration of the UI.  There's no way to resize or
8278         zoom in on elements of the details pane, and images likely aren't sized
8279         correctly, but it's a place to start.
8280
8281         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js:
8282             - Add constants for our data attributes so we don't typo them!
8283         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
8284             - Add DOM structure for the details pane.
8285         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
8286         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
8287             - Bind events for showing and hiding the details pane.
8288         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
8289             - Reduce the number of result types that we query for to avoid
8290               overfilling the details pane.  We'll probably need another
8291               solution here in the long-term.  For example, we could use tabs
8292               to pack more results into the pane.
8293         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
8294         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
8295             - Now that we're using the config package during testing, we need
8296               to include it in the testing HTML.
8297         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
8298         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
8299             - Change the DOM structure of the results detail to make it fit
8300               nicely in the details pane (rather than flowing freely in the
8301               body, as it did before).
8302
8303 2011-07-13  Adam Roben  <aroben@apple.com>
8304
8305         Make TestFailures show every time a possibly-flaky test failed, but hide it by default
8306
8307         It's useful to be able to see every time a flaky test failed to see whether it failed the
8308         same way every time. But doing so takes a lot of space, so the list of failures is now
8309         collapsed by default and can be revealed using a disclosure triangle.
8310
8311         Fixes <http://webkit.org/b/64455> TestFailures page doesn't show as much information for
8312         flaky tests as I would like, even though the page is already so long
8313
8314         Reviewed by Dimitri Glazkov.
8315
8316         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
8317         (FlakyLayoutTestDetector.prototype.allFailures): Replaced flakinessExamples with this
8318         function. Now returns all failures for the given test.
8319
8320         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector_unittests.js:
8321         Added. This just contains some simple tests of the FlakyLayoutTestDetector class. We'll add
8322         more over time.
8323
8324         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
8325         (LayoutTestHistoryAnalyzer.prototype.start): Updated the documentation comment to reflect
8326         that we no longer return passing builds for possibly-flaky tests.
8327
8328         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
8329         (.existing-bugs-list, .suspect-revisions-list, .flakiness-examples-list): Make the list of
8330         flakiness examples small, too, since it can get quite long.
8331
8332         (.disclosure-triangle):
8333         (.expanded > .disclosure-triangle):
8334         Simple styles for the disclosure triangle.
8335
8336         (.flakiness-examples-list): Collapse the list by default.
8337
8338         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
8339         (ViewController.prototype._displayBuilder): Pass the total number of builds analyzed to
8340         _domForPossiblyFlakyTests.
8341         (ViewController.prototype._domForPossiblyFlakyTests): Put a disclosure triangle to the left
8342         of each test name, and the number of failures to the right. When the disclosure triangle is
8343         clicked for the first time, we build up the list of failures and expand the element. After
8344         that we just collapse or expand the element on subsequent clicks.
8345
8346         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
8347         Pulled in new tests.
8348
8349 2011-07-13  Adam Roben  <aroben@apple.com>
8350
8351         Teach TestFailures to understand NRWT's output when it exits early due to too many failures
8352
8353         Fixes <http://webkit.org/b/64456> TestFailures page reports way too many failures when NRWT
8354         exits early
8355
8356         Reviewed by Dimitri Glazkov.
8357
8358         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
8359         (Builder.prototype.getNumberOfFailingTests): Relaxed the "Exiting early" test not to require
8360         it to be at the beginning of the line, since NRWT prints a bunch of junk earlier in the
8361         line. Tightened up the regex that's used to parse the number of failing tests to require the
8362         leading number to be followed by whitespace so that we won't parse the "2011" in
8363         "2011-07-13" as a number of failures.
8364
8365         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
8366         Added a test for the above.
8367
8368 2011-07-12  Dimitri Glazkov  <dglazkov@chromium.org>
8369
8370         Extract model-like TestExpectationLine and TestExpectationFile from TestExpectations.
8371         https://bugs.webkit.org/show_bug.cgi?id=64386
8372
8373         This is the first step in converting TestExpectations to a real model.
8374         * TestExpectationsLine represents a line in the test_expectations.txt file, and
8375         * TestExpectationsFile represents the file, which is a collection of lines.
8376
8377         Reviewed by Adam Barth.
8378
8379         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
8380         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
8381         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
8382
8383 2011-07-13  Xan Lopez  <xlopez@igalia.com>
8384
8385         [GTK] Do not grab focus too early in DRT.
8386
8387         Reviewed by Gustavo Noronha.
8388
8389         It causes a layout to happen and a progress signal to be emitted
8390         since r90900, but at this point we don't have a
8391         LayoutTestController object and we'll eventually crash. Since we
8392         already grab focus at the beginning of runTest() this is
8393         redundant, so get rid of it to fix the crash.
8394
8395         * DumpRenderTree/gtk/DumpRenderTree.cpp:
8396         (main): remove call to grab_focus
8397
8398 2011-07-13  Sheriff Bot  <webkit.review.bot@gmail.com>
8399
8400         Unreviewed, rolling out r90893 and r90894.
8401         http://trac.webkit.org/changeset/90893
8402         http://trac.webkit.org/changeset/90894
8403         https://bugs.webkit.org/show_bug.cgi?id=64441
8404
8405         NRWT still doesn't work on qt-mac platform (Requested by Ossy
8406         on #webkit).
8407
8408         * Scripts/run-webkit-tests:
8409         (useNewRunWebKitTests):
8410
8411 2011-07-13  Csaba Osztrogonác  <ossy@webkit.org>
8412
8413         [Qt] NRWT should pick up the right httpd config file
8414         https://bugs.webkit.org/show_bug.cgi?id=64086
8415
8416         * Scripts/run-webkit-tests: Enable NRWT on qt-mac platform after r90810.
8417         (useNewRunWebKitTests):
8418
8419 2011-07-12  Mark Rowe  <mrowe@apple.com>
8420
8421         Fix the 32-bit build.
8422
8423         * DumpRenderTree/mac/UIDelegate.mm:
8424         (-[UIDelegate webView:exceededApplicationCacheOriginQuotaForSecurityOrigin:totalSpaceNeeded:]):
8425         Cast the NSUInteger value to unsigned long to match the format specifier.
8426
8427 2011-07-12  Adam Barth  <abarth@webkit.org>
8428
8429         Improve garden-o-matic UI when the bots fail to report revision numbers
8430         https://bugs.webkit.org/show_bug.cgi?id=64427
8431
8432         Reviewed by Dimitri Glazkov.
8433
8434         At least Win (dbg)(2) seems to fail to report the SVN revision number
8435         in full_results.json.  This patch makes garden-o-matic more robust to
8436         missing revision numbers.
8437
8438         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
8439         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
8440         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
8441         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
8442         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
8443
8444 2011-07-12  Adam Barth  <abarth@webkit.org>
8445
8446         Adjust garden-o-matic layout to use fewer lines
8447         https://bugs.webkit.org/show_bug.cgi?id=64422
8448
8449         Reviewed by Ojan Vafai.
8450
8451         This patch effectively merges the test name line with the list of
8452         builders on which the test fails.  Each test failure now occupies two
8453         lines instead of three.
8454
8455         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
8456         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
8457         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
8458
8459 2011-07-12  Adam Barth  <abarth@webkit.org>
8460
8461         garden-o-matic should label tests failures of new tests as such
8462         https://bugs.webkit.org/show_bug.cgi?id=64421
8463
8464         Reviewed by Ojan Vafai.
8465
8466         Also, we shouldn't dim failures of new tests, even if we've only seen
8467         them once because they're likely to be real problems that need
8468         attention.
8469
8470         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
8471         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js: Added.
8472         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout_unittests.js: Added.
8473         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
8474         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
8475         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
8476         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
8477         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
8478         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
8479         * Scripts/webkitpy/tool/servers/gardeningserver.py:
8480
8481 2011-07-12  Adam Barth  <abarth@webkit.org>
8482
8483         garden-o-matic should display how many times we've seen a failure
8484         https://bugs.webkit.org/show_bug.cgi?id=64417
8485
8486         Reviewed by Ojan Vafai.
8487
8488         This patch adds some UI to display how many times we've seen a given
8489         failure, which can be helpful for determining whether that failure is a
8490         real failure or a flaky test.
8491
8492         When a failure has only been seen once (i.e., only a single run on a
8493         single bot), we set the opacity of to 50% to avoid distracting the
8494         gardener.
8495
8496         This patch also refactors the failure walker to have a simpler API
8497         internally by moving from an object-oriented paradigm to a functional
8498         paradigm.
8499
8500         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
8501         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
8502         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
8503         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
8504         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
8505         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
8506
8507 2011-07-12  Chris Rogers  <crogers@google.com>
8508
8509         Enable Web Audio for chromium DRT
8510         https://bugs.webkit.org/show_bug.cgi?id=64409
8511
8512         Reviewed by James Robinson.
8513
8514         * DumpRenderTree/chromium/TestShell.cpp:
8515         (TestShell::TestShell):
8516
8517 2011-07-12  Adam Barth  <abarth@webkit.org>
8518
8519         Tweak some UI in garden-o-matic now that the tree actually has a
8520         failure and I can see what this all looks like.
8521
8522         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
8523         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
8524         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
8525
8526 2011-07-12  Joseph Pecoraro  <joepeck@webkit.org>
8527
8528         Unreviewed. Skipping a few tests which fail due to differing output
8529         in recent ApplicationCache quota tests. Reenabling the tests is
8530         tracked by <http://webkit.org/b/64410>.
8531
8532         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
8533         (WebCore::DumpRenderTree::dumpApplicationCacheQuota): This should
8534         have been a space, not a newline.
8535
8536 2011-07-12  Adam Barth  <abarth@webkit.org>
8537
8538         garden-o-matic should display regression ranges
8539         https://bugs.webkit.org/show_bug.cgi?id=64407
8540
8541         Reviewed by Dimitri Glazkov.
8542
8543         This patch computes a regression range for a failure by intersecting
8544         the regression ranges seen by the various bots.  We make the underlying
8545         assumption that a test is only failing due to one revision at any given
8546         moment.  If that's not true, this code probably explodes.
8547
8548         The regression ranges appear asynchronously, which might be a jarring
8549         UI.  We'll have to experiment to see.
8550
8551         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
8552         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
8553         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
8554         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
8555         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
8556         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
8557         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
8558
8559 2011-07-12  Adam Barth  <abarth@webkit.org>
8560
8561         cr-linux-ews should run pixel tests
8562         https://bugs.webkit.org/show_bug.cgi?id=64394
8563
8564         Reviewed by Ojan Vafai.
8565
8566         This patch enabled pixel test on the cr-linux-ews.  Running a quick
8567         test on the bots, it looks like we have five pixel failures, related to
8568         some missing fonts.  I've also made the bots skip failing tests to
8569         avoid uploading huge zip files to bugs.webkit.org (with all the exected
8570         IMAGE failures).
8571
8572         * Scripts/webkitpy/common/config/ports.py:
8573         * Scripts/webkitpy/common/config/ports_unittest.py:
8574
8575 2011-07-12  Joseph Pecoraro  <joepeck@webkit.org>
8576
8577         ApplicationCache update should not immediately fail when reaching per-origin quota
8578         https://bugs.webkit.org/show_bug.cgi?id=64177
8579
8580         Reviewed by Alexey Proskuryakov.
8581
8582         Add layoutTestController.disallowIncreaseForApplicationCacheQuota()
8583         to disable the default behavior of raising the default per-origin
8584         quota to 5MB when reached. This allows us to test what happens
8585         when the quota is reached and not increased.
8586
8587         * DumpRenderTree/LayoutTestController.cpp:
8588         (LayoutTestController::LayoutTestController):
8589         (disallowIncreaseForApplicationCacheQuotaCallback):
8590         (LayoutTestController::staticFunctions):
8591         * DumpRenderTree/LayoutTestController.h:
8592         (LayoutTestController::disallowIncreaseForApplicationCacheQuota):
8593         (LayoutTestController::setDisallowIncreaseForApplicationCacheQuota):
8594         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
8595         (LayoutTestController::reset):
8596         * DumpRenderTree/qt/LayoutTestControllerQt.h:
8597         (LayoutTestController::disallowIncreaseForApplicationCacheQuota):
8598         (LayoutTestController::dumpApplicationCacheDelegateCallbacks):
8599         Expose a function to prevent automatically increasing the per-origin
8600         quota when the quota is reached. This allows us to test what happens
8601         when a user would "disallow" a quota increase.
8602
8603         * DumpRenderTree/mac/UIDelegate.mm:
8604         (-[UIDelegate webView:exceededApplicationCacheOriginQuotaForSecurityOrigin:spaceNeeded:]):
8605         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
8606         (WebCore::DumpRenderTree::dumpApplicationCacheQuota):
8607         Respect the new disallowIncrease flag and dump the spaceNeeded values
8608         for ports that implement application cache quotas.
8609
8610 2011-07-12  Ojan Vafai  <ojan@chromium.org>
8611
8612         fix flakiness dashboard to work with new crash log filenames
8613         https://bugs.webkit.org/show_bug.cgi?id=64393
8614
8615         Reviewed by Adam Barth.
8616
8617         * TestResultServer/static-dashboards/flakiness_dashboard.html:
8618
8619 2011-07-12  Adam Barth  <abarth@webkit.org>
8620
8621         Add a happy animation to garden-o-matic when there are no failures
8622         https://bugs.webkit.org/show_bug.cgi?id=64382
8623
8624         Reviewed by Ojan Vafai.
8625
8626         As requested by Ojan.
8627
8628         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
8629         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
8630         * Scripts/webkitpy/tool/servers/data/gardeningserver/partytime.gif: Added.
8631         * Scripts/webkitpy/tool/servers/gardeningserver.py:
8632
8633 2011-07-12  Adam Barth  <abarth@webkit.org>
8634
8635         sheriffbot can't perform rollouts
8636         https://bugs.webkit.org/show_bug.cgi?id=64370
8637
8638         Reviewed by Eric Seidel.
8639
8640         Now that we're not updating the working copy before processing every
8641         IRC message, the rollout command needs to update it manually.
8642
8643         * Scripts/webkitpy/tool/bot/irc_command.py:
8644         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
8645
8646 2011-07-12  Adam Barth  <abarth@webkit.org>
8647
8648         Fix the grammar in the FIXME comment.
8649
8650         * Scripts/webkitpy/common/net/resultsjsonparser.py:
8651
8652 2011-07-12  Eric Seidel  <eric@webkit.org>
8653
8654         NRWT should open test results page with Safari trunk, not the system provided one on Mac
8655         https://bugs.webkit.org/show_bug.cgi?id=64346
8656
8657         Reviewed by Adam Barth.
8658
8659         To fix this I implemented Port.show_results_html_file in Mac, Gtk and Qt ports with
8660         implementations (mostly) matching those found in old-run-webkit-tests.
8661         There are still some minor differences for Qt which Qt hackers may wish to tweak.
8662
8663         I had to add a WebKitPort._port_flag_for_scripts method (similar to flag() in common.config.ports.py)
8664         for the Qt/Gtk ports which always require a flag passed to scripts.
8665
8666         While trying to test this, I found that FactoryTest.test_chromium_gpu_linux
8667         was using a real filesystem (due to assert_platform) which due to
8668         global variables in config.py was causing set-webkit-configuration to have
8669         an affect on unit test results!  So I fixed this by making FactoryTest.assert_port
8670         pass a mock file system whenever calling factory.get.
8671
8672         Unfortunately TestPort was depending on always being passed a None filesystem
8673         and asserting filesystem._tests (only true for unit_test_filesystem()).
8674         So I just removed the FactoryTest.test_test and FactoryTest.test_dryrun tests
8675         deciding that they were pretty much useless anyway.  If others feel strongly
8676         I'm happy to fix this in a different way.
8677
8678         * Scripts/webkitpy/common/system/executive.py:
8679          - default arguments in python are screwy.  They use a single shared
8680            instance, so it's better to use argument=None and then argument = argument or Default()
8681            if you have any chance of mutating (or returning) the default argument.
8682         * Scripts/webkitpy/layout_tests/port/config.py:
8683          - This code is wrong.  We don't need to use a global variable here (as far as I can tell).
8684            I'm not fixing it in this patch, but I've marked it with a FIXME and we can convert to
8685            storing the results of the read on the Config object (which should only be created once during normal operation).
8686            Unit tests shouldn't be hitting the disk anyway.  It's possible Config should move off of Port and onto Tool/Host directly.
8687         * Scripts/webkitpy/layout_tests/port/factory.py:
8688         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
8689         * Scripts/webkitpy/layout_tests/port/gtk.py:
8690         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py: Added.
8691         * Scripts/webkitpy/layout_tests/port/mac.py:
8692         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
8693         * Scripts/webkitpy/layout_tests/port/qt.py:
8694         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
8695         * Scripts/webkitpy/layout_tests/port/webkit.py:
8696
8697 2011-07-12  Adam Roben  <aroben@apple.com>
8698
8699         Teach TestFailures to recognize when run-webkit-tests gets killed by buildbot
8700
8701         Fixes <http://webkit.org/b/64358> TestFailures page thinks all tests passed in
8702         http://build.webkit.org/builders/Windows%207%20Release%20(Tests)/builds/14672
8703
8704         Reviewed by Daniel Bates.
8705
8706         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
8707         (Builder.prototype.getNumberOfFailingTests): If run-webkit-tests exited with a non-zero
8708         exit status but we didn't find any failure counts, assume that there was some error that
8709         caused run-webkit-tests to die early (like being killed by buildbot due to a timeout).
8710
8711         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
8712         Added a new test that shows that we get a failingTestCount of -1 when run-webkit-tests dies
8713         early.
8714         (runGetNumberOfFailingTestsTest): Moved most code here from the only pre-existing test in this
8715         file.
8716
8717         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
8718         (LayoutTestResultsLoader.prototype.start): Bump the cache number so old cached data that was
8719         tainted by the bug fixed in this patch will be evicted.
8720
8721 2011-07-12  Adam Barth  <abarth@webkit.org>
8722
8723         cr-linux-ews complains about tests that aren't actually failing
8724         https://bugs.webkit.org/show_bug.cgi?id=64367
8725
8726         The underlying problem here is that full_results.json doesn't have
8727         enough information to determine whether a given test result was
8728         expected because whether an actual result is expected depends on
8729         whether full_results.json was generated duing a test run that included
8730         pixel tests.
8731
8732         The right long-term solution is to make full_results.json a complete
8733         description of what happened durning a test run.  In this patch
8734         however, to stop the spam, this patch makes the jsonresultsparser
8735         ignore pixel failures.
8736
8737         (I'm landing this patch unreviewed in order to stop the bots from
8738         spamming.  I'll happily address any review comments in a follow-up
8739         patch.)
8740
8741         * Scripts/webkitpy/common/net/resultsjsonparser.py:
8742         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
8743
8744 2011-07-12  Adam Barth  <abarth@webkit.org>
8745
8746         Re-work garden-o-matic UI to begin to look like the mocks
8747         https://bugs.webkit.org/show_bug.cgi?id=64334
8748
8749         Reviewed by Dimitri Glazkov.
8750
8751         This patch changes how we display failures to be more like the mock the
8752         dglazkov made. We now can display the same information much more
8753         compactly.
8754
8755         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
8756         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
8757         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
8758         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css: Added.
8759         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
8760         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
8761         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
8762         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
8763         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
8764         * Scripts/webkitpy/tool/servers/gardeningserver.py:
8765
8766 2011-07-11  Adam Roben  <aroben@apple.com>
8767
8768         Extract TestFailures's bug-filing code into two new classes
8769
8770         Fixes <http://webkit.org/b/64300> TestFailures page's new-bug-filing code is a mess!
8771
8772         Reviewed by Darin Adler and Adam Barth.
8773
8774         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js:
8775         (Buildbot.prototype.resultsDirectoryURL): Changed to return a URI-encoded URL. Otherwise the
8776         URL isn't valid (and it's harder to mock this function).
8777
8778         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot_unittests.js:
8779         Added. Just tests the above change (for now).
8780
8781         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
8782         Added a license header, enclosed everything in a closure to avoid polluting the global
8783         namespace, changed the test name to actually describe the passing condition, and replaced
8784         uses of equals() with equal(). (The latter is the actual name of the function, and matches
8785         deepEqual, etc.)
8786
8787         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm.js: Added.
8788         (NewBugForm): This class knows how to construct a <form> used for filing new bugs in
8789         Bugzilla based on some parameters.
8790         (NewBugForm.prototype.domElement): Creates and returns the <form> element. Code came from
8791         ViewController.prototype._domForNewAndExistingBugs.
8792
8793         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm_unittests.js:
8794         Added. Tests the above code.
8795
8796         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm.js: Added.
8797         (TestFailureBugForm): This class knows how to construct a <form> element used for filing new
8798         bugs specifically about test failures. Code came from
8799         ViewController.prototype._domForNewAndExistingBugs.
8800         (TestFailureBugForm.prototype.domElement): Slightly customizes the <form> element returned
8801         by NewBugForm.
8802         (TestFailureBugForm.prototype._computeOperatingSystem):
8803         (TestFailureBugForm.prototype._computePlatform):
8804         (TestFailureBugForm.prototype._createBugTitle):
8805         (TestFailureBugForm.prototype._failingResultsHTMLURL):
8806         (TestFailureBugForm.prototype._failingRevision):
8807         (TestFailureBugForm.prototype._passingRevision):
8808         (TestFailureBugForm.prototype._regressionRangeString):
8809         Code came from ViewController.prototype._domForNewAndExistingBugs. I broke it out into
8810         separate functions to break up the rat's nest a bit.
8811
8812         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm_unittests.js:
8813         Added. Tests the above code.
8814
8815         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
8816         (ViewController.prototype._domForNewAndExistingBugs): Moved code from here to
8817         TestFailureBugForm/NewBugForm, and changed this code to use a TestFailureBugForm.
8818
8819         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
8820         NewBugForm/TestFailureBugForm.js.
8821
8822         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
8823         Added new tests and required files.
8824
8825 2011-07-12  Adam Roben  <aroben@apple.com>
8826
8827         Test that no intermediate WTF::Strings are created when concatenating with string literals
8828
8829         Test for <http://webkit.org/b/63330> Concatenating string literals and WTF::Strings using
8830         operator+ is suboptimal
8831
8832         Reviewed by Darin Adler.
8833
8834         * TestWebKitAPI/Tests/WTF/StringOperators.cpp: Added.
8835         (TestWebKitAPI::TEST): Test that a bunch of different string concatenation expressions don't
8836         create any intermediate WTF::Strings while they're being evaluated.
8837
8838         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
8839         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
8840         Added new file.
8841
8842 2011-07-12  Eric Seidel  <eric@webkit.org>
8843
8844         [Qt] NRWT should pick up the right httpd config file
8845         https://bugs.webkit.org/show_bug.cgi?id=64086
8846
8847         Reviewed by Adam Barth.
8848
8849         This is more fall-out from the Port class inappropriately encapsulating
8850         both platform and port knowledge.  We need to split out some of this
8851         platform knowledge into a separate class which can be better shared
8852         between ports.
8853
8854         The fix was to move all the _path_to_apache_config_file logic down
8855         into the WebKitPort baseclass so that all of the ports can find the
8856         right apache config file, regardless of what platform they may be running on.
8857
8858         I did not move this down into base.Port (even though I considered it).
8859         Chromium duplicates some of this logic, but since they have separate
8860         subclasses for each port-OS combination (e.g. ChromiumMac, ChromiumWin)
8861         they wouldn't notice the change.  Eventually we'll move this logic
8862         out of Port entirely, and then it will be shared by all ports.
8863
8864         I also cleaned up the http-config logic for ORWT while I was there,
8865         although since we're killing that code, I'm happy to revert that part
8866         of the change if changing it is viewed as needlessly risky.
8867
8868         During this cleanup, I noticed that no ports use the "default" httpd.conf
8869         which is left over from Mac OS X Tiger (and old linux distros) which used
8870         Apache 1.3.  I've removed httpd.conf (and associated support in ORWT) since
8871         we no longer support any ports which use this httpd.conf.
8872
8873         * Scripts/webkitperl/httpd.pm:
8874         (getHTTPDConfigPathForTestDirectory):
8875         * Scripts/webkitpy/layout_tests/port/base.py:
8876         * Scripts/webkitpy/layout_tests/port/gtk.py:
8877         * Scripts/webkitpy/layout_tests/port/mac.py:
8878         * Scripts/webkitpy/layout_tests/port/qt.py:
8879         * Scripts/webkitpy/layout_tests/port/webkit.py:
8880         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
8881         * Scripts/webkitpy/layout_tests/port/win.py:
8882         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
8883
8884 2011-07-11  Hironori Bono  <hbono@chromium.org>
8885
8886         Reviewed by Adam Roben.
8887
8888         Implement layoutTestController.setTextDirection for WebKit2 and Windows.
8889         https://bugs.webkit.org/show_bug.cgi?id=61931
8890
8891         This change implements layoutTextController.setTextDirection for WebKit2
8892         and Windows so we can run a layout test added by r87770 on them. Each
8893         implementation adds a wrapper function for Editor::setBaseWritingDirection()
8894         so LayoutTestController can call it.
8895
8896         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
8897         (LayoutTestController::setTextDirection): Implemented the binding function
8898         for layoutTestController.setTextDirection (Windows).
8899         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
8900         Added a binding function for layoutTestController.setTextDirection (WebKit2).
8901         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp: ditto.
8902         (WTR::LayoutTestController::setTextDirection):
8903         * WebKitTestRunner/InjectedBundle/LayoutTestController.h: ditto.
8904
8905 2011-07-11  Raphael Kubo da Costa  <kubo@profusion.mobi>
8906
8907         [EFL] ImageDiff: Make sure gEcoreEvas is destroyed at the right time.
8908         https://bugs.webkit.org/show_bug.cgi?id=64293
8909
8910         Reviewed by Kent Tamura.
8911
8912         gEcoreEvas was being destroyed after shutdownEfl() was called, so the
8913         call to ecore_evas_free() failed.
8914
8915         We now shut it down manually before shutdownEfl() is called.
8916
8917         * DumpRenderTree/efl/ImageDiff.cpp:
8918         (main): Destroy gEcoreEvas before shutting down the EFL.
8919
8920 2011-07-11  Ryosuke Niwa  <rniwa@webkit.org>
8921
8922         Make Chromium Mac and Linux tests bot core builders
8923         https://bugs.webkit.org/show_bug.cgi?id=63196
8924
8925         Reviewed by Adam Barth.
8926
8927         Added Chromium Mac and Linux Release Tests bots core builders.
8928
8929         Also rearranged the order of bots so that test bots show up right next to
8930         respective builders.
8931
8932         * BuildSlaveSupport/build.webkit.org-config/config.json:
8933         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
8934         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
8935
8936 2011-07-11  Raphael Kubo da Costa  <kubo@profusion.mobi>
8937
8938         [EFL] DRT: Add code related to font management.
8939         https://bugs.webkit.org/show_bug.cgi?id=63989
8940
8941         Reviewed by Kent Tamura.
8942
8943         The added files are responsible for managing fontconfig and adding the
8944         fonts required by some layout tests to fontconfig's path.
8945
8946         * DumpRenderTree/efl/FontManagement.cpp: Added.
8947         (getFontDirectories):
8948         (getFontFiles):
8949         (addFontDirectories):
8950         (addFontFiles):
8951         (addFontsToEnvironment):
8952         * DumpRenderTree/efl/FontManagement.h: Added.
8953
8954 2011-07-08  Dirk Pranke  <dpranke@chromium.org>
8955
8956         nrwt: linting fixes
8957         https://bugs.webkit.org/show_bug.cgi?id=64225
8958
8959         Reviewed by Eric Siedel.
8960
8961         Miscellaneous linting fixes. The most notable change is that
8962         we add public attributes for user, executive, filesystem, and
8963         options on the Port object, so we don't have to refer to the
8964         "protected" versions all over the place".
8965
8966         * Scripts/webkitpy/layout_tests/controllers/manager.py:
8967         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
8968         * Scripts/webkitpy/layout_tests/controllers/message_broker.py:
8969         * Scripts/webkitpy/layout_tests/controllers/message_broker_unittest.py:
8970         * Scripts/webkitpy/layout_tests/controllers/worker.py:
8971         * Scripts/webkitpy/layout_tests/port/base.py:
8972         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
8973
8974 2011-07-11  Ryosuke Niwa  <rniwa@webkit.org>
8975
8976         webkit-patch roll-chromium-deps no longer works
8977         https://bugs.webkit.org/show_bug.cgi?id=64324
8978
8979         Reviewed by James Robinson.
8980
8981         Lower the minimum commit message length to 10.
8982
8983         * Scripts/webkitpy/tool/steps/commit.py:
8984
8985 2011-07-11  Alice Boxhall  <aboxhall@chromium.org>
8986
8987         Convert json_results_generator.py to output version 4 JSON.
8988         https://bugs.webkit.org/show_bug.cgi?id=60869
8989
8990         Reviewed by Ojan Vafai.
8991
8992         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
8993         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
8994         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
8995
8996 2011-07-11  Martin Robinson  <mrobinson@igalia.com>
8997
8998         [GTK] [NRWT] Pixel tests do not work
8999         https://bugs.webkit.org/show_bug.cgi?id=64091
9000
9001         Reviewed by Eric Seidel.
9002
9003         Fix pixel test runs for GTK+ new-run-webkit-tests by exposing the
9004         path to the ImageDiff binary.
9005
9006         * Scripts/webkitpy/layout_tests/port/gtk.py: Implement _path_to_image_diff.
9007
9008 2011-07-11  Dirk Pranke  <dpranke@chromium.org>
9009
9010         Fix hang on win32 in a multiprocessing test that shouldn't have been running.
9011
9012         Unreviewed, build fix.
9013
9014         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
9015
9016 2011-07-11  Adam Roben  <aroben@apple.com>
9017
9018         Don't count leaks as test failures on TestFailures's front page
9019
9020         As a bonus, this patch adds our first unit test.
9021
9022         Fixes <http://webkit.org/b/64303> REGRESSION (r90489): TestFailures page says far more tests
9023         are failing on the Leaks bot than actually are
9024
9025         Reviewed by Adam Barth.
9026
9027         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
9028         (Builder.prototype.getNumberOfFailingTests): Exclude lines containing "leak" when summing
9029         failure counts.
9030
9031         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
9032         Added. Contains a simple test to show that leaks aren't counted as test failures.
9033
9034         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
9035         Added. This is the test harness.
9036
9037 2011-06-14  Mike Stegeman  <mrstegeman@gmail.com>
9038
9039         [GTK] Add API to allow setting local storage database path
9040         https://bugs.webkit.org/show_bug.cgi?id=62091
9041
9042         Reviewed by Martin Robinson and Gustavo Noronha.
9043
9044         Expose the path of the localStorage databases through a setting
9045         to allow HTML5 localStorage to be persistent. New setting is:
9046         html5-local-storage-database-path
9047
9048         * DumpRenderTree/gtk/DumpRenderTree.cpp:
9049         (resetDefaultsToConsistentValues):
9050
9051 2011-07-11  Adam Barth  <abarth@webkit.org>
9052
9053         garden-o-matic should be able to rebaseline tests
9054         https://bugs.webkit.org/show_bug.cgi?id=64186
9055
9056         Reviewed by Eric Seidel.
9057
9058         This patch contains a simple HTTP server binding for the
9059         rebaseline-test command.  The frontend will appear in another patch.
9060
9061         * Scripts/webkitpy/tool/servers/gardeningserver.py:
9062         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
9063         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
9064
9065 2011-07-11  Adam Barth  <abarth@webkit.org>
9066
9067         Add some basic UI for showing regression ranges
9068         https://bugs.webkit.org/show_bug.cgi?id=64243
9069
9070         Reviewed by Eric Seidel.
9071
9072         This UI is all temporary.  It exists mostly just to have some buttons
9073         to click to exercise the code.
9074
9075         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
9076         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
9077         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
9078         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
9079         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
9080
9081 2011-07-11  Adam Barth  <abarth@webkit.org>
9082
9083         garden-o-matic should be able to determine which revisions caused a given failure
9084         https://bugs.webkit.org/show_bug.cgi?id=64189
9085
9086         Reviewed by Adam Roben.
9087
9088         Walking the failure history looking for failures turns out to be
9089         slightly tricky because the network requests are asynchronous.
9090         Currently we do all the fetches serially and our cache is unbounded.
9091         We'll probably optimize both those parameters eventually.
9092
9093         This patch also generalizes some functionality in the unit testing
9094         framework to make testing this sort of code easier.
9095
9096         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
9097         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
9098         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
9099
9100 2011-07-11  Adam Barth  <abarth@webkit.org>
9101
9102         Add a webkit-patch command for rebaselining an individual test
9103         https://bugs.webkit.org/show_bug.cgi?id=64246
9104
9105         Reviewed by Eric Seidel.
9106
9107         This patch introduces a command that's able to rebaseline a single
9108         test.  Currently, the command works only with the build.chromium.org
9109         buildbots, but extending it to work with the build.webkit.org bots
9110         shouldn't be that hard.
9111
9112         A complete rebaseling tool should also include an "optimize baselines"
9113         command (which moves/deletes baselines in order to reduce the number of
9114         expected results files), but that will come in a future patch.
9115
9116         Really BuilderToPort should be merged into builders.py, but I'm going
9117         to save that for a future patch as well.  (We need to stop shaving yaks
9118         at some point.)
9119
9120         * Scripts/webkitpy/tool/commands/rebaseline.py:
9121
9122 2011-07-11  Adam Barth  <abarth@webkit.org>
9123
9124         garden-o-matic should be able to roll out patches
9125         https://bugs.webkit.org/show_bug.cgi?id=64185
9126
9127         Reviewed by Eric Seidel.
9128
9129         This gardening server API is simply a binding to the rollout machinery
9130         already present in webkit-patch.
9131
9132         * Scripts/webkitpy/tool/commands/gardenomatic.py:
9133         * Scripts/webkitpy/tool/servers/gardeningserver.py:
9134         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Added.
9135
9136 2011-07-11  Csaba Osztrogonác  <ossy@webkit.org>
9137
9138         [Qt][Mac] Unreviewed fix after r90746.
9139
9140         * Scripts/run-webkit-tests:
9141         (useNewRunWebKitTests): Really disable NRWT for qt-mac platform.
9142
9143 2011-07-11  Adam Roben  <aroben@apple.com>
9144
9145         Teach TestFailures that ORWT's results.html file might be missing due to all tests passing
9146
9147         There are three reasons why we might fail to fetch ORWT's results.html:
9148           - All tests passed, so no results.html was generated
9149           - Some error during the test run caused results.html not to be generated (e.g., ORWT
9150             timed out)
9151           - Some network error occurred when fetching results.html
9152
9153         We were failing to account for the first possibility in some cases. For test runs before
9154         r89610, we first check build.webkit.org/json to determine how many tests failed and whether
9155         ORWT exited early due to too many failures; if all tests passed then we don't fetch
9156         results.html at all. r89610 changed ORWT to put information in results.html about exiting
9157         early due to too many failures, so we no longer needed to check build.webkit.org/json to get
9158         that information, and in r89619 I changed TestFailures to do just that. But I forgot that we
9159         still needed to check build.webkit.org/json to find out if all tests passed!
9160
9161         Now, for test runs after r89610, we check results.html first, and then check
9162         build.webkit.org/json if we fail to fetch results.html. This lets us distinguish between all
9163         tests passing and the error cases.
9164
9165         Fixes <http://webkit.org/b/64280> TestFailures page can't pinpoint that r90699 caused 13
9166         tests to fail on Windows 7 Release (WebKit2 Tests)
9167
9168         Reviewed by David Kilzer.
9169
9170         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
9171         (LayoutTestResultsLoader.prototype.start): Bumped the cache version so that old, buggy
9172         cached data will get evicted. We were marking builds where all tests passed as errors!
9173         (LayoutTestResultsLoader.prototype._fetchAndParseORWTResults): Added success/error callback
9174         parameters to the fetchAndParseResultsHTML helper function, and added a similar
9175         fetchNumberOfFailingTests function that fetches data from build.webkit.org/json (code came
9176         from later in the function). For test runs before r89610, we first check
9177         build.webkit.org/json then check results.html. For builds after r89610, we first check
9178         results.html and then check build.webkit.org/json if we couldn't fetch results.html.
9179
9180 2011-07-11  Csaba Osztrogonác  <ossy@webkit.org>
9181
9182         new-run-webkit-tests does not support qt-arm or qt-4.8 results
9183         https://bugs.webkit.org/show_bug.cgi?id=64071
9184
9185         [Qt] NRWT should pick up the right httpd config file
9186         https://bugs.webkit.org/show_bug.cgi?id=64086
9187
9188         Reviewed by Zoltán Herczeg.
9189
9190         * Scripts/run-webkit-tests: Disable NRWT for qt-mac, qt-arm and qt-4.8 platforms until fix.
9191         (useNewRunWebKitTests):
9192
9193 2011-07-11  Kenichi Ishibashi  <bashi@chromium.org>
9194
9195         Add TestNetscapePlugIn/Tests/FormValue.cpp to DRT build files
9196         https://bugs.webkit.org/show_bug.cgi?id=64248
9197
9198         Reviewed by Kent Tamura.
9199
9200         * DumpRenderTree/DumpRenderTree.gypi: Added FormValue.cpp.
9201         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Ditto.
9202         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro: Ditto.
9203         * GNUmakefile.am: Ditto.
9204
9205 2011-07-10  Kenichi Ishibashi  <bashi@chromium.org>
9206
9207         Let plugins participate in form submission
9208         https://bugs.webkit.org/show_bug.cgi?id=13061
9209
9210         Adds NPPVformValue support to the testing plugin.
9211
9212         Reviewed by Anders Carlsson.
9213
9214         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added FormValue.cpp
9215         * DumpRenderTree/TestNetScapePlugIn/Tests/FormValue.cpp: Added.
9216         (FormValue::FormValue): Ditto.
9217         (FormValue::NPP_GetValue): Ditto.
9218
9219 2011-07-11  Roland Steiner  <rolandsteiner@chromium.org>
9220
9221         Reviewed by Tony Chang.
9222
9223         Complete functions in filesystem.py
9224         https://bugs.webkit.org/show_bug.cgi?id=63528
9225
9226         - missing text file functions added
9227         - functions sorted
9228         - removed 'append' optional parameters (were unused)
9229         - adapted filesystem_mock in the same way
9230
9231         * Scripts/webkitpy/common/system/filesystem.py:
9232         * Scripts/webkitpy/common/system/filesystem_mock.py:
9233         * Scripts/webkitpy/common/system/filesystem_unittest.py:
9234
9235 2011-07-10  Adam Barth  <abarth@webkit.org>
9236
9237         Clean up style in fallback path calculation
9238         https://bugs.webkit.org/show_bug.cgi?id=64239
9239
9240         Reviewed by Daniel Bates.
9241
9242         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
9243         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
9244         * Scripts/webkitpy/layout_tests/port/mac.py:
9245         * Scripts/webkitpy/layout_tests/port/win.py:
9246
9247 2011-07-10  Balazs Kelemen  <kbalazs@webkit.org>
9248
9249         [Qt][WK2] Implement the rest of WTR::PlatformWebView
9250         https://bugs.webkit.org/show_bug.cgi?id=63630
9251
9252         Reviewed by Kenneth Rohde Christiansen.
9253
9254         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
9255         (WTR::PlatformWebView::windowFrame):
9256         (WTR::PlatformWebView::setWindowFrame):
9257
9258 2011-07-10  Mark Rowe  <mrowe@apple.com>
9259
9260         Teach build-webkit how to find the latest WebKitSystemInterface binary.
9261
9262         * Scripts/build-webkit:
9263
9264 2011-07-09  Adam Roben  <aroben@apple.com>
9265
9266         Teach TestFailures to abbreviate the examples of test flakiness
9267
9268         These lists can get quite long, and it's not really helpful in most cases to have soooooo
9269         many examples of flakiness.
9270
9271         Fixes <http://webkit.org/b/64203> Lists of flaky revisions on TestFailures page can get so
9272         long they're hard to navigate
9273
9274         Reviewed by Dan Bates.
9275
9276         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
9277         (FlakyLayoutTestDetector.prototype.flakinessExamples): If we have more than a certain number
9278         of examples, replace the middle items with a separator.
9279
9280         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
9281         (.flakiness-example-separator): Added styles for the separator.
9282
9283         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
9284         (ViewController.prototype._domForPossiblyFlakyTests): Use a vertical ellipsis to represent
9285         the separator.
9286
9287 2011-07-09  Dirk Pranke  <dpranke@chromium.org>
9288
9289         nrwt: stack traces from worker-side exceptions aren't very useful inside test-webkitpy
9290         https://bugs.webkit.org/show_bug.cgi?id=64218
9291
9292         Reviewed by Eric Seidel.
9293
9294         Exceptions aren't picklable and can't be sent across the
9295         manager/worker message queue without losing information. NRWT
9296         handles this by turning the stack trace into a set of strings,
9297         and logging the strings when we receive an exception from the
9298         worker. However, when you are running tests and something
9299         crashes on the worker side, test-webkitpy prints the
9300         manager-side stack trace, which is just confusing and useless.
9301
9302         This patch changes the logic so that exceptions are passed
9303         through as-is when the worker and manager are in the same
9304         process (the --worker-model=inline option). This increases the
9305         code paths slightly but makes crashes much more useful.
9306
9307         * Scripts/webkitpy/layout_tests/controllers/manager.py:
9308         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
9309         * Scripts/webkitpy/layout_tests/controllers/message_broker.py:
9310         * Scripts/webkitpy/layout_tests/controllers/worker.py:
9311         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
9312
9313 2011-07-09  Darin Fisher  <darin@chromium.org>
9314
9315         Eliminate bad dependency on gfx::Point.
9316         https://bugs.webkit.org/show_bug.cgi?id=64228
9317
9318         Reviewed by Kent Tamura.
9319
9320         * DumpRenderTree/chromium/EventSender.cpp:
9321         (initMouseEvent):
9322
9323 2011-07-08  Ojan Vafai  <ojan@chromium.org>
9324
9325         pull static dashboard files into the appengine server from the chromium repository
9326         https://bugs.webkit.org/show_bug.cgi?id=64208
9327
9328         Reviewed by Dirk Pranke.
9329
9330         These files belong in the WebKit repo since they are tied primarily to the webkit tests.
9331         They have some extra bits to support chromium's gtests, but that seems fine.
9332         Mainly, this will allow other WebKit hackers to hack on the dashboards.
9333
9334         As a nice side-effect, we can now push the dashboard files when we do appengine pushes
9335         instead of the weird thing we used to do of pulling them from the Chromium repository
9336         and storing them in the appengine datastore. This allows for cleaning up a lot of code
9337         and will likely make the dashboards load a bit faster.
9338
9339         The new JS files don't fully match WebKit style, but I'd like to clean that up in a
9340         followup patch if possible to maintain my sanity with this patch.
9341
9342         * TestResultServer/app.yaml:
9343         * TestResultServer/handlers/dashboardhandler.py: Removed.
9344         * TestResultServer/handlers/menu.py:
9345         * TestResultServer/main.py:
9346         * TestResultServer/model/dashboardfile.py: Removed.
9347         * TestResultServer/static-dashboards/LICENSE.dygraph.txt: Added.
9348         * TestResultServer/static-dashboards/README.dygraph.txt: Added.
9349         * TestResultServer/static-dashboards/README.webtreemap.txt: Added.
9350         * TestResultServer/static-dashboards/aggregate_results.html: Added.
9351         * TestResultServer/static-dashboards/builders.js: Added.
9352         * TestResultServer/static-dashboards/dashboard_base.js: Added.
9353         * TestResultServer/static-dashboards/dygraph-combined.js: Added.
9354         * TestResultServer/static-dashboards/flakiness_dashboard.html: Added.
9355         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js: Added.
9356         * TestResultServer/static-dashboards/timeline_explorer.html: Added.
9357         * TestResultServer/static-dashboards/treemap.html: Added.
9358         * TestResultServer/static-dashboards/webtreemap.css: Added.
9359         * TestResultServer/static-dashboards/webtreemap.js: Added.
9360         * TestResultServer/stylesheets/dashboardfile.css: Removed.
9361         * TestResultServer/stylesheets/menu.css:
9362         * TestResultServer/templates/dashboardfilelist.html: Removed.
9363         * TestResultServer/templates/menu.html:
9364
9365 2011-07-08  Adam Roben  <aroben@apple.com>
9366
9367         Make TestFailures's list of flaky tests look more like the list of non-flaky tests
9368
9369         Fixes <http://webkit.org/b/64204> TestFailures page's flaky tests list is ugly!
9370
9371         Reviewed by Daniel Bates.
9372
9373         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
9374         (#failure-history, #possibly-flaky-tests): Expanded this rule to apply to the list of flaky
9375         tests.
9376         (#failure-history > li, #possibly-flaky-tests > li): Ditto, but moved the 50px left padding
9377         from here...
9378         (#failure-history > li): ...to here.
9379
9380         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
9381         (ViewController.prototype._domForPossiblyFlakyTests): Give the list an id attribute for
9382         styling purposes.
9383
9384 2011-07-08  Jeffrey Pfau  <jpfau@apple.com>
9385
9386         Unreviewed, add myself as committer.
9387
9388         * Scripts/webkitpy/common/config/committers.py:
9389
9390 2011-07-08  Adam Barth  <abarth@webkit.org>
9391
9392         TestResultsServer should keep old test results
9393         https://bugs.webkit.org/show_bug.cgi?id=64199
9394
9395         Reviewed by Ojan Vafai.
9396
9397         Having historical data will help us do failure archeology.
9398
9399         * TestResultServer/handlers/testfilehandler.py:
9400         * TestResultServer/model/testfile.py:
9401         * TestResultServer/templates/showfilelist.jsonp: Added.
9402
9403 2011-07-08  Adam Roben  <aroben@apple.com>
9404
9405         Remove commit-log-editor's dependency on Module::Load::Conditional
9406
9407         This module isn't available in Perl 5.8.8 (the version used on Leopard).
9408
9409         Fixes <http://webkit.org/b/64198> REGRESSION (r90583):
9410         webkitpy.common.checkout.checkout_unittest failing on Leopard
9411
9412         Reviewed by Daniel Bates.
9413
9414         * Scripts/commit-log-editor: Use the new loadTermReadKey() function instead of
9415         Module::Load::Conditional::can_load.
9416         (loadTermReadKey): Added. Tries to load Term::ReadKey and returns true if it's successful.
9417
9418 2011-07-08  Adam Roben  <aroben@apple.com>
9419
9420         Make checkout_unittest more robust against files moving around
9421
9422         Fixes <http://webkit.org/b/64197> checkout_unittest contains ugly, fragile code to find the
9423         Scripts directory
9424
9425         Reviewed by Adam Barth.
9426
9427         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
9428         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Instantiate a real SCM
9429         object and use it to get the path to the Scripts directory, rather than hard-coding the
9430         relative path from this file to Scripts.
9431
9432 2011-07-08  Adam Roben  <aroben@apple.com>
9433
9434         Make TestFailures remember that run-webkit-tests timed out, even across reloads
9435
9436         Fixes <http://webkit.org/b/64193> TestFailures page incorrectly thinks all tests passed in
9437         http://build.webkit.org/builders/Windows%207%20Release%20(Tests)/builds/14589 after a reload
9438
9439         Reviewed by Daniel Bates.
9440
9441         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
9442         (LayoutTestResultsLoader.prototype.start): Store an "error" attribute in the cached data.
9443         When true, it indicates that there was an error retrieving the results for this build and
9444         that the errorCallback should be called.
9445
9446 2011-07-08  Adam Barth  <abarth@webkit.org>
9447
9448         Teach garden-o-matic how to display test results
9449         https://bugs.webkit.org/show_bug.cgi?id=64141
9450
9451         Reviewed by Ojan Vafai.
9452
9453         This patch includes basic infrastructure for probing build.chromium.org
9454         for test results.  We only handle text and image tests, not anything
9455         complicated like reftests.  Also, we're using the revision/build
9456         independent results store on the server, so we're avoiding that
9457         complication for now.
9458
9459         It's slightly hacky that we need to probe the server to see what kinds
9460         of results exist.  A better solution would be to add CORS support to
9461         the server or to use the local server to help.
9462
9463         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
9464         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
9465         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
9466         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
9467         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
9468         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
9469
9470 2011-07-08  Dirk Pranke  <dpranke@chromium.org>
9471
9472         REGRESSION(90419) NRWT's httpd locking is broken for --child-processes=1
9473         https://bugs.webkit.org/show_bug.cgi?id=64092
9474
9475         Reviewed by Tony Chang.
9476
9477         The code was incorrectly creating empty shards in the
9478         shard_in_two case.
9479
9480         * Scripts/webkitpy/layout_tests/controllers/manager.py:
9481         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
9482
9483 2011-07-08  Adam Roben  <aroben@apple.com>
9484
9485         Teach buildbot to figure out how many webkitpy/webkitperl tests failed
9486
9487         Fixes <http://webkit.org/b/64192> It's hard to tell how many test-webkitpy/test-webkitperl
9488         tests failed when looking at build.webkit.org
9489
9490         Reviewed by Eric Seidel.
9491
9492         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
9493         (TestWithFailureCount): New class that represents a test build step which has an associated
9494         failure count. Eventually we should move more of our test classes to deriving from this.
9495         (TestWithFailureCount.countFailures): Method for subclasses to override to say how many
9496         failures occurred.
9497
9498         (TestWithFailureCount.commandComplete):
9499         (TestWithFailureCount.evaluateCommand):
9500         (TestWithFailureCount.getText):
9501         (TestWithFailureCount.getText2):
9502         These were all based on RunGtkAPITests.
9503
9504         (RunPythonTests): Changed to inherit from TestWithFailureCount.
9505         (RunPythonTests.countFailures): Parses the test-webkitpy output looking for the count of
9506         failures.
9507         (RunPerlTests): Changed to inherit from TestWithFailureCount.
9508         (RunPerlTests.countFailures): Parses the test-webkitperl output looking for the count of
9509         failures.
9510
9511 2011-07-08  Adam Roben  <aroben@apple.com>
9512
9513         Ensure $CHANGE_LOG_EMAIL_ADDRESS is set when testing webkitpy's commit-log-editor integration
9514
9515         Fixes <http://webkit.org/b/64180> REGRESSION (r90564): test-webkitpy failing on multiple
9516         bots due to commit-log-editor errors
9517
9518         Reviewed by Adam Barth.
9519
9520         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
9521         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Set
9522         $CHANGE_LOG_EMAIL_ADDRESS to the patch author's email address. This ensures that
9523         commit-log-editor can find a value for the committer's email, and that the committer and
9524         author email addresses match, which will prevent commit-log-editor from inserting a "Patch
9525         by" line in the commit message.
9526
9527 2011-07-08  Andreas Kling  <kling@webkit.org>
9528
9529         [Qt][WK2] Views should expose QActions for basic navigation.
9530         https://bugs.webkit.org/show_bug.cgi?id=64174
9531
9532         Reviewed by Benjamin Poulain.
9533
9534         Bring the toolbar in Qt's MiniBrowser back to life.
9535
9536         * MiniBrowser/qt/BrowserView.cpp:
9537         (BrowserView::navigationAction):
9538         * MiniBrowser/qt/BrowserView.h:
9539         * MiniBrowser/qt/BrowserWindow.cpp:
9540         (BrowserWindow::BrowserWindow):
9541
9542 2011-07-08  Adam Roben  <aroben@apple.com>
9543
9544         Teach TestFailures how to find test names in commit-log-editor-style commit messages
9545
9546         TestFailures was relying on Trac turning the list of modified files in our commit messages
9547         into an HTML list. But Trac only does this when the list of modified files is indented.
9548         commit-log-editor doesn't indent the file list, so the list wasn't being turned into an HTML
9549         list, which was confusing TestFailures.
9550
9551         TestFailures now does much simpler parsing of the commit message (i.e., just a substring
9552         search) without relying at all on its structure.
9553
9554         Fixes <http://webkit.org/b/64173> TestFailures page fails to blame r90608 for breaking
9555         fast/dom/HTMLProgressElement/progress-element-markup.html on Windows
9556
9557         Reviewed by David Kilzer.
9558
9559         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
9560         (Trac.prototype.getCommitDataForRevisionRange): Instead of trying to parse the commit
9561         message, just return its text.
9562
9563         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
9564         (ViewController.prototype._domForRegressionRange): Instead of searching for test names in
9565         each commit's list of modified files, just search for test names anywhere in the commit's
9566         message.
9567
9568 2011-07-08  Adam Barth  <abarth@webkit.org>
9569
9570         sheriffbot is too spammy in IRC
9571         https://bugs.webkit.org/show_bug.cgi?id=64153
9572
9573         Reviewed by Eric Seidel.
9574
9575         Reporting failures in IRC worked well when the tree was greener than it
9576         is today.  Nowadays, this feature mostly just results in spam about
9577         false positives.  If we reach a greener state, we can bring this code
9578         back.
9579
9580         (Another possibility is to restrict this feature to builder bots,
9581         rather than including testers, as we did before this patch.)
9582
9583         * Scripts/webkitpy/tool/commands/sheriffbot.py:
9584         * Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
9585
9586 2011-07-08  Adam Roben  <aroben@apple.com>
9587
9588         Ensure commit-log-editor adds a "Patch by" line when the author and committer are different
9589
9590         Previously we were only adding a "Patch by" line when the ChangeLog contained a "Reviewed
9591         by" line. But some patches (like rollout patches) don't contain that line. Now we always add
9592         "Patch by" to the commit log regardless of the ChangeLog's contents.
9593
9594         Fixes <http://webkit.org/b/64127> Committer for r90588 is commit-queue@webkit.org, but
9595         should have been abarth@webkit.org
9596
9597         Reviewed by Anders Carlsson.
9598
9599         * Scripts/commit-log-editor:
9600         (createCommitMessage): Try to put the "Patch by" line just above the "Reviewed by" line, as
9601         before. If there is no "Reviewed by" line, try to put it just above the first modified file.
9602         If all else fails, put it at the end of the commit message.
9603         (patchAuthorshipString): Added. Code came from createCommitMessage.
9604
9605 2011-07-08  Herczeg Zoltan  <zherczeg@webkit.org>
9606
9607         Reviewed by Andreas Kling.
9608
9609         Adding myself as a reviewer.
9610
9611         * Scripts/webkitpy/common/config/committers.py:
9612
9613 2011-07-07  Kevin Ollivier  <kevino@theolliviers.com>
9614
9615         [wx] Unreviewed build fix, remove old files from the tree and update paths to new ones.
9616
9617         * wx/build: Removed.
9618         * wx/build/build_utils.py: Removed.
9619         * wx/build/settings.py: Removed.
9620         * wx/build/waf_extensions.py: Removed.
9621         * wx/build/wxpresets.py: Removed.
9622         * wx/packaging/build-mac-installer.py:
9623
9624 2011-07-07  Kevin Ollivier  <kevino@theolliviers.com>
9625
9626         [wx] Unreviewed build fix, more fixes for install name issues.
9627
9628         * wx/packaging/build-mac-installer.py:
9629
9630 2011-07-07  Dirk Pranke  <dpranke@chromium.org>
9631
9632         test-webkitpy fails on chromium win
9633         https://bugs.webkit.org/show_bug.cgi?id=64137
9634
9635         Reviewed by Eric Seidel.
9636
9637         test-webkitpy doesn't play nicely with the multiprocessing
9638         module on win32. An earlier change actually reenabled the tests
9639         on win32 by mistake.
9640
9641         This patch also fixes a few cases where path names will trip
9642         things up on win32 (e.g., by testing unix paths on windows).
9643         We do not lose any real coverage here with those fixes.
9644
9645         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
9646         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
9647         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
9648         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
9649
9650 2011-07-07  Leandro Pereira  <leandro@profusion.mobi>
9651
9652         [ImageDiff] Calculate/print difference right after reading baseline image.
9653         https://bugs.webkit.org/show_bug.cgi?id=64117
9654
9655         Reviewed by Kent Tamura.
9656
9657         Otherwise, ImageDiff will block on fgets() until the universe ends or you ^C
9658         it.  Whichever happens first.
9659
9660         * DumpRenderTree/efl/ImageDiff.cpp:
9661         (main):
9662
9663 2011-07-07  Raphael Kubo da Costa  <kubo@profusion.mobi>
9664
9665         [EFL] DRT: Leak cairo_t to make the pixel tests stop crashing
9666         https://bugs.webkit.org/show_bug.cgi?id=64107
9667
9668         Reviewed by Kent Tamura.
9669
9670         The pointer is later managed by BitmapContext, but it was being
9671         de-refed and deleted earlier by our RefPtr, causing crashes in all
9672         pixel tests.
9673
9674         We now call leakRef() to make sure it is not removed when our
9675         createBitmapContextFromWebView goes out of context.
9676
9677         * DumpRenderTree/efl/PixelDumpSupportEfl.cpp:
9678         (createBitmapContextFromWebView):
9679
9680 2011-07-07  Adam Roben  <aroben@apple.com>
9681
9682         Update TestFailures's title and header immediately upon navigation
9683
9684         Fixes <http://webkit.org/b/64125> TestFailures page seems slow to react on link clicks
9685
9686         Reviewed by Daniel Bates.
9687
9688         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
9689         (ViewController.prototype.loaded): Create and store an h1 element for displaying the page's
9690         title.
9691         (ViewController.prototype._displayBuilder): Set the page's title and clear out any old
9692         content immediately instead of waiting for the first set of results for the history
9693         analyzer. As we receive new results from the analyzer we'll just clear out the main content
9694         area and replace it with the new info.
9695         (ViewController.prototype._displayTesters): Set the page's title and clear out any old
9696         content immediately instead of waiting for the list of testers from the buildbot. Sprinkled
9697         in a little use of our removeAllChildren helper, too.
9698         (ViewController.prototype._setTitle): Added. Set's the page's title and the header text.
9699
9700 2011-07-07  Adam Roben  <aroben@apple.com>
9701
9702         Make Checkout use SCM's Executive instead of conjuring up its own
9703
9704         This will improve integration with the rest of webkitpy, particularly when invoked via
9705         webkit-patch.
9706
9707         Fixes <http://webkit.org/b/64115> REGRESSION (r90564): webkitpy's Checkout class uses
9708         Executive inappropriately
9709
9710         Reviewed by Adam Barth.
9711
9712         * Scripts/webkitpy/common/checkout/checkout.py:
9713         (Checkout.commit_message_for_this_commit): Use SCM.run instead of creating an Executive for
9714         our own use. SCM might have some super-special Executive that it uses under the covers, and
9715         we want to use it, too!
9716
9717         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
9718         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Mock the SCM.run
9719         method to call through to Executive.run_command.
9720
9721 2011-07-07  Adam Roben  <aroben@apple.com>
9722
9723         Make Term/ReadKey.pm an optional dependency of commit-log-editor
9724
9725         When Term/ReadKey.pm isn't present, the --regenerate-log switch will be non-functional.
9726
9727         Fixes <http://webkit.org/b/64113> REGRESSION (r90564):
9728         webkitpy.common.checkout.checkout_unittest failing on GTK bots due to missing
9729         Term/ReadKey.pm module
9730
9731         Reviewed by Xan Lopez.
9732
9733         * Scripts/commit-log-editor: Use Module::Load::Conditional::can_load to only load
9734         Term/ReadKey.pm if it's present. If it isn't present, ignore the --regenerate-log switch so
9735         that existing commit messages will be preserved. (Users can manually remove existing commit
9736         messages using git-reset.)
9737
9738 2011-07-07  Adam Roben  <aroben@apple.com>
9739
9740         Keep testing that commit messages containing Unicode are handled properly by webkitpy
9741
9742         Fixes <http://webkit.org/b/64109> REGRESSION (r90571): test-webkitpy no longer tests that we
9743         correctly handle Unicode in commit messages
9744
9745         Reviewed by Eric Seidel.
9746
9747         * Scripts/webkitpy/common/checkout/checkout_unittest.py: Put some Unicode characters in
9748         places in the ChangeLog that will end up being included in the commit message.
9749
9750 2011-07-07  Andreas Kling  <kling@webkit.org>
9751
9752         [Qt][WK2] Remove Symbian code.
9753         https://bugs.webkit.org/show_bug.cgi?id=64101
9754
9755         Reviewed by Benjamin Poulain.
9756
9757         * MiniBrowser/qt/BrowserWindow.cpp:
9758         (BrowserWindow::BrowserWindow):
9759         (BrowserWindow::screenshot):
9760         * MiniBrowser/qt/MiniBrowser.pro:
9761         * MiniBrowser/qt/MiniBrowserApplication.h:
9762         (WindowOptions::WindowOptions):
9763
9764 2011-07-07  Andreas Kling  <kling@webkit.org>
9765
9766         [Qt][WK2] Don't support ridiculous matrix of QT_NO_FEATURE combinations.
9767         https://bugs.webkit.org/show_bug.cgi?id=64099
9768
9769         Reviewed by Benjamin Poulain.
9770
9771         * MiniBrowser/qt/BrowserWindow.cpp:
9772         (BrowserWindow::openFile):
9773         (BrowserWindow::screenshot):
9774         (BrowserWindow::loadURLListFromFile):
9775
9776 2011-07-07  Adam Roben  <aroben@apple.com>
9777
9778         Completely ignore too-many-failures builds in TestFailures in most circumstances
9779
9780         Because a semi-arbitrary subset of tests are run in a too-many-failures build, we can't
9781         really use them to perform regression analysis. The only time we want to pay attention to
9782         too-many-failures builds is when we're trying to explain when the current bout of
9783         too-many-failures started.
9784
9785         Fixes <http://webkit.org/b/64106> TestFailures page sometimes claims a test started failing
9786         in a build that didn't even run it (because it exited early due to too many failues)
9787
9788         Reviewed by David Kilzer.
9789
9790         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
9791         (LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): Removed old, broken
9792         too-many-failures handling that would cause us to blame builds that didn't even run a given
9793         test for breaking it. Instead, skip over all too-many-failures builds unless the most recent
9794         build was itself a too-many-failures build.
9795
9796 2011-07-07  Adam Roben  <aroben@apple.com>
9797
9798         Teach webkitpy's Checkout class to use commit-log-editor to create commit messages
9799
9800         Fixes <http://webkit.org/b/26755> webkit-patch's commit messages are less readable than
9801         commit-log-editor's
9802
9803         Reviewed by David Kilzer.
9804
9805         * Scripts/webkitpy/common/checkout/checkout.py:
9806         (Checkout.commit_message_for_this_commit): Run commit-log-editor, passing it the paths of
9807         the modified ChangeLogs, to generate the commit message, rather than trying to generate one
9808         ourselves.
9809
9810         * Scripts/webkitpy/common/checkout/checkout_unittest.py: Updated the expected commit message
9811         to match commit-log-editor's format.
9812         (CommitMessageForThisCommitTest.setUp): Write the ChangeLogs into Tools and LayoutTests
9813         directories so we can see how the various entries get labeled in the commit message.
9814         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Create a mock SCM
9815         instance that knows how to find commit-log-editor and pass it to our Checkout instance.
9816         Don't bother capturing output, since there shouldn't be any.
9817
9818 2011-07-07  Adam Roben  <aroben@apple.com>
9819
9820         Add a --print-log option to commit-log-editor
9821
9822         When specified, commit-log-editor takes a set of ChangeLog paths on the command line,
9823         generates a commit message from those ChangeLogs, prints it to stdout, and exits.
9824
9825         Prep work for fixing <http://webkit.org/b/26755> webkit-patch's commit messages are less
9826         readable than commit-log-editor's
9827
9828         Reviewed by David Kilzer.
9829
9830         * Scripts/commit-log-editor: Changed to use Getopt::Long to parse options. Added --print-log
9831         option, which calls through to createCommitMessage, prints the result, and exits.
9832         (printUsageAndExit): Renamed from "usage" for clarity. Beefed up the usage statement to
9833         explain commit-log-editor's 2.5 (normal, --print-log, and --help) modes of operation.
9834
9835 2011-07-07  Adam Roben  <aroben@apple.com>
9836
9837         Extract commit-log-editor's commit-message-generation code into a separate function
9838
9839         As a bonus, we now skip a bunch of work in the case where we are generating a commit message
9840         for staged git changes.
9841
9842         Prep work for fixing <http://webkit.org/b/26755> webkit-patch's commit messages are less
9843         readable than commit-log-editor's
9844
9845         Reviewed by David Kilzer.
9846
9847         * Scripts/commit-log-editor: Moved the default definition of $endl up toward the top of the
9848         file so that it will be defined even if we aren't operating on an existing log message.
9849         Moved a bunch of top-level code to generate the commit message from here...
9850         (createCommitMessage): ...to here.
9851
9852 2011-07-07  Carlos Garcia Campos  <cgarcia@igalia.com>
9853
9854         [GTK] Open links in a new window when clicking with the middle button in MiniBrowser
9855         https://bugs.webkit.org/show_bug.cgi?id=63988
9856
9857         Reviewed by Martin Robinson.
9858
9859         * MiniBrowser/gtk/BrowserWindow.c:
9860         (browserWindowConstructed):
9861         (decidePolicyForNavigationAction): Ignore the action if a link is
9862         clicked with the middle buttonm and open the link in a new window.
9863         (browserWindowPolicyClientInit):
9864
9865 2011-07-07  Adam Barth  <abarth@webkit.org>
9866
9867         Remove style_references.py
9868         https://bugs.webkit.org/show_bug.cgi?id=64070
9869
9870         Reviewed by Eric Seidel.
9871
9872         We never really adopted the _references design, and this file is a bit
9873         of an orphan at this point.
9874
9875         * Scripts/check-webkit-style:
9876         * Scripts/webkitpy/style/checker.py:
9877         * Scripts/webkitpy/style/checker_unittest.py:
9878         * Scripts/webkitpy/style/checkers/python.py:
9879         * Scripts/webkitpy/style/checkers/test_expectations.py:
9880         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
9881         * Scripts/webkitpy/style/main_unittest.py:
9882         * Scripts/webkitpy/style_references.py: Removed.
9883
9884 2011-07-07  Eric Seidel  <eric@webkit.org>
9885
9886         Fix WebKit2 expected results search paths for Mac and Qt under new-run-webkit-tests
9887         https://bugs.webkit.org/show_bug.cgi?id=64056
9888
9889         Unreviewed.
9890
9891         I had written this unittest prior to landing my change, but
9892         had named it qt_unitest.py, so it wasn't running (nor was
9893         it added to my git repo).
9894
9895         Once I fixe the name of the unittest file, it was very easy
9896         to fix the error in webkit.py (which already had a FIXME).
9897
9898         * Scripts/webkitpy/layout_tests/port/qt_unittest.py: Added.
9899         * Scripts/webkitpy/layout_tests/port/webkit.py:
9900
9901 2011-07-07  Eric Seidel  <eric@webkit.org>
9902
9903         Unreviewed.  Add more debug logging about which test expectations we're using.
9904
9905         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
9906         * Scripts/webkitpy/layout_tests/port/webkit.py:
9907
9908 2011-07-06  Adam Barth  <abarth@webkit.org>
9909
9910         Remove unexpected_results.json
9911         https://bugs.webkit.org/show_bug.cgi?id=64065
9912
9913         Reviewed by Eric Seidel.
9914
9915         No one should be using this file any more.  This patch removes it from
9916         the face of the earth.  This patch also removes two NRWT-specific
9917         command line options that were implemented using
9918         unexpected_results.json.
9919
9920         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
9921         * Scripts/webkitpy/layout_tests/controllers/manager.py:
9922         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
9923         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
9924         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
9925         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
9926         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
9927
9928 2011-07-07  Adam Barth  <abarth@webkit.org>
9929
9930         REGRESSION(90520): chromium GPU tests produce output in the LayoutTests directory
9931         https://bugs.webkit.org/show_bug.cgi?id=64067
9932
9933         Reviewed by Eric Seidel.
9934
9935         The author of r90520 forgot to change the chromium_gpu.py implemenation
9936         of this method as well.
9937
9938         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
9939
9940 2011-07-06  Eric Seidel  <eric@webkit.org>
9941
9942         Fix WebKit2 expected results search paths for Mac and Qt under new-run-webkit-tests
9943         https://bugs.webkit.org/show_bug.cgi?id=64056
9944
9945         Reviewed by Adam Barth.
9946
9947         This fixes the MacPort's version of baseline_search_path to include
9948         wk2 fallback like WebKitPort's does.  This also re-works the QtPort
9949         to set self._name (as other ports do) so that WebKitPort's default
9950         implementations will understand Qt operating system flavors correctly
9951         (including when running wk2 tests).
9952
9953         In trying to test this, I realized that nearly all of our port-tests
9954         use the default FileSystem(), User() and Executive() objects, which
9955         is really bad, because they expose details of your system in the unittest results!
9956
9957         Once I fixed many of the Port() calls to pass MockFileSystem(), then
9958         MacPort.test_tests_for_other_platforms started failing, and I had to
9959         teach MockFileSystem.glob how to handle directories to make it work again.
9960
9961         I removed the useless PortTestCase.test_baseline_search_path and replaced
9962         it with more useful test_baseline_search_path tests in MacPortTest and QtPortTest.
9963
9964         * Scripts/webkitpy/common/config/ports.py:
9965         * Scripts/webkitpy/common/system/filesystem_mock.py:
9966         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
9967         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
9968         * Scripts/webkitpy/layout_tests/port/mac.py:
9969         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
9970         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
9971         * Scripts/webkitpy/layout_tests/port/qt.py:
9972         * Scripts/webkitpy/layout_tests/port/test_files.py:
9973         * Scripts/webkitpy/layout_tests/port/win.py:
9974         * Scripts/webkitpy/to_be_moved/deduplicate_tests.py:
9975         * Scripts/webkitpy/tool/mocktool.py:
9976
9977 2011-07-06  Adam Barth  <abarth@webkit.org>
9978
9979         Wean rebaseline-server off unexpected_results.json
9980         https://bugs.webkit.org/show_bug.cgi?id=64063
9981
9982         Reviewed by Eric Seidel.
9983
9984         In the process of changing this code to use full_results.json, I
9985         noticed that the code was broken (because it wasn't tested).  This
9986         patch also adds test coverage for the broken code.
9987
9988         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
9989         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
9990         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
9991
9992 2011-07-06  Adam Barth  <abarth@webkit.org>
9993
9994         Wean resultsjsonparser off unexpected_results.json
9995         https://bugs.webkit.org/show_bug.cgi?id=64061
9996
9997         Reviewed by Eric Seidel.
9998
9999         We're about to delete unexpected_results.json, so we need to remove all
10000         the users.
10001
10002         * Scripts/webkitpy/common/config/ports.py:
10003         * Scripts/webkitpy/common/net/resultsjsonparser.py:
10004         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
10005         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
10006
10007 2011-07-06  Adam Barth  <abarth@webkit.org>
10008
10009         Cleanup result_summary.py
10010         https://bugs.webkit.org/show_bug.cgi?id=64057
10011
10012         Reviewed by Eric Seidel.
10013
10014         This class had a bunch of out-of-date docstrings that no longer make
10015         any sense.
10016
10017         * Scripts/webkitpy/layout_tests/models/result_summary.py:
10018
10019 2011-07-06  Adam Barth  <abarth@webkit.org>
10020
10021         Tweak the ChangeLog for DEPS rolls to say "Unreviewed" so that the
10022         patches can be landed by the commit-queue.
10023
10024         * Scripts/webkitpy/tool/steps/preparechangelogfordepsroll.py:
10025
10026 2011-07-06  Adam Barth  <abarth@webkit.org>
10027
10028         Normalize callers of logging.getLogger to use __name__
10029         https://bugs.webkit.org/show_bug.cgi?id=64054
10030
10031         Reviewed by Eric Seidel.
10032
10033         Most callers were already using __name__.  This patch makes everyone
10034         consistent.
10035
10036         * Scripts/webkitpy/common/checkout/diff_parser.py:
10037         * Scripts/webkitpy/common/checkout/scm/svn.py:
10038         * Scripts/webkitpy/common/net/statusserver.py:
10039         * Scripts/webkitpy/common/system/executive.py:
10040         * Scripts/webkitpy/common/system/file_lock.py:
10041         * Scripts/webkitpy/common/system/user.py:
10042         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
10043         * Scripts/webkitpy/layout_tests/models/result_summary.py:
10044         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
10045         * Scripts/webkitpy/layout_tests/port/chromium.py:
10046         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
10047         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
10048         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
10049         * Scripts/webkitpy/layout_tests/port/gtk.py:
10050         * Scripts/webkitpy/layout_tests/port/http_lock.py:
10051         * Scripts/webkitpy/layout_tests/port/mac.py:
10052         * Scripts/webkitpy/layout_tests/port/qt.py:
10053         * Scripts/webkitpy/layout_tests/port/server_process.py:
10054         * Scripts/webkitpy/layout_tests/port/webkit.py:
10055         * Scripts/webkitpy/layout_tests/port/win.py:
10056         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
10057         * Scripts/webkitpy/layout_tests/servers/http_server.py:
10058         * Scripts/webkitpy/layout_tests/servers/http_server_base.py:
10059         * Scripts/webkitpy/layout_tests/servers/websocket_server.py:
10060         * Scripts/webkitpy/layout_tests/views/printing.py:
10061         * Scripts/webkitpy/python24/versioning.py:
10062         * Scripts/webkitpy/style/checker.py:
10063         * Scripts/webkitpy/style/checkers/test_expectations.py:
10064         * Scripts/webkitpy/style/patchreader.py:
10065         * Scripts/webkitpy/to_be_moved/update_webgl_conformance_tests.py:
10066
10067 2011-07-06  Adam Barth  <abarth@webkit.org>
10068
10069         Move model classes out of layout_package and into models
10070         https://bugs.webkit.org/show_bug.cgi?id=64053
10071
10072         Reviewed by Eric Seidel.
10073
10074         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
10075         * Scripts/webkitpy/common/net/layouttestresults.py:
10076         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
10077         * Scripts/webkitpy/common/net/resultsjsonparser.py:
10078         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
10079         * Scripts/webkitpy/layout_tests/controllers/manager.py:
10080         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
10081         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
10082         * Scripts/webkitpy/layout_tests/controllers/worker.py:
10083         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
10084         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
10085         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
10086         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py: Removed.
10087         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py: Removed.
10088         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py: Removed.
10089         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py: Removed.
10090         * Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py: Removed.
10091         * Scripts/webkitpy/layout_tests/layout_package/test_input.py: Removed.
10092         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
10093         * Scripts/webkitpy/layout_tests/layout_package/test_results.py: Removed.
10094         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py: Removed.
10095         * Scripts/webkitpy/layout_tests/models: Added.
10096         * Scripts/webkitpy/layout_tests/models/__init__.py: Added.
10097         * Scripts/webkitpy/layout_tests/models/result_summary.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/result_summary.py.
10098         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_expectations.py.
10099         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py.
10100         * Scripts/webkitpy/layout_tests/models/test_failures.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_failures.py.
10101         * Scripts/webkitpy/layout_tests/models/test_failures_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py.
10102         * Scripts/webkitpy/layout_tests/models/test_input.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_input.py.
10103         * Scripts/webkitpy/layout_tests/models/test_results.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_results.py.
10104         * Scripts/webkitpy/layout_tests/models/test_results_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py.
10105         * Scripts/webkitpy/layout_tests/port/chromium.py:
10106         * Scripts/webkitpy/layout_tests/views/printing.py:
10107         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
10108         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
10109         * Scripts/webkitpy/style_references.py:
10110         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
10111         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
10112         * Scripts/webkitpy/tool/bot/flakytestreporter_unittest.py:
10113         * Scripts/webkitpy/tool/commands/queues_unittest.py:
10114         * Scripts/webkitpy/tool/commands/rebaseline.py:
10115
10116 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
10117
10118         nrwt: fix typo preventing http startup on windows
10119         https://bugs.webkit.org/show_bug.cgi?id=64050
10120
10121         Reviewed by Eric Siedel.
10122
10123         NRWT checks whether it needs to start the servers by looking
10124         for "/http/" in the test names. After r90520, the leading slash
10125         is not present, so the check isn't quite right.
10126
10127         * Scripts/webkitpy/layout_tests/controllers/manager.py:
10128         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
10129
10130 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
10131
10132         nrwt: remove --use-apache from the command line
10133         https://bugs.webkit.org/show_bug.cgi?id=63358
10134
10135         Reviewed by Adam Barth.
10136
10137         Re-land the fix in bug 63358 - the prior fix was flawed in that
10138         we were trying to start the http server during check_sys_deps()
10139         but not actually setting up the environment properly. This
10140         broken prior to this change, but exposed by it, since the patch
10141         changes the chromium port to call the base class's checks.
10142
10143         * Scripts/webkitpy/common/system/executive.py:
10144         * Scripts/webkitpy/layout_tests/port/base.py:
10145         * Scripts/webkitpy/layout_tests/port/chromium.py:
10146         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
10147         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
10148         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
10149         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
10150         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
10151         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
10152
10153 2011-07-06  Adam Barth  <abarth@webkit.org>
10154
10155         Move view-related code out of layout_package and into views
10156         https://bugs.webkit.org/show_bug.cgi?id=64051
10157
10158         Reviewed by Eric Seidel.
10159
10160         * Scripts/webkitpy/layout_tests/controllers/manager.py:
10161         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
10162         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
10163         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
10164         * Scripts/webkitpy/layout_tests/layout_package/metered_stream.py: Removed.
10165         * Scripts/webkitpy/layout_tests/layout_package/metered_stream_unittest.py: Removed.
10166         * Scripts/webkitpy/layout_tests/layout_package/printing.py: Removed.
10167         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py: Removed.
10168         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
10169         * Scripts/webkitpy/layout_tests/views: Added.
10170         * Scripts/webkitpy/layout_tests/views/__init__.py: Added.
10171         * Scripts/webkitpy/layout_tests/views/metered_stream.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/metered_stream.py.
10172         * Scripts/webkitpy/layout_tests/views/metered_stream_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/metered_stream_unittest.py.
10173         * Scripts/webkitpy/layout_tests/views/printing.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/printing.py.
10174         * Scripts/webkitpy/layout_tests/views/printing_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py.
10175
10176 2011-07-06  Adam Barth  <abarth@webkit.org>
10177
10178         Add goofy comments to explain why we have these files.
10179
10180         * Scripts/webkitpy/layout_tests/controllers/__init__.py:
10181         * Scripts/webkitpy/layout_tests/layout_package/__init__.py:
10182
10183 2011-07-06  Adam Barth  <abarth@webkit.org>
10184
10185         Move a bunch of classes out of layout_tests.layout_package and into layout_tests.controllers
10186         https://bugs.webkit.org/show_bug.cgi?id=64048
10187
10188         Reviewed by Eric Seidel.
10189
10190         * Scripts/webkitpy/layout_tests/controllers: Added.
10191         * Scripts/webkitpy/layout_tests/controllers/__init__.py: Added.
10192         * Scripts/webkitpy/layout_tests/controllers/manager.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager.py.
10193         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py.
10194         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py.
10195         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py.
10196         * Scripts/webkitpy/layout_tests/controllers/message_broker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/message_broker.py.
10197         * Scripts/webkitpy/layout_tests/controllers/message_broker_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py.
10198         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py.
10199         * Scripts/webkitpy/layout_tests/controllers/worker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/worker.py.
10200         * Scripts/webkitpy/layout_tests/layout_package/manager.py: Removed.
10201         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py: Removed.
10202         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py: Removed.
10203         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py: Removed.
10204         * Scripts/webkitpy/layout_tests/layout_package/message_broker.py: Removed.
10205         * Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py: Removed.
10206         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
10207         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py: Removed.
10208         * Scripts/webkitpy/layout_tests/layout_package/worker.py: Removed.
10209         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
10210
10211 2011-07-06  Adam Barth  <abarth@webkit.org>
10212
10213         Attempt to fix unittest on Leopard.  Let's enumerate the keys in sorted
10214         order so the test results are predictable.
10215
10216         * Scripts/webkitpy/tool/bot/irc_command.py:
10217         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
10218
10219 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
10220
10221         Reviewed by Eric Siedel.
10222
10223         nrwt: make the code be consistent about using test names instead of filenames or paths
10224
10225         https://bugs.webkit.org/show_bug.cgi?id=63597
10226
10227         It can be hard to figure out where in the code we're assuming
10228         test names are listed as unix-style relative filenames and where
10229         they are either absolute paths or relative paths following the
10230         host filesystem convention.
10231
10232         This patch changes things so that everything outside of the
10233         Port object uses (and must assume) unix-style relative
10234         filenames (with one exception, which is specifying host-local
10235         filenames as a list of test arguments on the command line).
10236
10237         This should make things clearer and more consistent, and also
10238         removes a lot of path conversion calls.
10239
10240         The changes in this patch outside of port/base.py are largely
10241         cosmetic changes of "filename" to "test". We add a few routines
10242         to the base Port implementation to manipulate test names to
10243         split them, figure out if we're referring to a directory of
10244         tests, etc.
10245
10246         * Scripts/webkitpy/common/net/layouttestresults.py:
10247         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
10248         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
10249         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
10250         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
10251         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
10252         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
10253         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
10254         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
10255         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
10256         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
10257         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
10258         * Scripts/webkitpy/layout_tests/layout_package/test_input.py:
10259         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
10260         * Scripts/webkitpy/layout_tests/layout_package/test_results.py:
10261         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py:
10262         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
10263         * Scripts/webkitpy/layout_tests/port/base.py:
10264         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
10265         * Scripts/webkitpy/layout_tests/port/chromium.py:
10266         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
10267         * Scripts/webkitpy/layout_tests/port/dryrun.py:
10268         * Scripts/webkitpy/layout_tests/port/test.py:
10269         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
10270         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
10271         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
10272         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
10273         * Scripts/webkitpy/tool/bot/flakytestreporter.py:
10274         * Scripts/webkitpy/tool/commands/rebaseline.py:
10275         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
10276
10277 2011-07-06  Chris Guillory  <ctguil@chromium.org>
10278
10279         No review necessary.
10280
10281         Adding myself to committers list in committers.py.
10282         https://bugs.webkit.org/show_bug.cgi?id=63478
10283
10284         * Scripts/webkitpy/common/config/committers.py:
10285
10286 2011-07-06  Johnny Ding  <jnd@chromium.org>
10287
10288         https://bugs.webkit.org/show_bug.cgi?id=63857
10289         Implement createPopupMenu for Chromium DRT.
10290
10291         Reviewed by Tony Chang.
10292
10293         * DumpRenderTree/chromium/WebViewHost.cpp:
10294         (WebViewHost::createPopupMenu):
10295         (WebViewHost::~WebViewHost):
10296         * DumpRenderTree/chromium/WebViewHost.h:
10297
10298 2011-07-06  Mihai Parparita  <mihaip@chromium.org>
10299
10300         Unreviewed, rolling out r90503.
10301         http://trac.webkit.org/changeset/90503
10302         https://bugs.webkit.org/show_bug.cgi?id=63358
10303
10304         Makes Chromium Windows canary bots hang.
10305
10306         * Scripts/webkitpy/layout_tests/port/base.py:
10307         * Scripts/webkitpy/layout_tests/port/chromium.py:
10308         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
10309         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
10310         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
10311         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
10312         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
10313         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
10314
10315 2011-07-06  Adam Barth  <abarth@webkit.org>
10316
10317         Move buildbot URL in webkitpy to config.urls
10318         https://bugs.webkit.org/show_bug.cgi?id=64041
10319
10320         Reviewed by Eric Seidel.
10321
10322         This patch prepares for using the buildbot class to talk to the
10323         Chromium buildbots.
10324
10325         * Scripts/webkitpy/common/config/urls.py:
10326         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
10327         * Scripts/webkitpy/tool/commands/download_unittest.py:
10328         * Scripts/webkitpy/tool/commands/queries.py:
10329         * Scripts/webkitpy/tool/mocktool.py:
10330         * Scripts/webkitpy/tool/steps/ensurebuildersaregreen.py:
10331
10332 2011-07-06  Kevin Ollivier  <kevino@theolliviers.com>
10333
10334         [wx] Unreviewed build fix, add the correct path to the waf build settings.
10335
10336         * wx/packaging/build-win-installer.py:
10337
10338 2011-07-06  Adam Barth  <abarth@webkit.org>
10339
10340         Add missing include.
10341
10342         * Scripts/webkitpy/tool/bot/irc_command.py:
10343
10344 2011-07-06  Adam Barth  <abarth@webkit.org>
10345
10346         Add roll-chromium-deps command to sheriff-bot
10347         https://bugs.webkit.org/show_bug.cgi?id=64037
10348
10349         Reviewed by Eric Seidel.
10350
10351         Lo, and the DEPS were rolled on command.
10352
10353         * Scripts/webkitpy/tool/bot/irc_command.py:
10354         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
10355         * Scripts/webkitpy/tool/bot/sheriff.py:
10356         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
10357         * Scripts/webkitpy/tool/commands/roll.py:
10358         * Scripts/webkitpy/tool/commands/roll_unittest.py:
10359         * Scripts/webkitpy/tool/steps/updatechromiumdeps.py:
10360
10361 2011-07-06  Eric Seidel  <eric@webkit.org>
10362
10363         Split Driver, DriverOutput and DriverInput out into a new driver.py file
10364         https://bugs.webkit.org/show_bug.cgi?id=64031
10365
10366         Reviewed by Adam Barth.
10367
10368         This ended up kinda large and I'm happy to break it up if desired.
10369
10370         As part of cleaning up the Port module, I split Driver, DriverInput and DriverOutput
10371         out into its own driver.py file.
10372
10373         I also cleaned up how we import Port objects, by making port/__init__.py
10374         expose Port, Driver, DriverInput, and DriverOutput.  I removed exposure of
10375         some various test objects as they were causing circular dependencies.
10376         base.py is now an implementation detail of port.  No longer is "base" referenced
10377         anywhere in the code.
10378
10379         I also made Driver.__init__ non-virtual so that sub-classers could all share
10380         the data storage on Driver.  This more closely matches how other subclasses
10381         work throughout our python code.
10382
10383         Standard line-unwrapping and doc-string removal as I went through the code.
10384
10385         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
10386         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
10387         * Scripts/webkitpy/layout_tests/port/__init__.py:
10388         * Scripts/webkitpy/layout_tests/port/base.py:
10389         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
10390         * Scripts/webkitpy/layout_tests/port/chromium.py:
10391         * Scripts/webkitpy/layout_tests/port/driver.py: Added.
10392         * Scripts/webkitpy/layout_tests/port/dryrun.py:
10393         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
10394         * Scripts/webkitpy/layout_tests/port/test.py:
10395         * Scripts/webkitpy/layout_tests/port/webkit.py:
10396         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
10397         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
10398         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
10399
10400 2011-07-06  Adam Roben  <aroben@apple.com>
10401
10402         Clear TestFailures's getNumberOfFailingTests cache
10403
10404         Now that we support NRWT, we need to clear out old zero values for the NRWT bots that might
10405         have been cached by old versions of TestFailures.
10406
10407         Fixes <http://webkit.org/b/64029> TestFailures page still doesn't show NRWT bots if you've
10408         used it recently
10409
10410         Reviewed by Adam Barth.
10411
10412         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
10413         (Builder.prototype.getNumberOfFailingTests): Added a .version property to the cache data
10414         that must match the current version. Otherwise we'll ignore the cached data.
10415
10416 2011-06-24  Dirk Pranke  <dpranke@chromium.org>
10417
10418         Reviewed by Adam Barth.
10419
10420         nrwt: remove --use-apache from the command line
10421         https://bugs.webkit.org/show_bug.cgi?id=63358
10422
10423         This change removes the --use-apache command line argument. It
10424         was initially put in when we were trying to get the cygwin
10425         apache instance to work with Chromium win, but that code has
10426         bitrotted and doesn't work at all now.
10427
10428         Arguably we should remove all of the code to allow for a choice
10429         of web servers, but since we may still want to switch off of
10430         LigHTTPd at some point on Windows, I'll leave the rest in for
10431         now (or at least yank it in a different change.)
10432
10433         * Scripts/webkitpy/layout_tests/port/base.py:
10434         * Scripts/webkitpy/layout_tests/port/chromium.py:
10435         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
10436         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
10437         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
10438         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
10439
10440 2011-07-06  Dimitri Glazkov  <dglazkov@chromium.org>
10441
10442         garden-o-matic should use a favicon to indicate current state.
10443         https://bugs.webkit.org/show_bug.cgi?id=64027
10444
10445         Reviewed by Adam Barth.
10446
10447         * Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-green.png: Added.
10448         * Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-red.png: Added.
10449         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html: Added blurb about icons.
10450         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js: Added setting icon state.
10451         * Scripts/webkitpy/tool/servers/gardeningserver.py: Allowed the use of favicons.
10452
10453 2011-07-06  Daniel Bates  <dbates@rim.com>
10454
10455         Fix webkitpy error after changeset 90483 <http://trac.webkit.org/changeset/90483>; Add missing
10456         parameter to _fetch_file_from_results lamdba function.
10457
10458         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
10459
10460 2011-07-06  Kevin Ollivier  <kevino@theolliviers.com>
10461
10462         [wx] Unreviewed build fix. Generate the proper install_name for dylibs on Mac.
10463
10464         * wx/packaging/build-mac-installer.py:
10465
10466 2011-07-06  Xan Lopez  <xlopez@igalia.com>
10467
10468         [GTK] Only set env. variables in start when strictly needed
10469         https://bugs.webkit.org/show_bug.cgi?id=64026
10470
10471         Reviewed by Eric Seidel.
10472
10473         * Scripts/webkitpy/layout_tests/port/gtk.py: add a
10474         setup_environ_for_server method in GtkPort and set as many
10475         env. variables as possible there.
10476
10477 2011-07-06  Adam Roben  <aroben@apple.com>
10478
10479         Teach TestFailures how to load, parse, and interpret NRWT test results
10480
10481         Fixes <http://webkit.org/b/61877> TestFailures page doesn't show testers that use
10482         new-run-webkit-tests
10483
10484         Reviewed by Adam Barth.
10485
10486         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
10487         (Builder.prototype.failureDiagnosisTextAndURL): Added support for the new 'flaky' failure
10488         type. For now we don't account for image-only flakes (but TestFailures doesn't deal with
10489         pixel tests at all currently).
10490         (Builder.prototype.getNumberOfFailingTests): Relaxed the regex used for parsing the number
10491         of failing tests from the buildbot output. Make sure not to count "new passes" (a new
10492         category introduced by NRWT) as failures.
10493
10494         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
10495         (LayoutTestResultsLoader.prototype.start): Moved a bunch of code from here to
10496         _fetchAndParseORWTResults. This function now attempts to load NRWT results, then falls back
10497         to loading ORWT results.
10498         (LayoutTestResultsLoader.prototype._fetchAndParseNRWTResults): Added. Tries to load and
10499         parse the full_results.json file for the given build.
10500         (LayoutTestResultsLoader.prototype._fetchAndParseORWTResults): Added. Code came from start.
10501         Fixed a bug along the way where we were sometimes calling the error callback instead of the
10502         success callback when all tests passed.
10503
10504         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NRWTResultsParser.js: Added.
10505         (NRWTResultsParser): Do-nothing constructor.
10506         (NRWTResultsParser.prototype.parse): Uses eval() (eww!) to get the test results out of the
10507         JS string, then iterates over all the tests in the results data and builds up a data
10508         structure matching what ORWTResultsParser returns.
10509
10510         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
10511         (Array.prototype.contains):
10512         (String.prototype.contains):
10513         Added these simple helper functions.
10514
10515         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
10516         NRWTResultsParser.js.
10517
10518 2011-07-06  Adam Roben  <aroben@apple.com>
10519
10520         Extract code to parse ORWT's results.html file into its own class
10521
10522         Prep work for fixing <http://webkit.org/b/61877> TestFailures page doesn't show testers that
10523         use new-run-webkit-tests
10524
10525         Reviewed by Adam Barth.
10526
10527         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
10528         (LayoutTestResultsLoader.prototype.start): Moved a bunch of code from here...
10529
10530         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ORWTResultsParser.js: Added.
10531         (ORWTResultsParser):
10532         (ORWTResultsParser.prototype.parse):
10533         ...to here.
10534
10535         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
10536         ORWTResultsParser.js.
10537
10538 2011-07-06  Eric Seidel  <eric@webkit.org>
10539
10540         Unreviewed.  Fix a unittest from hitting the network. :)
10541
10542         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
10543
10544 2011-07-06  Eric Seidel  <eric@webkit.org>
10545
10546         Unreviewed.  Fixing an exception in test-webkitpy, caused by bug 64006.
10547
10548         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
10549
10550 2011-07-06  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
10551
10552         Run an Xvfb instance to run the API tests on, like we do for the
10553         layout tests.
10554
10555         Reviewed by Xan Lopez.
10556
10557         * Scripts/run-gtk-tests:
10558
10559 2011-07-06  Daniel Bates  <dbates@webkit.org>
10560
10561         Make SCM unit tests faster
10562         https://bugs.webkit.org/show_bug.cgi?id=63883
10563
10564         Speeds up the SCM unit tests by a factor of 4. Currently, we create a
10565         mock SVN repo for each test_ method in SVNTest and GitTest and creating
10566         this repo is expensive.
10567
10568         Instead, it is significantly faster to create the mock SVN repo once
10569         and then perform a filesystem copy of it for each test_ method.
10570
10571         Note, Python 2.7's unittest module implements support for per class and
10572         per module setup and tear down methods which could be used to implement
10573         similar functionality. At the time of writing, test-webkitpy is designed
10574         to support Python 2.5. So, we can't take advantage of these Python 2.7
10575         features :(
10576
10577         Reviewed by Eric Seidel.
10578
10579         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
10580
10581 2011-07-06  Adam Barth  <abarth@webkit.org>
10582
10583         Remove unused parameter from _add_test in test_expectations.py
10584         https://bugs.webkit.org/show_bug.cgi?id=64008
10585
10586         Reviewed by Eric Seidel.
10587
10588         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
10589
10590 2011-07-06  Adam Barth  <abarth@webkit.org>
10591
10592         Use old-run-webkit-tests when --leaks is present
10593         https://bugs.webkit.org/show_bug.cgi?id=64012
10594
10595         Reviewed by Adam Roben.
10596
10597         We'll remove this branch once support of --leaks is added to
10598         new-run-webkit-tests.
10599
10600         * Scripts/run-webkit-tests:
10601
10602 2011-07-06  Benjamin Poulain  <benjamin@webkit.org>
10603
10604         [Qt] [WK2] Add touch mocking to Qt's MiniBrowser
10605         https://bugs.webkit.org/show_bug.cgi?id=63995
10606
10607         Reviewed by Andreas Kling.
10608
10609         The mouse events are intercepted through QCoreApplication::notify(),
10610         and fake touch events are generated for mouse events when necessary.
10611
10612         If touch events are received from the platform, we assume the current
10613         hardware has a touch screen and we stop generating fake events.
10614
10615         * MiniBrowser/qt/MiniBrowserApplication.cpp:
10616         (isTouchEvent):
10617         (isMouseEvent):
10618         (MiniBrowserApplication::MiniBrowserApplication):
10619         (MiniBrowserApplication::notify):
10620         * MiniBrowser/qt/MiniBrowserApplication.h:
10621
10622 2011-07-06  Xan Lopez  <xlopez@igalia.com>
10623
10624         Reviewed by Gustavo Noronha.
10625
10626         * Scripts/webkitpy/layout_tests/port/gtk.py: use Popen correctly.
10627
10628 2011-07-06  Eric Seidel  <eric@webkit.org>
10629
10630         webkit-patch failure-reason does not understand NRWT results
10631         https://bugs.webkit.org/show_bug.cgi?id=64006
10632
10633         Reviewed by Adam Barth.
10634
10635         Adam Barth tells me failure-reason should be deleted soon,
10636         but aroben's fancy new TestFailures/ page doesn't work for
10637         NRWT yet, and I needed to know when
10638         third-party-cookie-relaxing started failing.
10639
10640         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
10641         * Scripts/webkitpy/tool/commands/queries.py:
10642
10643 2011-07-06  Xan Lopez  <xlopez@igalia.com>
10644
10645         [GTK] Add missing environment variables in NWRT
10646         https://bugs.webkit.org/show_bug.cgi?id=64004
10647
10648         Reviewed by Gustavo Noronha Silva.
10649
10650         Add missing environment variables needed by the GTK+ port.
10651
10652         * Scripts/webkitpy/layout_tests/port/gtk.py: ditto.
10653
10654 2011-07-06  Adam Roben  <aroben@apple.com>
10655
10656         Make run-api-tests output intelligible on the bots
10657
10658         run-api-tests's output on the bots was very hard to read for two reasons:
10659           1) It was mixing its own output with gtest's
10660           2) It was using ANSI escape sequences to print colored text, but the bots don't support
10661              that
10662
10663         Now, the --verbose flag turns off almost all of run-api-tests's own output so that gtest's
10664         output will not be obscured. We still print "Timeout" messages even in verbose mode, since
10665         gtest doesn't have any native support for timeouts. Also, when our output is being
10666         redirected to a file, we don't print ANSI escape sequences.
10667
10668         Fixes <http://webkit.org/b/63996> It's very hard to read run-api-tests output on the bots
10669
10670         Reviewed by Sam Weinig.
10671
10672         * Scripts/run-api-tests: Don't use the :constants interface of Term::ANSIColor. We want to
10673         use the colored function instead.
10674         (runAllTestsInSuite): Don't print out suite names in verbose mode; gtest will give us enough
10675         context that they aren't needed.
10676         (runTest): Don't print out test names or pass/fail messages in verbose mode; gtest will do
10677         that for us. Also, use the new possiblyColored function instead of always coloring output.
10678         (possiblyColored): Added. When printing to a tty, returns the string with the appropriate
10679         ANSI color escape sequences added. Otherwise just returns the string unmodified.
10680
10681 2011-07-06  Xan Lopez  <xlopez@igalia.com>
10682
10683         Reviewed by Gustavo Noronha.
10684
10685         Launch Xvfb (mostly) the same way we were launching it in the
10686         bots, for maximum compatibility.
10687
10688         * Scripts/webkitpy/layout_tests/port/gtk.py:
10689
10690 2011-07-06  Andreas Kling  <kling@webkit.org>
10691
10692         [Qt][WK2] Split Qt API into two different web views (touch and desktop)
10693         https://bugs.webkit.org/show_bug.cgi?id=63950
10694
10695         Co-authored with Benjamin Poulain <benjamin@webkit.org>.
10696
10697         Reviewed by Kenneth Rohde Christiansen.
10698
10699         Port WTR/Qt to use QDesktopWebView and force a 800x600 viewport.
10700
10701         The Qt MiniBrowser now defaults to QDesktopWebView but can be started
10702         in QTouchWebView mode by passing -touch on the command line.
10703
10704         A lot of functionality has been removed and/or disabled from
10705         MiniBrowser until we reintroduce the necessary API's.
10706
10707         * MiniBrowser/qt/BrowserView.cpp:
10708         (BrowserView::BrowserView):
10709         (BrowserView::~BrowserView):
10710         (BrowserView::load):
10711         (BrowserView::view):
10712         (BrowserView::touchWebView):
10713         (BrowserView::desktopWebView):
10714         * MiniBrowser/qt/BrowserView.h:
10715         * MiniBrowser/qt/BrowserWindow.cpp:
10716         (newPageFunction):
10717         (BrowserWindow::BrowserWindow):
10718         (BrowserWindow::webView):
10719         (BrowserWindow::newWindow):
10720         (BrowserWindow::showUserAgentDialog):
10721         (BrowserWindow::updateUserAgentList):
10722         * MiniBrowser/qt/BrowserWindow.h:
10723         * MiniBrowser/qt/MiniBrowserApplication.cpp:
10724         (MiniBrowserApplication::handleUserOptions):
10725         * MiniBrowser/qt/MiniBrowserApplication.h:
10726         (WindowOptions::WindowOptions):
10727         * MiniBrowser/qt/UrlLoader.cpp:
10728         (UrlLoader::UrlLoader):
10729         * MiniBrowser/qt/main.cpp:
10730         (main):
10731         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
10732         (WTR::WebView::wkView):
10733         (WTR::WebView::pageRef):
10734         (WTR::WebView::WebView):
10735         (WTR::PlatformWebView::page):
10736         (WTR::PlatformWebView::windowFrame):
10737         (WTR::PlatformWebView::setWindowFrame):
10738
10739 2011-07-06  Eric Seidel  <eric@webkit.org>
10740
10741         Unreviewed..
10742
10743         * Scripts/run-webkit-tests:
10744         (useNewRunWebKitTests):
10745          - Fix the pass-thru of --gtk and --qt for old-run-webkit-tests,
10746            it was previously broken by the existence of isGtk/isQt checks which
10747            removed the --gtk/--qt args from @ARGV.
10748
10749 2011-07-06  Eric Seidel  <eric@webkit.org>
10750
10751         new-run-webkit-tests uses a 35s timeout for Mac and 6s for all other ports
10752         https://bugs.webkit.org/show_bug.cgi?id=63983
10753
10754         Reviewed by Xan Lopez.
10755
10756         All WebKit ports want to use a 35s timeout.  Only chromium currently
10757         uses a 6s timeout.  Eventually we'll move all WebKit ports to 6s too
10758         to further speed up new-run-webkit-tests runs, but that's going to be a while.
10759
10760         * Scripts/webkitpy/layout_tests/port/mac.py:
10761         * Scripts/webkitpy/layout_tests/port/webkit.py:
10762         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
10763
10764 2011-07-06  Xan Lopez  <xlopez@igalia.com>
10765
10766         Unreviewed.
10767
10768         NWRT actually seems to work locally, so it should work in the bots
10769         as long as we take care of the TIMEOUT issue (which is being fixed
10770         as I write this). Let's give this another shot and fix whatever
10771         breaks.
10772
10773         * Scripts/run-webkit-tests:
10774         (useNewRunWebKitTests):
10775
10776 2011-07-06  Xan Lopez  <xlopez@igalia.com>
10777
10778         Unreviewed.
10779
10780         Disable NWRT on GTK. At the very least it needs an updated exected
10781         results file since NWRT is more sensitive to timeouts, not sure
10782         what the exact problem is.
10783
10784         * Scripts/run-webkit-tests:
10785         (useNewRunWebKitTests):
10786
10787 2011-07-06  Adam Barth  <abarth@webkit.org>
10788
10789         Fix global variable leak detected by noglobals.
10790
10791         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
10792
10793 2011-07-06  Adam Barth  <abarth@webkit.org>
10794
10795         garden-o-matic shouldn't show flaky tests by default
10796         https://bugs.webkit.org/show_bug.cgi?id=63979
10797
10798         Reviewed by Ojan Vafai.
10799
10800         If any of the actual test results are PASS, the test is flaky and we
10801         can ignore it (at least for now).  If we get awesome at dealing with
10802         failing tests, we might want to tackle flaky tests by refering the user
10803         to the flakiness dashboard, which is way more awesome at dealing with
10804         flaky tests that this tool ever will be.
10805
10806         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
10807         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
10808
10809 2011-07-05  Adam Barth  <abarth@webkit.org>
10810
10811         garden-o-matic should list the failing tests
10812         https://bugs.webkit.org/show_bug.cgi?id=63978
10813
10814         Reviewed by Eric Seidel.
10815
10816         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js: Added.
10817         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
10818         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
10819         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
10820         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
10821         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
10822         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js: Added.
10823         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js: Added.
10824         * Scripts/webkitpy/tool/servers/gardeningserver.py:
10825
10826
10827 2011-07-05  Eric Seidel  <eric@webkit.org>
10828
10829         Re-enable NRWT on Leopard now that bug 63973 is fixed.
10830
10831         Reviewed by Adam Barth.
10832
10833         I also made the should-enable expression a black-list instead
10834         of a white-list since every bot on build.webkit.org is
10835         now using NRWT except windows and WebKit2 bots.
10836
10837         * Scripts/run-webkit-tests:
10838         (useNewRunWebKitTests):
10839
10840 2011-07-05  Eric Seidel  <eric@webkit.org>
10841
10842         Perl string concat failure on Leopard (causes NRWT to fail)
10843         https://bugs.webkit.org/show_bug.cgi?id=63973
10844
10845         Reviewed by Daniel Bates.
10846
10847         Leopard's "xcodebuild -version" has a different format from Snow Leopard/Lion
10848         thus the version parsing logic in webkitdirs.pm seems to fail.
10849
10850         Default to "3.0" instead of undef when we can't parse the version number.
10851         This should send us down the correct path for Leopard.
10852
10853         * Scripts/webkitdirs.pm:
10854         (determineBaseProductDir):
10855
10856 2011-07-05  Kevin Ollivier  <kevino@theolliviers.com>
10857
10858         [wx] Unreviewed build fix after addition of Source/WebCore/xml/parser dir.
10859
10860         * waf/build/settings.py:
10861
10862 2011-07-05  Eric Seidel  <eric@webkit.org>
10863
10864         Disable NRWT for Leopard until https://bugs.webkit.org/show_bug.cgi?id=63973 can be fixed.
10865
10866         Reviewed by Adam Barth.
10867
10868         * Scripts/run-webkit-tests:
10869         (useNewRunWebKitTests):
10870
10871 2011-07-05  Eric Seidel  <eric@webkit.org>
10872
10873         Make run-webkit-tests pass --verbose to new-run-webkit-tests on qt/gtk bots.
10874
10875         Reviewed by Adam Barth.
10876
10877         * Scripts/run-webkit-tests:
10878         (runningOnBuildBot):
10879
10880 2011-07-05  Adam Barth  <abarth@webkit.org>
10881
10882         Add results_unittests.js to garden-o-matic
10883         https://bugs.webkit.org/show_bug.cgi?id=63971
10884
10885         Reviewed by Eric Seidel.
10886
10887         Adding these unit tests forced me to commit to an API for the results
10888         module.  I'm not sure if this is the API we'll end up with, but it's
10889         something not entirely unreasonable.
10890
10891         Also, I've changed results.js over to WebKit style, as discussed previously.
10892
10893         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
10894         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
10895         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js: Added.
10896         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
10897
10898 2011-07-05  Eric Seidel  <eric@webkit.org>
10899
10900         Teach run-webkit-tests how to translate --qt and --gtk for new-run-webkit-tests.
10901
10902         Reviewed by Adam Barth.
10903
10904         * Scripts/run-webkit-tests:
10905
10906 2011-07-05  Adam Barth  <abarth@webkit.org>
10907
10908         Add first unit tests for garden-o-matic
10909         https://bugs.webkit.org/show_bug.cgi?id=63969
10910
10911         Reviewed by Eric Seidel.
10912
10913         I've also reformated base.js to match WebKit style, as discussed in the
10914         previous bug.
10915
10916         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
10917         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js: Added.
10918         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html: Added.
10919
10920 2011-07-05  Eric Seidel  <eric@webkit.org>
10921
10922         Move Qt, Gtk and Leopard to NRWT.
10923
10924         We'll roll out any of these if we see problems.
10925
10926         Reviewed by Adam Barth.
10927
10928         * Scripts/run-webkit-tests:
10929         (useNewRunWebKitTests):
10930
10931 2011-07-05  Adam Barth  <abarth@webkit.org>
10932
10933         garden-o-matic should understand which tests have unexpected failures
10934         https://bugs.webkit.org/show_bug.cgi?id=63965
10935
10936         Reviewed by Eric Seidel.
10937
10938         Currently, this code just logs the list of unexpected failures to the
10939         console.  A future patch will do something useful with this
10940         information.  A testing harness will also come in the next patch.
10941
10942         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js: Added.
10943         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
10944         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
10945         * Scripts/webkitpy/tool/servers/gardeningserver.py:
10946
10947 2011-07-05  Adam Barth  <abarth@webkit.org>
10948
10949         garden-o-matic should know how to fetch test results from the (NRWT) bots
10950         https://bugs.webkit.org/show_bug.cgi?id=63959
10951
10952         Reviewed by Eric Seidel.
10953
10954         Currently, this code just fetches the results and expectations JSON
10955         blob for Linux from the server and logs it to the console.
10956
10957         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
10958         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js: Added.
10959         * Scripts/webkitpy/tool/servers/gardeningserver.py:
10960
10961 2011-07-05  Dirk Pranke  <dpranke@chromium.org>
10962
10963         Reviewed by Eric Seidel.
10964
10965         Python tests are failing on leopard
10966         https://bugs.webkit.org/show_bug.cgi?id=63842
10967
10968         Fix the missing flush() call that appears to be needed by the
10969         logging package in python 2.5. Also fix the regressions
10970         introduced in the run_webkit_test tests when we switched from
10971         thread to processes (since processes aren't available on 2.5).
10972
10973         Also fix a missing "from __future__ import with_statement' in
10974         testfilehandler (unreviewed).
10975
10976         * Scripts/webkitpy/layout_tests/layout_package/metered_stream.py:
10977         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
10978         * TestResultServer/handlers/testfilehandler.py:
10979
10980 2011-07-05  Eric Seidel  <eric@webkit.org>
10981
10982         Reviewed by Adam Barth.
10983
10984         re-enabling new-run-webkit-tests on the Snow Leopard WebKit1 bots
10985         now that we believe the stray http-server issue to be solved.
10986
10987         * Scripts/run-webkit-tests:
10988         (useNewRunWebKitTests):
10989
10990 2011-07-05  Adam Barth  <abarth@webkit.org>
10991
10992         Turns out we need to support numerals too because that's what jQuery
10993         uses by default.
10994
10995         * TestResultServer/handlers/testfilehandler.py:
10996
10997 2011-07-05  Dirk Pranke  <dpranke@chromium.org>
10998
10999         nrwt: allow for multiple http shards
11000         https://bugs.webkit.org/show_bug.cgi?id=63116
11001
11002         Reviewed by Tony Chang.
11003
11004         This modifies the sharding logic to support multiple http
11005         shards, but for now we clamp to one shard until we can test
11006         perf impact and flakiness impact.
11007
11008         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
11009         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
11010
11011 2011-07-05  Dirk Pranke  <dpranke@chromium.org>
11012
11013         Re-land nrwt: make sharding tests needing locks less hard-coded
11014         https://bugs.webkit.org/show_bug.cgi?id=63112
11015
11016         Reviewed by Ojan Vafai.
11017
11018         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
11019         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
11020
11021 2011-07-05  Adam Barth  <abarth@webkit.org>
11022
11023         TestResultServer should support callback parameter for JSON
11024         https://bugs.webkit.org/show_bug.cgi?id=63961
11025
11026         Reviewed by Ojan Vafai.
11027
11028         This makes using jQuery.ajax so much more pleasant.
11029
11030         * TestResultServer/handlers/testfilehandler.py:
11031
11032 2011-07-05  Eric Seidel  <eric@webkit.org>
11033
11034         buildbot needs to understand whether NRWT exited early after having too many failures
11035         https://bugs.webkit.org/show_bug.cgi?id=63839
11036
11037         Reviewed by Adam Barth.
11038
11039         Fix-up the exited early messages printed by NRWT so that
11040         the buildbot can parse them as expected.
11041         It looks for lines using "if line.find('Exiting early') >= 0:"
11042
11043         I also plumbed the "early exit" status through to results.json
11044         in the form of an "interrupted" bool.  It was unclear to me
11045         if results.json already had enough information to compute this bool
11046         itself.  It's possible Ojan could come up with a better fix.
11047
11048         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
11049         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
11050         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
11051         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
11052
11053 2011-07-05  Eric Seidel  <eric@webkit.org>
11054
11055         new-run-webkit-tests fails to start http server if one is already running
11056         https://bugs.webkit.org/show_bug.cgi?id=63956
11057
11058         Reviewed by Adam Barth.
11059
11060         * BuildSlaveSupport/kill-old-processes:
11061          - Removed ^M line endings and sorted the process names.
11062          - There are no functional changes to this file.
11063            (except that now that its using unix line endings it can be executed directly!)
11064         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
11065          - Change the name to 'httpd' to match old-run-webkit-test paths.
11066
11067 2011-07-05  Adam Barth  <abarth@webkit.org>
11068
11069         Add basic ajax support to garden-o-matic
11070         https://bugs.webkit.org/show_bug.cgi?id=63874
11071
11072         Reviewed by Eric Seidel.
11073
11074         This patch adds some basic infrastructure to garden-o-matic.
11075         Currently, the infrastructure is just used to make the "quit" command
11076         use Ajax, but in the future, this infrastructure will be used to do
11077         more sophistocated remote proceedure calls.
11078
11079         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
11080         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js: Added.
11081         * Scripts/webkitpy/tool/servers/gardeningserver.py:
11082         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
11083
11084 2011-07-05  Adam Barth  <abarth@webkit.org>
11085
11086         Add trivial garden-o-matic command (with server)
11087         https://bugs.webkit.org/show_bug.cgi?id=63872
11088
11089         Reviewed by Eric Seidel.
11090
11091         This patch adds a trivial "Hello, world" garden-o-matic command,
11092         complete with an HTTP server.  This command re-uses a bunch of code
11093         from the existing rebaseline-server command.  Over time, this command
11094         will grow to be a tool useful for gardening the WebKit tree.
11095
11096         * Scripts/webkitpy/tool/commands/__init__.py:
11097         * Scripts/webkitpy/tool/commands/abstractservercommand.py: Added.
11098         * Scripts/webkitpy/tool/commands/gardenomatic.py: Added.
11099         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
11100         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html: Added.
11101         * Scripts/webkitpy/tool/servers/gardeningserver.py: Added.
11102         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
11103
11104 2011-07-05  Adam Roben  <aroben@apple.com>
11105
11106         Make prepare-ChangeLog include modified Perl functions in its ChangeLog template
11107
11108         This is a very simple first cut. Functions must start with a line that starts with "sub "
11109         and end with a line that starts with a closing brace. No leading whitespace is allowed.
11110         Package names aren't parsed at all.
11111
11112         Fixes <http://webkit.org/b/21591> prepare-ChangeLog should know how to find functions in
11113         Perl files
11114
11115         Reviewed by David Kilzer.
11116
11117         * Scripts/prepare-ChangeLog: Removed redundant code that ignored certain files when
11118         searching for function line ranges. This is already done inside the get_function_line_ranges
11119         function.
11120         (get_function_line_ranges): Cleaned up coding style a little bit. Call
11121         get_function_line_ranges_for_perl for files with .pl and .pm extensions. For files with an
11122         unknown extension or no extension, read the shebang line to try to determine the script
11123         interpreter. Call get_function_line_ranges_for_perl if the interpreter seems to be Perl.
11124         (get_function_line_ranges_for_perl): Added. Does extremely basic parsing of the file to find
11125         lines starting with "sub " or "}".
11126
11127 2011-07-05  Adam Roben  <aroben@apple.com>
11128
11129         Clean up run-api-tests output on Windows
11130
11131         We were mixing run-api-tests output with gtest output, and the result was a mess.
11132
11133         Fixes <http://webkit.org/b/63954> run-api-tests output is very confusing on Windows
11134
11135         Reviewed by Dan Bates.
11136
11137         * Scripts/run-api-tests:
11138         (runTest):
11139         (populateTests):
11140         Made Windows use the formerly-Mac-only codepaths which correctly suppress gtest output
11141         except in verbose mode. Most changes are due to indentation. Use of the arch utility is now
11142         the only Mac-specific thing in these functions. Some other platform-specific code was moved
11143         from here...
11144
11145         (prepareEnvironmentForRunningTestTool):
11146         (testToolPath):
11147         ...to here.
11148
11149 2011-07-04  Gavin Barraclough  <barraclough@apple.com>
11150
11151         Reviewed by Sam Weinig.
11152
11153         https://bugs.webkit.org/show_bug.cgi?id=16652
11154         Firefox and JavaScriptCore differ in Number.toString(integer)
11155
11156         Added forwarding header.
11157
11158         * DumpRenderTree/ForwardingHeaders/wtf/StdLibExtras.h: Added.
11159
11160 2011-07-04  Anders Carlsson  <andersca@apple.com>
11161
11162         Try to fix Qt build.
11163
11164         * DumpRenderTree/TestNetscapePlugIn/Tests/NPRuntimeRemoveProperty.cpp:
11165
11166 2011-07-04  Anders Carlsson  <andersca@apple.com>
11167
11168         NP_RemoveProperty is not called back by Safari when delete npObject.prop is encountered in JavaScript
11169         https://bugs.webkit.org/show_bug.cgi?id=63915
11170         <rdar://problem/7124300>
11171
11172         Reviewed by Sam Weinig.
11173
11174         Extend the NPRuntimeRemoveProperty to handle delete object.property from JavaScript.
11175
11176         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
11177         (PluginTest::NPN_IdentifierIsString):
11178         (PluginTest::NPN_UTF8FromIdentifier):
11179         (PluginTest::NPN_IntFromIdentifier):
11180         (PluginTest::NPN_RetainObject):
11181         (PluginTest::NPN_ReleaseObject):
11182         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
11183         (PluginTest::Object::removeProperty):
11184         (PluginTest::Object::identifierIs):
11185         (PluginTest::Object::NP_RemoveProperty):
11186         (PluginTest::Object::npClass):
11187         * DumpRenderTree/TestNetscapePlugIn/Tests/NPRuntimeRemoveProperty.cpp:
11188         (NPRuntimeRemoveProperty::TestObject::TestObject):
11189         (NPRuntimeRemoveProperty::TestObject::hasProperty):
11190         (NPRuntimeRemoveProperty::TestObject::getProperty):
11191         (NPRuntimeRemoveProperty::TestObject::removeProperty):
11192         (NPRuntimeRemoveProperty::PluginObject::PluginObject):
11193         (NPRuntimeRemoveProperty::PluginObject::~PluginObject):
11194         (NPRuntimeRemoveProperty::PluginObject::hasMethod):
11195         (NPRuntimeRemoveProperty::PluginObject::invoke):
11196         (NPRuntimeRemoveProperty::PluginObject::hasProperty):
11197         (NPRuntimeRemoveProperty::PluginObject::getProperty):
11198         (NPRuntimeRemoveProperty::NPP_GetValue):
11199
11200 2011-07-04  Leandro Pereira  <leandro@profusion.mobi>
11201
11202         Reviewed by Kent Tamura.
11203
11204         [EFL] Add LayoutTestControllerEfl.
11205         https://bugs.webkit.org/show_bug.cgi?id=61974
11206
11207         Only the core features are implemented. Most functions are actually
11208         stubs and will be implemented as soon as its features are needed.
11209         Build system changes will be made as soon as DRT is completely
11210         upstreamed.
11211
11212         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp: Added.
11213         (LayoutTestController::~LayoutTestController):
11214         (LayoutTestController::addDisallowedURL):
11215         (LayoutTestController::clearBackForwardList):
11216         (LayoutTestController::copyDecodedHostName):
11217         (LayoutTestController::copyEncodedHostName):
11218         (LayoutTestController::dispatchPendingLoadRequests):
11219         (LayoutTestController::display):
11220         (LayoutTestController::counterValueForElementById):
11221         (LayoutTestController::keepWebHistory):
11222         (LayoutTestController::computedStyleIncludingVisitedInfo):
11223         (LayoutTestController::nodesFromRect):
11224         (LayoutTestController::layerTreeAsText):
11225         (LayoutTestController::pageNumberForElementById):
11226         (LayoutTestController::numberOfPages):
11227         (LayoutTestController::pageProperty):
11228         (LayoutTestController::isPageBoxVisible):
11229         (LayoutTestController::pageSizeAndMarginsInPixels):
11230         (LayoutTestController::webHistoryItemCount):
11231         (LayoutTestController::workerThreadCount):
11232         (LayoutTestController::notifyDone):
11233         (LayoutTestController::pathToLocalResource):
11234         (LayoutTestController::queueLoad):
11235         (LayoutTestController::setAcceptsEditing):
11236         (LayoutTestController::setAlwaysAcceptCookies):
11237         (LayoutTestController::setCustomPolicyDelegate):
11238         (LayoutTestController::waitForPolicyDelegate):
11239         (LayoutTestController::setScrollbarPolicy):
11240         (LayoutTestController::addOriginAccessWhitelistEntry):
11241         (LayoutTestController::removeOriginAccessWhitelistEntry):
11242         (LayoutTestController::setMainFrameIsFirstResponder):
11243         (LayoutTestController::setTabKeyCyclesThroughElements):
11244         (LayoutTestController::setUseDashboardCompatibilityMode):
11245         (LayoutTestController::setUserStyleSheetEnabled):
11246         (LayoutTestController::setUserStyleSheetLocation):
11247         (LayoutTestController::setValueForUser):
11248         (LayoutTestController::setViewModeMediaFeature):
11249         (LayoutTestController::setWindowIsKey):
11250         (LayoutTestController::setSmartInsertDeleteEnabled):
11251         (waitToDumpWatchdogFired):
11252         (LayoutTestController::setWaitToDump):
11253         (LayoutTestController::windowCount):
11254         (LayoutTestController::setPrivateBrowsingEnabled):
11255         (LayoutTestController::setJavaScriptCanAccessClipboard):
11256         (LayoutTestController::setXSSAuditorEnabled):
11257         (LayoutTestController::setFrameFlatteningEnabled):
11258         (LayoutTestController::setSpatialNavigationEnabled):
11259         (LayoutTestController::setAllowUniversalAccessFromFileURLs):
11260         (LayoutTestController::setAllowFileAccessFromFileURLs):
11261         (LayoutTestController::setAuthorAndUserStylesEnabled):
11262         (LayoutTestController::setAutofilled):
11263         (LayoutTestController::disableImageLoading):
11264         (LayoutTestController::setMockDeviceOrientation):
11265         (LayoutTestController::setMockGeolocationPosition):
11266         (LayoutTestController::setMockGeolocationError):
11267         (LayoutTestController::setGeolocationPermission):
11268         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
11269         (LayoutTestController::addMockSpeechInputResult):
11270         (LayoutTestController::setIconDatabaseEnabled):
11271         (LayoutTestController::setJavaScriptProfilingEnabled):
11272         (LayoutTestController::setSelectTrailingWhitespaceEnabled):
11273         (LayoutTestController::setPopupBlockingEnabled):
11274         (LayoutTestController::setPluginsEnabled):
11275         (LayoutTestController::elementDoesAutoCompleteForElementWithId):
11276         (LayoutTestController::execCommand):
11277         (LayoutTestController::findString):
11278         (LayoutTestController::isCommandEnabled):
11279         (LayoutTestController::setCacheModel):
11280         (LayoutTestController::setPersistentUserStyleSheetLocation):
11281         (LayoutTestController::clearPersistentUserStyleSheet):
11282         (LayoutTestController::clearAllApplicationCaches):
11283         (LayoutTestController::setApplicationCacheOriginQuota):
11284         (LayoutTestController::clearApplicationCacheForOrigin):
11285         (LayoutTestController::localStorageDiskUsageForOrigin):
11286         (LayoutTestController::originsWithApplicationCache):
11287         (LayoutTestController::applicationCacheDiskUsageForOrigin):
11288         (LayoutTestController::clearAllDatabases):
11289         (LayoutTestController::setDatabaseQuota):
11290         (LayoutTestController::originsWithLocalStorage):
11291         (LayoutTestController::deleteAllLocalStorage):
11292         (LayoutTestController::deleteLocalStorageForOrigin):
11293         (LayoutTestController::observeStorageTrackerNotifications):
11294         (LayoutTestController::syncLocalStorage):
11295         (LayoutTestController::setDomainRelaxationForbiddenForURLScheme):
11296         (LayoutTestController::goBack):
11297         (LayoutTestController::setDefersLoading):
11298         (LayoutTestController::setAppCacheMaximumSize):
11299         (LayoutTestController::pauseAnimationAtTimeOnElementWithId):
11300         (LayoutTestController::pauseTransitionAtTimeOnElementWithId):
11301         (LayoutTestController::sampleSVGAnimationForElementAtTime):
11302         (LayoutTestController::numberOfActiveAnimations):
11303         (LayoutTestController::suspendAnimations):
11304         (LayoutTestController::resumeAnimations):
11305         (LayoutTestController::overridePreference):
11306         (LayoutTestController::addUserScript):
11307         (LayoutTestController::addUserStyleSheet):
11308         (LayoutTestController::setDeveloperExtrasEnabled):
11309         (LayoutTestController::setAsynchronousSpellCheckingEnabled):
11310         (LayoutTestController::showWebInspector):
11311         (LayoutTestController::closeWebInspector):
11312         (LayoutTestController::evaluateInWebInspector):
11313         (LayoutTestController::evaluateScriptInIsolatedWorld):
11314         (LayoutTestController::removeAllVisitedLinks):
11315         (LayoutTestController::callShouldCloseOnWebView):
11316         (LayoutTestController::apiTestNewWindowDataLoadBaseURL):
11317         (LayoutTestController::apiTestGoToCurrentBackForwardItem):
11318         (LayoutTestController::setWebViewEditable):
11319         (LayoutTestController::markerTextForListItem):
11320         (LayoutTestController::authenticateSession):
11321         (LayoutTestController::setEditingBehavior):
11322         (LayoutTestController::abortModal):
11323         (LayoutTestController::hasSpellingMarker):
11324         (LayoutTestController::hasGrammarMarker):
11325         (LayoutTestController::dumpConfigurationForViewport):
11326         (LayoutTestController::setSerializeHTTPLoads):
11327         (LayoutTestController::setMinimumTimerInterval):
11328         (LayoutTestController::setTextDirection):
11329         (LayoutTestController::allowRoundingHacks):
11330
11331 2011-06-24  Vsevolod Vlasov  <vsevik@chromium.org>
11332
11333         Unreviewed, add myself as committer.
11334
11335         * Scripts/webkitpy/common/config/committers.py:
11336
11337 2011-07-03  Sheriff Bot  <webkit.review.bot@gmail.com>
11338
11339         Unreviewed, rolling out r90347.
11340         http://trac.webkit.org/changeset/90347
11341         https://bugs.webkit.org/show_bug.cgi?id=63886
11342
11343         Build breaks on Leopard, Chromium-win, WinCairo, and WinCE.
11344         (Requested by tkent on #webkit).
11345
11346         * DumpRenderTree/ForwardingHeaders/wtf/StdLibExtras.h: Removed.
11347
11348 2011-07-03  Daniel Bates  <dbates@webkit.org>
11349
11350         Import Checkout in scm_unittests.py and logging in svn.py to fix
11351         Python missing global variable errors
11352
11353         Rubber-stamped by Adam Barth.
11354
11355         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py: Import Checkout.
11356         * Scripts/webkitpy/common/checkout/scm/svn.py: Import logging.
11357
11358 2011-07-02  Gavin Barraclough  <barraclough@apple.com>
11359
11360         Reviewed by Sam Weinig.
11361
11362         https://bugs.webkit.org/show_bug.cgi?id=16652
11363         Firefox and JavaScriptCore differ in Number.toString(integer)
11364
11365         Added forwarding header.
11366
11367         * DumpRenderTree/ForwardingHeaders/wtf/StdLibExtras.h: Added.
11368
11369 2011-04-02  Robert Hogan  <robert@webkit.org>
11370
11371         Reviewed by Benjamin Poulain.
11372
11373         [Qt] Support third-party cookie policy for Qt clients
11374         https://bugs.webkit.org/show_bug.cgi?id=45455
11375
11376         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
11377         (LayoutTestController::reset):
11378         (LayoutTestController::setAlwaysAcceptCookies):
11379         (LayoutTestController::setAlwaysBlockCookies):
11380         * DumpRenderTree/qt/LayoutTestControllerQt.h:
11381
11382 2011-07-03  Adam Barth  <abarth@webkit.org>
11383
11384         Factor ReflectionHandler out of the rebaseline server.  Just code motion.
11385
11386         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
11387         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
11388         * Scripts/webkitpy/tool/servers/reflectionhandler.py: Added.
11389
11390 2011-07-03  Adam Barth  <abarth@webkit.org>
11391
11392         Move the rebaseline server out of the commands package and into the
11393         (new) servers package.  This patch prepares for refactoring this code
11394         to share server infrastructure with the (forthcoming) garden-o-matic
11395         command.
11396
11397         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
11398         * Scripts/webkitpy/tool/servers/__init__.py: Added.
11399         * Scripts/webkitpy/tool/servers/data/rebaselineserver/loupe.js:
11400         * Scripts/webkitpy/tool/servers/data/rebaselineserver/main.css:
11401         * Scripts/webkitpy/tool/servers/data/rebaselineserver/queue.js:
11402         * Scripts/webkitpy/tool/servers/rebaselineserver.py: Added.
11403
11404 2011-07-01  Adam Barth  <abarth@webkit.org>
11405
11406         Fix typo in master.cfg, hoping to trigger a master restart.
11407
11408         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
11409
11410 2011-07-01  Adam Barth  <abarth@webkit.org>
11411
11412         Turn off new-run-webkit-tests and remove some debug logging.
11413
11414         * Scripts/run-webkit-tests:
11415         * Scripts/webkitpy/layout_tests/port/http_lock.py:
11416
11417 2011-07-01  Adam Barth  <abarth@webkit.org>
11418
11419         Add some more debug logging to try to understand why we're getting
11420         stuck acquring the HTTP lock.
11421
11422         * Scripts/webkitpy/layout_tests/port/http_lock.py:
11423
11424 2011-07-01  Adam Barth  <abarth@webkit.org>
11425
11426         Refactor NRWT parsing logic in master.cfg to understand that run-webkit-tests might run NRWT
11427         https://bugs.webkit.org/show_bug.cgi?id=63854
11428
11429         Reviewed by Dirk Pranke.
11430
11431         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
11432
11433 2011-07-01  Eric Seidel  <eric@webkit.org>
11434
11435         new-run-webkit-tests results does not understand that mac uses test_expectations files
11436         https://bugs.webkit.org/show_bug.cgi?id=63838
11437
11438         Reviewed by Adam Barth.
11439
11440         We've decided to "enable" test_expectations for the Mac port for now.
11441         It makes the results.html results much more confusing to read, but at least they're
11442         no longer lying to us.
11443
11444         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
11445         * Scripts/webkitpy/layout_tests/port/base.py:
11446         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
11447
11448 2011-07-01  Adam Barth  <abarth@webkit.org>
11449
11450         Temporarily disable waiting for ReportCrash to see how this code
11451         affects performance on the bots.
11452
11453         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
11454         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
11455
11456 2011-07-01  Dmitry Lomov  <dslomov@google.com>
11457
11458         Reviewed by Adam Roben.
11459
11460         https://bugs.webkit.org/show_bug.cgi?id=63818
11461         Test bots shouldn't attempt to build TestWebKitAPI on Windows.
11462         TestWebKitAPI builds as a part of general webkit build and is a
11463         part of build products archive.
11464
11465         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
11466         * Scripts/run-api-tests:
11467
11468 2011-07-01  Adam Barth  <abarth@webkit.org>
11469
11470         We shouldn't wait for ReportCrash while the timeout is running
11471         https://bugs.webkit.org/show_bug.cgi?id=63837
11472
11473         Reviewed by Eric Seidel.
11474
11475         Instead, we should move this check to worker.py.
11476
11477         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
11478         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
11479
11480 2011-07-01  Adam Roben  <aroben@apple.com>
11481
11482         Stop setting unsafe headers on POSTs via XHR from TestFailures page
11483
11484         Fixes <http://webkit.org/b/63833> TestFailures page causes a bunch of spew in the
11485         Inspector's console about unsafe headers
11486
11487         Reviewed by Dan Bates.
11488
11489         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
11490         (fetchResource): Removed unnecessary code to set Content-Length and Connection headers.
11491
11492 2011-07-01  Eric Seidel  <eric@webkit.org>
11493
11494         Reviewed by Adam Barth.
11495
11496         Switch Snow Leopard over to new-run-webkit-tests (except for wk2)
11497         https://bugs.webkit.org/show_bug.cgi?id=63786
11498
11499         We're down to one bug blocking the "move to NRWT" bug:
11500         https://bugs.webkit.org/show_bug.cgi?id=34984
11501         so we're trying transitioning Snow Leopard to NRWT as an experiment.
11502
11503         If we run into any blocking issues we'll of course roll this back out
11504         (or anyone reading this should feel encouraged to).
11505
11506         * Scripts/run-webkit-tests:
11507
11508 2011-07-01  Adam Barth  <abarth@webkit.org>
11509
11510         Add a compat shim for --leaks.  Support for --leaks will be added in
11511         https://bugs.webkit.org/show_bug.cgi?id=63832.
11512
11513         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
11514
11515 2011-07-01  Adam Barth  <abarth@webkit.org>
11516
11517         Reviewed by Darin Adler.
11518
11519         ReportCrash destabilizes new-run-webkit-tests
11520         https://bugs.webkit.org/show_bug.cgi?id=63795
11521
11522         ReportCrash chews up too many system resources that it destablizes
11523         tests run concurrently.  This patch causes us to spin down all the
11524         worker processes when ReportCrash is running.
11525
11526         Also, this patch causes the master process to tell the user we're
11527         waiting for the crash reporter rather than having that be the worker's
11528         job.  This stops the user from getting spammed with too many "waiting
11529         for crash reporter" messages and also cleans up some minor UI issues
11530         with the meter.
11531
11532         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
11533         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
11534         * Scripts/webkitpy/layout_tests/port/base.py:
11535         * Scripts/webkitpy/layout_tests/port/server_process.py:
11536
11537 2011-07-01  Adam Roben  <aroben@apple.com>
11538
11539         Teach webkitpy about the new format of our ChangeLog template
11540
11541         r90229 moved the "Need a short description and bug URL" line to the top of the ChangeLog
11542         template. But webkitpy didn't know this, so the various rollout-related commands were
11543         leaving that line in the ChangeLog.
11544
11545         Fixes <http://webkit.org/b/63815> REGRESSION (r90229): webkit-patch rollout and sheriffbot
11546         rollout are broken
11547
11548         Reviewed by Dan Bates.
11549
11550         * Scripts/webkitpy/common/checkout/changelog.py:
11551         (ChangeLog.update_with_unreviewed_message): Changed to look for the "Need a short
11552         description and bug URL" line instead of the "Reviewed by" line as the first line of the
11553         ChangeLog template.
11554
11555         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
11556         (ChangeLogTest): Updated _new_entry_boilerplate to match what prepare-ChangeLog now
11557         generates.
11558
11559 2011-07-01  Adam Roben  <aroben@apple.com>
11560
11561         Make TestFailures stop loading old builds once we've explained all current failures
11562
11563         We used to do this, but some refactoring in r90165 broke this logic.
11564
11565         Fixes <http://webkit.org/b/63808> REGRESSION (r90165): TestFailures page always loads all
11566         builds back to the beginning of time
11567
11568         Reviewed by Dan Bates.
11569
11570         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
11571         (LayoutTestHistoryAnalyzer.prototype.start): Stop fetching old builds once we're told that
11572         all current failures have been explained.
11573
11574 2011-07-01  Adam Roben  <aroben@apple.com>
11575
11576         commit-log-editor reorders ChangeLog entries in unexpected ways
11577         https://bugs.webkit.org/show_bug.cgi?id=63804
11578
11579         Reviewed by Darin Adler.
11580
11581         commit-log-editor was changed in r46899 to move the bug title and URL to the top of the
11582         commit message in order to make git-based tools, which expect the first line of the commit
11583         message to be a summary of the change, to work better. But this reordering of the ChangeLog
11584         entry was unpredictable and unexpected.
11585
11586         I changed prepare-ChangeLog to put the bug title and URL at the top of the ChangeLog entry
11587         so that commit-log-editor doesn't have to move them. Then I changed commit-log-editor not to
11588         try to move them.
11589
11590         * Scripts/commit-log-editor: Partially reverted r46899. We no longer save the "Reviewed by"
11591         line and try to move it down below the bug title and URL. The order of the text in the
11592         ChangeLog is now preserved, except that we insert a "Patch by" line just above the "Reviewed
11593         by" line in cases where the committer and author are not the same person.
11594
11595         * Scripts/prepare-ChangeLog: Moved the "Reviewed by" line down below the bug title and URL.
11596
11597 2011-07-01  Mihnea Ovidenie  <mihnea@adobe.com>
11598
11599         Reviewed by Adam Roben.
11600
11601         [CSSRegions] Add build slave
11602         https://bugs.webkit.org/show_bug.cgi?id=62372
11603
11604         * BuildSlaveSupport/build.webkit.org-config/config.json:
11605         Add an entry for a release build of Mac.
11606
11607 2011-07-01  Sheriff Bot  <webkit.review.bot@gmail.com>
11608
11609         Unreviewed, rolling out r90213.
11610         http://trac.webkit.org/changeset/90213
11611         https://bugs.webkit.org/show_bug.cgi?id=63792
11612
11613         The SL bot has an errant HTTP server holding port 8080
11614         (Requested by abarth on #webkit).
11615
11616         * Scripts/run-webkit-tests:
11617
11618 2011-07-01  Carlos Garcia Campos  <cgarcia@igalia.com>
11619
11620         Unreviewed. Fix WebKit2 GTK build after r90163.
11621
11622         * MiniBrowser/gtk/WebBundle/WebBundleMain.c:
11623         (didCreatePage):
11624
11625 2011-07-01  Adam Barth  <abarth@webkit.org>
11626
11627         Exit faster when we get a keyboard interrupt.  Previously, we'd open
11628         Safari if you ^C NRWT after it found a failure.
11629
11630         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
11631
11632 2011-07-01  Eric Seidel  <eric@webkit.org>
11633
11634         Reviewed by Adam Barth.
11635
11636         Switch Snow Leopard over to new-run-webkit-tests (except for wk2)
11637         https://bugs.webkit.org/show_bug.cgi?id=63786
11638
11639         We're down to one bug blocking the "move to NRWT" bug:
11640         https://bugs.webkit.org/show_bug.cgi?id=34984
11641         so we're trying transitioning Snow Leopard to NRWT as an experiment.
11642
11643         If we run into any blocking issues we'll of course roll this back out
11644         (or anyone reading this should feel encouraged to).
11645
11646         * Scripts/run-webkit-tests:
11647
11648 2011-07-01  Sheriff Bot  <webkit.review.bot@gmail.com>
11649
11650         Unreviewed, rolling out r90192.
11651         http://trac.webkit.org/changeset/90192
11652         https://bugs.webkit.org/show_bug.cgi?id=63788
11653
11654         Appears to have caused NRWT on Chromium WebKit Vista to hang
11655         (Requested by abarth on #webkit).
11656
11657         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
11658         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
11659
11660 2011-06-30  Adam Barth  <abarth@webkit.org>
11661
11662         Reviewed by Eric Seidel.
11663
11664         Remove the concept of "being wedged" from new-run-webkit-tests
11665         https://bugs.webkit.org/show_bug.cgi?id=63767
11666
11667         Worker processes shouldn't ever become wedged.  My understanding is
11668         that this code was originally motivated by the old threading-based
11669         design but no longer servers any purpose.
11670
11671         Note: If we actually have a problem with the test harness getting
11672         stuck, buildbot will kill us.
11673
11674         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
11675         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
11676         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
11677
11678 2011-06-30  Adam Barth  <abarth@webkit.org>
11679
11680         The win32 implementation of fork is sad about forking test-webkitpy.
11681         Hopefully this will heal the bot.
11682
11683         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
11684
11685 2011-06-30  Eric Seidel  <eric@webkit.org>
11686
11687         Reviewed by Adam Barth.
11688
11689         Update new-run-webkit-tests now that the wk2 shared skipped list has moved
11690         https://bugs.webkit.org/show_bug.cgi?id=63780
11691
11692         The shared list was moved from mac-wk2 to wk2 as part of
11693         http://trac.webkit.org/changeset/90125.
11694
11695         * Scripts/webkitpy/layout_tests/port/webkit.py:
11696         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
11697
11698 2011-06-30  Kenichi Ishibashi  <bashi@chromium.org>
11699
11700         Reviewed by Darin Adler.
11701
11702         Style Checker should flag removal of "developmentRegion = English;" from project.pbxproj.
11703         https://bugs.webkit.org/show_bug.cgi?id=62022
11704
11705         Adds XcodeProjectFileChecker to check removal of "developmentRegion".
11706
11707         * Scripts/webkitpy/style/checker.py: Added XcodeProjectFileChecker as a checker for *.pbxproj.
11708         * Scripts/webkitpy/style/checkers/xcodeproj.py: Added.
11709         * Scripts/webkitpy/style/checkers/xcodeproj_unittest.py: Added.
11710
11711 2011-06-30  Dirk Pranke  <dpranke@chromium.org>
11712
11713         Reviewed by Ojan Vafai.
11714
11715         nrwt: make sharding tests needing locks less hard-coded
11716         https://bugs.webkit.org/show_bug.cgi?id=63112
11717
11718         This change also changes the manager logic so that it will
11719         drop the server lock as soon as all of the shards requiring
11720         the lock have completed.
11721
11722         This change includes some minor namespace/import changes in the
11723         unit tests, and also makes the Manager a new-style object, which
11724         it should've been all along.
11725
11726         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
11727         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
11728
11729 2011-06-30  Adam Barth  <abarth@webkit.org>
11730
11731         Reviewed by Eric Seidel.
11732
11733         Remove threaded mode from new-run-webkit-tests
11734         https://bugs.webkit.org/show_bug.cgi?id=63771
11735
11736         This mode is not used and is buggy.  Rather than carry around a bunch
11737         of unused buggy code, we should rip it out and focus on the
11738         multiprocess implementation.
11739
11740         * Scripts/webkitpy/common/system/executive.py:
11741         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
11742         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
11743         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
11744         * Scripts/webkitpy/layout_tests/port/base.py:
11745         * Scripts/webkitpy/layout_tests/port/chromium.py:
11746         * Scripts/webkitpy/layout_tests/port/mac.py:
11747         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
11748         * Scripts/webkitpy/layout_tests/port/server_process.py:
11749         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
11750         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
11751
11752 2011-06-30  Adam Barth  <abarth@webkit.org>
11753
11754         Reviewed by Darin Adler.
11755
11756         Clean up output from new-run-webkit-tests
11757         https://bugs.webkit.org/show_bug.cgi?id=63759
11758
11759         Printing messages from the child process looks super ugly because of
11760         the way the pretty-printer works.  Printing a blank line first is a
11761         hack, but it makes things at least partially sane.
11762
11763         Also, handle the case where calling sample throws an exception.
11764
11765         * Scripts/webkitpy/layout_tests/port/server_process.py:
11766         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
11767
11768 2011-06-27  Diego Gonzalez  <diegohcg@webkit.org>
11769
11770         Reviewed by Antonio Gomes.
11771
11772         [Qt] DRT: Provide LayoutTestController::setDefersLoading and ::goBack support
11773         https://bugs.webkit.org/show_bug.cgi?id=63489
11774
11775         Unskipped:
11776             loader/navigation-while-deferring-loads.html
11777
11778         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
11779         (LayoutTestController::goBack):
11780         (LayoutTestController::setDefersLoading):
11781         * DumpRenderTree/qt/LayoutTestControllerQt.h:
11782
11783 2011-06-30  Adam Barth  <abarth@webkit.org>
11784
11785         Reviewed by Dirk Pranke.
11786
11787         new-run-webkit-tests should stop when I ^C it
11788         https://bugs.webkit.org/show_bug.cgi?id=63763
11789
11790         It still doesn't stop fast enough, but at least with this patch it
11791         doesn't run more tests.
11792
11793         Also, I tweaked the UI that happens when there's a keyboard interrupt
11794         to not be amazingly ugly.  There's still a garbage "Te" left on the
11795         console, but that's a problem for another day.
11796
11797         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
11798
11799 2011-06-30  Adam Roben  <aroben@apple.com>
11800
11801         Make TestFailures never treat a too-many-failures build as a last-passing build for a test
11802
11803         Fixes <http://webkit.org/b/63758> TestFailures sometimes incorrectly says that a build was
11804         the last passing build for a test, when in fact the test wasn't run due to too many failures
11805         causing old-run-webkit-tests to exit early
11806
11807         Reviewed by Anders Carlsson.
11808
11809         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
11810         (LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): If there were too many
11811         failures in this build, assume that all the tests failed in the last build would have failed
11812         in this build, too, had they been run. Introduced some new local variables to reduce typing,
11813         and used them throughout.
11814
11815         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
11816         (LayoutTestResultsLoader.prototype.start): Fix a regression from r89841 where we wouldn't
11817         correctly remember whether old-run-webkit-tests exited early due to too many failures for
11818         builds older than r89610. Also removed an unused parameter while I was at it.
11819
11820 2011-06-30  Adam Roben  <aroben@apple.com>
11821
11822         Make TestFaiulres only load old-results directories as needed
11823
11824         Fixes <http://webkit.org/b/63752> Tester pages on TestFailures page load very slowly
11825
11826         Reviewed by Anders Carlsson.
11827
11828         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
11829         (Builder.prototype.getBuildNames): Moved up into the API section of the class. Now just
11830         calls through to _getBuildNamesFromResultsDirectory.
11831         (Builder.prototype.getOldBuildNames): Added. Just calls through to
11832         _getBuildNamesFromResultsDirectory.
11833         (Builder.prototype._getBuildNamesFromResultsDirectory): Renamed from getBuildNames. Now
11834         takes the directory URL as an argument and only fetches that single URL.
11835
11836         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
11837         (LayoutTestHistoryAnalyzer.prototype.start): Moved most logic to _analyzeBuilds. First
11838         analyzes builds from Builder.getBuildNames, then from Builder.getOldBuildNames if needed.
11839         (LayoutTestHistoryAnalyzer.prototype._analyzeBuilds): Moved logic here from start. (Most
11840         changes are just indentation.) Now takes a callback to call when we've finished analyzing
11841         all builds in buildNames so that we can try to fetch more build names if needed.
11842
11843 2011-06-30  Mark Rowe  <mrowe@apple.com>
11844
11845         Reviewed by Anders Carlsson.
11846
11847         Update MiniBrowser and WebKitTestRunner to use the client version constants.
11848
11849         * MiniBrowser/gtk/BrowserWindow.c:
11850         (browserWindowLoaderClientInit):
11851         (browserWindowUIClientInit):
11852         * MiniBrowser/gtk/WebBundle/WebBundleMain.c:
11853         (didCreatePage):
11854         (WKBundleInitialize):
11855         * MiniBrowser/gtk/main.c:
11856         (createWKContextWithInjectedBundle):
11857         * MiniBrowser/mac/AppDelegate.m:
11858         (-[BrowserAppDelegate init]):
11859         * MiniBrowser/mac/BrowserWindowController.m:
11860         (-[BrowserWindowController awakeFromNib]):
11861         * MiniBrowser/mac/WebBundle/WebBundleMain.m:
11862         (WKBundleInitialize):
11863         * MiniBrowser/win/BrowserView.cpp:
11864         (BrowserView::create):
11865         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
11866         (WTR::InjectedBundle::initialize):
11867         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
11868         (WTR::InjectedBundlePage::InjectedBundlePage):
11869         * WebKitTestRunner/TestController.cpp:
11870         (WTR::TestController::createOtherPage):
11871         (WTR::TestController::initialize):
11872
11873 2011-06-30  Adam Roben  <aroben@apple.com>
11874
11875         Identify revisions that modified failing tests on TestFailures page
11876
11877         Fixes <http://webkit.org/b/63716> TestFailures page should try to infer which commit caused
11878         a failure by looking at commit logs
11879
11880         Reviewed by David Kilzer.
11881
11882         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
11883         (.existing-bugs-list, .suspect-revisions-list): Made this rule apply to the suspect
11884         revisions list, too.
11885
11886         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
11887         (Trac): Initialize our memory cache.
11888         (Trac.prototype.getCommitDataForRevisionRange): Added. Fetches the commit log in RSS form,
11889         then parses out the revision number, Trac's idea of the commit title, and the files modified
11890         by the commit. (To get the modified files, we rely on the commit message including a
11891         prepare-ChangeLog-style file list.)
11892         (Trac.prototype.logURL): Added a new formatAsRSS parameter.
11893
11894         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
11895         (removePathExtension): Added. Returns a new string with the last dot and everything after it
11896         removed.
11897         (sorted): Added sortFunction parameter.
11898         (Node.prototype.removeAllChildren): Added. Does what it says.
11899
11900         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
11901         (ViewController.prototype._displayBuilder): Updated for change to _domForRegressionRange.
11902         (ViewController.prototype._domForRegressionRange): Changed argument order to match
11903         _domForNewAndExistingBugs. Now takes the list of failing tests and uses it to figure out
11904         which revisions modified the failing tests and displays the suspect revisions in a list.
11905
11906 2011-06-30  Eric Seidel  <eric@webkit.org>
11907
11908         Reviewed by Adam Barth.
11909
11910         Rename should_kill_dump_render_tree to driver_needs_restart in preparation for supporting WebKitTestRunner
11911         https://bugs.webkit.org/show_bug.cgi?id=63743
11912
11913         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
11914         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
11915
11916 2011-06-30  Martin Robinson  <mrobinson@igalia.com>
11917
11918         Reviewed by Anders Carlsson.
11919
11920         [GTK] Crash observed with nspluginwrapper and flash
11921         https://bugs.webkit.org/show_bug.cgi?id=62249
11922
11923         Added a TestNetscapePlugin test which verifies that WebKit properly
11924         handles situations where InvalidateRect is called with a null instance.
11925
11926         * DumpRenderTree/TestNetscapePlugIn/Tests/x11/CallInvalidateRectWithNullNPPArgument.cpp: Added.
11927         (CallInvalidateRectWithNullNPPArgument::CallInvalidateRectWithNullNPPArgument):
11928         (CallInvalidateRectWithNullNPPArgument::NPP_New):
11929         * GNUmakefile.am: Add the new file to sources list.
11930
11931 2011-06-30  Eric Seidel  <eric@webkit.org>
11932
11933         Reviewed by Adam Barth.
11934
11935         Remove redundant tests in TestFailuresTest
11936         https://bugs.webkit.org/show_bug.cgi?id=63742
11937
11938         These tests are all already covered by TestFailuresTest.test_loads.
11939         I also added the missing "T" from TestFailuresest. :)
11940
11941         * Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py:
11942
11943 2011-06-30  Adam Barth  <abarth@webkit.org>
11944
11945         Reviewed by Eric Seidel.
11946
11947         REGRESSION(r84294): new-run-webkit-tests results.html generate links to diffs.html or diff.png that don't exist
11948         https://bugs.webkit.org/show_bug.cgi?id=63494
11949
11950         This patch does two things:
11951
11952         1) Refactor diff_image to return the image diff instead of writing the
11953         diff to a file.  This fixes a bunch of hacks and disentangles a bunch
11954         of code.
11955
11956         2) When there's a checksum mismatch but not image diff, we no longer
11957         report an IMAGE failure to results.html.  That fixes the bug in
11958         question because results.html won't try to link to a non-existent image
11959         diff.
11960
11961         * Scripts/webkitpy/common/system/executive_mock.py:
11962         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
11963         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
11964         * Scripts/webkitpy/layout_tests/port/base.py:
11965         * Scripts/webkitpy/layout_tests/port/chromium.py:
11966         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
11967         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
11968         * Scripts/webkitpy/layout_tests/port/test.py:
11969         * Scripts/webkitpy/layout_tests/port/webkit.py:
11970         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
11971         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
11972
11973 2011-06-28  Hans Wennborg  <hans@chromium.org>
11974
11975         Reviewed by Tony Chang.
11976
11977         IndexedDB: Prepare for running layout tests with LevelDB
11978         https://bugs.webkit.org/show_bug.cgi?id=63593
11979
11980         The TestShell must always provide a temporary folder for LevelDB,
11981         not just when the backing store type is overridden.
11982
11983         * DumpRenderTree/chromium/LayoutTestController.cpp:
11984         (LayoutTestController::setOverrideIndexedDBBackingStore):
11985         (LayoutTestController::clearAllDatabases):
11986         * DumpRenderTree/chromium/LayoutTestController.h:
11987         * DumpRenderTree/chromium/TestShell.cpp:
11988         (TestShell::TestShell):
11989         * DumpRenderTree/chromium/TestShell.h:
11990
11991 2011-06-30  Zsolt Fehér  <h490267@stud.u-szeged.hu>
11992
11993         Reviewed by Csaba Osztrogonác.
11994
11995         [WK2] Create a general wk2 specific LayoutTests/platform/wk2 platform
11996         qt-wk2, win-wk2, mac-wk2 use this common Skipped list too
11997         https://bugs.webkit.org/show_bug.cgi?id=61785
11998
11999         * Scripts/old-run-webkit-tests:
12000
12001 2011-06-30  Adam Roben  <aroben@apple.com>
12002
12003         Use a colon to separate test names from failure info on TestFailures
12004
12005         This matches how the "Possibly Flaky Tests" list works, and is much prettier when the
12006         failure info itself includes parentheses.
12007
12008         Fixes <http://webkit.org/b/63722> Crashing symbols on TestFailures page look ugly due to
12009         double parentheses
12010
12011         Reviewed by Darin Adler.
12012
12013         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
12014         (ViewController.prototype._domForFailedTest): Use a colon instead of surrounding the failure
12015         info in parentheses.
12016
12017 2011-06-30  Adam Roben  <aroben@apple.com>
12018
12019         Robustify PersistentCache against cached data that contains the string ': '
12020
12021         Fixes <http://webkit.org/b/63721> Exceptions thrown while loading TestFailures page due to
12022         passing a bad string to JSON.parse
12023
12024         Reviewed by Darin Adler.
12025
12026         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/PersistentCache.js:
12027         (PersistentCache._parseDateAndJSONFromString): Manually split the string on the separator
12028         using indexOf and substring. (String.prototype.split always splits on all instances of the
12029         separator.)
12030
12031 2011-06-30  Adam Roben  <aroben@apple.com>
12032
12033         Include the crashing symbol in crash logs links on TestFailures
12034
12035         Fixes <http://webkit.org/b/63465> Links to crash logs on TestFailures page should include
12036         the crashing symbol (like links in results.html do)
12037
12038         Reviewed by David Kilzer.
12039
12040         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
12041         (LayoutTestResultsLoader.prototype.start): Bumped the cache version because we now store
12042         crashing symbols for crashing tests. Renamed testsForResultTable to parseResultTable because
12043         it now returns more than just the test names. Specifically, it now looks for crash log links
12044         and extracts the crashing symbol name from them. Updated callers of parseResultTable to
12045         match its new behavior. Changed to store the crashing symbol along with the failure type in
12046         the data we pass to the callback for tests which crashed.
12047
12048         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
12049         (code): Make <code> elements a little smaller because their contents can be quite long.
12050
12051         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
12052         (ViewController.prototype._domForFailureDiagnosis): Include the crashing symbol inside a
12053         <code> element in the link, if there is a crashing symbol.
12054
12055 2011-06-30  Adam Roben  <aroben@apple.com>
12056
12057         Use objects instead of strings to represent a test result in TestFailures code
12058
12059         This will eventually allow us to store more than just the type of failure for each test.
12060         (E.g., we can store the name of the crashing symbol for tests which crashed.)
12061
12062         Prep work for <http://webkit.org/b/63465> Links to crash logs on TestFailures page should
12063         include the crashing symbol (like links in results.html do)
12064
12065         Reviewed by David Kilzer.
12066
12067         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
12068         (Builder.prototype.failureDiagnosisTextAndURL): Changed to expect a testResult object
12069         instead of just a failureType string.
12070
12071         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
12072         (FlakyLayoutTestDetector.prototype.incorporateTestResults): Changed to store a
12073         testResult-like object for passing tests.
12074         (FlakyLayoutTestDetector.prototype.flakinessExamples): Changed to expect testResult-like
12075         objects.
12076
12077         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
12078         (LayoutTestResultsLoader.prototype.start): Store a version number along with the cached data
12079         so we can throw away cached data that's in an old format. Store a testResult object for each
12080         test instead of just its failure type.
12081
12082         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
12083         (ViewController.prototype._domForFailedTest):
12084         (ViewController.prototype._domForFailureDiagnosis):
12085         Changed to expect testResult objects instead of failureType strings.
12086
12087 2011-06-30  Adam Roben  <aroben@apple.com>
12088
12089         Show full commit logs when visiting Trac from TestFailures page
12090
12091         Fixes <http://webkit.org/b/63715> Links to Trac on TestFailures page
12092         should show full commit logs
12093
12094         Reviewed by David Kilzer.
12095
12096         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
12097         (Trac.prototype.logURL): Added new 'showFullCommitLogs' parameter. When specified, we add
12098         verbose=on to the log URL to tell Trac to show full commit logs.
12099
12100         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
12101         (ViewController.prototype._domForRegressionRange): Specify true for showFullCommitLogs.
12102
12103 2011-06-30  Adam Roben  <aroben@apple.com>
12104
12105         Add a link to file a bug about TestFailures to the TestFailures page
12106
12107         Fixes <http://webkit.org/b/63466> TestFailures page should include a link to file a bug
12108         about TestFailures itself
12109
12110         Reviewed by Anders Carlsson.
12111
12112         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
12113         (aside): Added some styles for the new link. It is positioned in the upper-right and
12114         hopefully not to glaringly obvious.
12115
12116         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
12117         (ViewController.prototype.loaded): Set up our main content element and auxiliary UI.
12118
12119         (ViewController.prototype._displayBuilder):
12120         (ViewController.prototype._displayTesters):
12121         Modify the main content element instead of the entire body, as there are other things in the
12122         body that we want to preserve.
12123
12124         (ViewController.prototype._domForAuxiliaryUIElements): Create and return an <aside> element
12125         that contains a link to file a bug about the TestFailures page.
12126
12127 2011-06-30  Adam Roben  <aroben@apple.com>
12128
12129         Ignore callbacks that correspond to old navigations of the TestFailures page
12130
12131         Fixes <http://webkit.org/b/63711> TestFailures page sometimes ignores back/forward
12132         navigations
12133
12134         Reviewed by Anders Carlsson.
12135
12136         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
12137         (LayoutTestHistoryAnalyzer.prototype.start): Check the return value of the callback. If the
12138         callback returns false, stop fetching more data.
12139
12140         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
12141         (ViewController): Initialize new _navigationID member. This member is used to assign each
12142         navigation by the user a unique ID.
12143         (ViewController.prototype.parseHash): Increment the navigation ID.
12144
12145         (ViewController.prototype._displayBuilder):
12146         (ViewController.prototype._displayTesters):
12147         At the start of these functions, record the current navigation ID. Bail out from callbacks
12148         if the navigation ID has changed, indicating that the user has navigated elsewhere.
12149
12150 2011-06-30  Carlos Garcia Campos  <cgarcia@igalia.com>
12151
12152         Reviewed by Martin Robinson.
12153
12154         [GTK] Add back/forward menu to MiniBrowser toolbar
12155         https://bugs.webkit.org/show_bug.cgi?id=63445
12156
12157         * MiniBrowser/gtk/BrowserWindow.c:
12158         (browser_window_init): Use a GtkMenuToolButton for navigation
12159         widgets.
12160         (browserWindowConstructed): Get the page BackForwardList.
12161         (browserWindowHistoryItemActivated): Go to activated history item.
12162         (browserWindowHistoryItemSelected): Show the url of currently
12163         selected history item in the statusbar.
12164         (createGtkActionFromBackForwardItem): Create a GtkAction from a
12165         WKBackForwardListItemRef.
12166         (browserWindowCreateMenuItemFromBackForwardItem): Create a
12167         GtkMenuItem from a WKBackForwardListItemRef.
12168         (browserWindowCreateBackForwardMenu): Create a GtkMenu with the
12169         given history items.
12170         (browserWindowUpdateNavigationActions): Update navigation widgets
12171         sensitivity and history menus.
12172         (didChangeBackForwardList): Call
12173         browserWindowUpdateNavigationActions().
12174         (browserWindowLoaderClientInit): Add didChangeBackForwardList
12175         callback.
12176
12177 2011-06-30  Dmitry Lomov  <dslomov@google.com>
12178
12179         Reviewed by Adam Roben.
12180
12181         https://bugs.webkit.org/show_bug.cgi?id=63660
12182         This enables unit-tests step on Windows and MacOS SnowLeopard.
12183
12184         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
12185
12186 2011-06-30  Balazs Kelemen  <kbalazs@webkit.org>
12187
12188         Reviewed by Kenneth Rohde Christiansen.
12189
12190         [Qt][WK2] Implement the rest of WTR::PlatformWebView
12191         https://bugs.webkit.org/show_bug.cgi?id=63630
12192
12193         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
12194         (WTR::PlatformWebView::windowFrame):
12195         (WTR::PlatformWebView::setWindowFrame):
12196
12197 2011-06-29  Dirk Pranke  <dpranke@chromium.org>
12198
12199         Reviewed by Adam Barth.
12200
12201         test-webkitpy is spamming lots of messages about updating test expectations
12202         https://bugs.webkit.org/show_bug.cgi?id=63680
12203
12204         This patch changes the passing_run() function to capture the
12205         output to stdout and stderr and assert that it is empty, and
12206         also specifies --no-new-test-results by default to avoid getting
12207         "generating baseline" messages for the expected-missing results.
12208
12209         Also check that we are logging messages when we generate new
12210         baselines.
12211
12212         Also fix a minor bug in the way test_tolerance() was
12213         checking for whether the tests were passing or not.
12214
12215         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
12216
12217 2011-06-29  Yuta Kitamura  <yutak@chromium.org>
12218
12219         Reviewed by Kent Tamura.
12220
12221         DRT: Make Hixie76WebSocketProtocolEnabled preference flag configurable from LayoutTestController
12222         https://bugs.webkit.org/show_bug.cgi?id=63532
12223
12224         LayoutTestController::overridePreference() will be used to change the value of
12225         the Hixie76WebSocketProtocolEnabled preference flag from layout tests. In this way,
12226         we can avoid writing a lot of boilerplate code in each port's DumpRenderTree, but
12227         at least we need to reset preference values in DumpRenderTree every time we start
12228         a new test because preference values changed in some test should not be leaked
12229         to the next test.
12230
12231         * DumpRenderTree/chromium/LayoutTestController.cpp:
12232         (LayoutTestController::overridePreference):
12233         The name of preference key is defined in WebKit/{mac/WebView,win}/WebPreferenceKeysPrivate.h.
12234         * DumpRenderTree/chromium/WebPreferences.cpp:
12235         (WebPreferences::reset):
12236         Set the default value. This will be called from TestShell::resetTestController().
12237         (WebPreferences::applyTo):
12238         * DumpRenderTree/chromium/WebPreferences.h:
12239         * DumpRenderTree/mac/DumpRenderTree.mm:
12240         (resetDefaultsToConsistentValues): Set the default value.
12241         * DumpRenderTree/win/DumpRenderTree.cpp:
12242         (resetDefaultsToConsistentValues): Ditto.
12243
12244 2011-06-29  Kenichi Ishibashi  <bashi@chromium.org>
12245
12246         Reviewed by Adam Barth.
12247
12248         check-webkit-style fails to detect https://webkit.org/b/XXXXX as a bug number
12249         https://webkit.org/b/57360
12250
12251         Modify the regular expression to match https scheme.
12252
12253         * Scripts/webkitpy/common/checkout/changelog_unittest.py: Added a test case for https://webkit.org/b/XXXXX.
12254         * Scripts/webkitpy/common/config/urls.py: Modified bug_url_short to match https scheme.
12255
12256 2011-06-29  Adam Barth  <abarth@webkit.org>
12257
12258         Reviewed by Eric Seidel.
12259
12260         new-run-webkit-tests: images page switch actual and expected images before they're loaded
12261         https://bugs.webkit.org/show_bug.cgi?id=63199
12262
12263         We now preload the images and keep the DOM nodes in memory instead of
12264         going back to the MemoryCache all the time.  (Also, remove some nutty
12265         </img> HTML.)
12266
12267         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
12268
12269 2011-06-29  Adam Barth  <abarth@webkit.org>
12270
12271         Reviewed by Eric Seidel.
12272
12273         new-run-webkit-tests should tell you when it creates new expectations
12274         https://bugs.webkit.org/show_bug.cgi?id=63675
12275
12276         Unforunately, this patch creates a bunch of test-webkitpy spam because
12277         we don't understand how Python logging works well enough to stop the
12278         spam.  :(
12279
12280         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
12281
12282 2011-06-29  Eric Seidel  <eric@webkit.org>
12283
12284         Reviewed by Adam Barth.
12285
12286         Refactor WebKitDriver.run_test into smaller pieces
12287         https://bugs.webkit.org/show_bug.cgi?id=63673
12288
12289         There is no functional change here.  Just moving code around.
12290         This is in preparation for adding support for #CRASHED and #CRASHED - WebProcess
12291
12292         * Scripts/webkitpy/layout_tests/port/webkit.py:
12293
12294 2011-06-29  Eric Seidel  <eric@webkit.org>
12295
12296         Reviewed by Adam Barth.
12297
12298         Simplify skipped file finding in preparation for adding wk2 skipped list fallback
12299         https://bugs.webkit.org/show_bug.cgi?id=63501
12300
12301         Re-landing after fixing a bug in WebKitPort.baseline_search_path which broke
12302         the rebaselineserver unit tests.  Also fixed nits brought up by Dirk after previous landing.
12303
12304         The goal was to add support for wk2 skipped lists.
12305         However, I found that our skipped list computation was a
12306         manual hack (which only worked for the mac port).
12307
12308         So I fixed a FIXME to move the skipped list finding
12309         into WebKitPort instead of MacPort.
12310         Doing so required the concept of a "port_name", since previously
12311         the only name accessible from a port object was name()
12312         which includes many things beyond "mac" or "chromium", etc.
12313
12314         Eventually I believe we'll want to re-think the way that we pass
12315         in a port_name argument to Port subclasses and expect them to parse
12316         it.  But for now I just added a cls.port_name variable which contains
12317         the static information needed to compute wk2 names as well as
12318         compute Skipped list fallback which works for Mac/Win/Qt and Gtk.
12319
12320         In order to test my new _skipped_file_search_paths method, I
12321         fixed another FIXME by making it return relative paths.
12322
12323         I also fixed the test_expectations_path code in WebKitPort to use port_name.
12324         It was using name() which would return PORT-VERSION so MacPort was overriding
12325         it to use just PORT.  After fixing test_expectations_path to use port_name
12326         (and making it aware of webkit2) I was able to remove the MacPort implementation.
12327
12328         * Scripts/webkitpy/layout_tests/port/base.py:
12329          - Add port_name() to access "mac", since name() returns "mac-leopard" etc.
12330          - Document that real_name() seems to have no purpose.
12331         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
12332          - Add _parse_port_name(), eventually we might call this from WebKitPort directly.
12333         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
12334          - Add _parse_port_name.
12335         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
12336          - Add _parse_port_name.
12337         * Scripts/webkitpy/layout_tests/port/gtk.py:
12338         * Scripts/webkitpy/layout_tests/port/mac.py:
12339          - Move Skipped-file finding code down to WebKitPort
12340         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
12341         * Scripts/webkitpy/layout_tests/port/qt.py:
12342         * Scripts/webkitpy/layout_tests/port/webkit.py:
12343         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
12344
12345 2011-06-29  Sheriff Bot  <webkit.review.bot@gmail.com>
12346
12347         Unreviewed, rolling out r90070.
12348         http://trac.webkit.org/changeset/90070
12349         https://bugs.webkit.org/show_bug.cgi?id=63676
12350
12351         Broke test-webkitpy rebaseline tests (Requested by eseidel2 on
12352         #webkit).
12353
12354         * Scripts/webkitpy/layout_tests/port/base.py:
12355         * Scripts/webkitpy/layout_tests/port/chromium.py:
12356         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
12357         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
12358         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
12359         * Scripts/webkitpy/layout_tests/port/gtk.py:
12360         * Scripts/webkitpy/layout_tests/port/mac.py:
12361         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
12362         * Scripts/webkitpy/layout_tests/port/qt.py:
12363         * Scripts/webkitpy/layout_tests/port/webkit.py:
12364         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
12365         * Scripts/webkitpy/layout_tests/port/win.py:
12366         * Scripts/webkitpy/tool/commands/rebaselineserver_unittest.py:
12367
12368 2011-06-29  Adam Barth  <abarth@webkit.org>
12369
12370         Reviewed by Eric Seidel.
12371
12372         new-run-webkit-tests doesn't support sample-on-timeout
12373         https://bugs.webkit.org/show_bug.cgi?id=56731
12374
12375         * Scripts/webkitpy/layout_tests/port/server_process.py:
12376         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
12377         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
12378
12379 2011-06-29  Eric Seidel  <eric@webkit.org>
12380
12381         Reviewed by Adam Barth.
12382
12383         Simplify skipped file finding in preparation for adding wk2 skipped list fallback
12384         https://bugs.webkit.org/show_bug.cgi?id=63501
12385
12386         The goal was to add support for wk2 skipped lists.
12387         However, I found that our skipped list computation was a
12388         manual hack (which only worked for the mac port).
12389
12390         So I fixed a FIXME to move the skipped list finding
12391         into WebKitPort instead of MacPort.
12392         Doing so required the concept of a "port_name", since previously
12393         the only name accessible from a port object was name()
12394         which includes many things beyond "mac" or "chromium", etc.
12395
12396         Eventually I believe we'll want to re-think the way that we pass
12397         in a port_name argument to Port subclasses and expect them to parse
12398         it.  But for now I just added a cls.port_name variable which contains
12399         the static information needed to compute wk2 names as well as
12400         compute Skipped list fallback which works for Mac/Win/Qt and Gtk.
12401
12402         In order to test my new _skipped_file_search_paths method, I
12403         fixed another FIXME by making it return relative paths.
12404
12405         I also fixed the test_expectations_path code in WebKitPort to use port_name.
12406         It was using name() which would return PORT-VERSION so MacPort was overriding
12407         it to use just PORT.  After fixing test_expectations_path to use port_name
12408         (and making it aware of webkit2) I was able to remove the MacPort implementation.
12409
12410         * Scripts/webkitpy/layout_tests/port/base.py:
12411          - Add port_name() to access "mac", since name() returns "mac-leopard" etc.
12412          - Document that real_name() seems to have no purpose.
12413         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
12414          - Add _parse_port_name(), eventually we might call this from WebKitPort directly.
12415         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
12416          - Add _parse_port_name.
12417         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
12418          - Add _parse_port_name.
12419         * Scripts/webkitpy/layout_tests/port/gtk.py:
12420         * Scripts/webkitpy/layout_tests/port/mac.py:
12421          - Move Skipped-file finding code down to WebKitPort
12422         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
12423         * Scripts/webkitpy/layout_tests/port/qt.py:
12424         * Scripts/webkitpy/layout_tests/port/webkit.py:
12425         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
12426
12427 2011-06-29  Adam Barth  <abarth@webkit.org>
12428
12429         Reviewed by Dirk Pranke.
12430
12431         new-run-webkit-tests complains about missing pixel results instead of plopping down new expectations
12432         https://bugs.webkit.org/show_bug.cgi?id=38063
12433
12434         This patch changes new-run-webkit-tests to match old-run-webkit-tests
12435         in generating new expected results when the expected results are
12436         missing.
12437
12438         There are still a couple details that are different:
12439
12440         1) Image baselines aren't generated unless you run with -p.
12441         2) Render tree dumps are places in the cross-platform directory instead
12442         of the platform-specific directory.
12443
12444         I'm inclined to deal with both of these issues in follow-up patches.
12445
12446         * Scripts/webkitpy/common/net/layouttestresults.py:
12447         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
12448         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
12449         * Scripts/webkitpy/layout_tests/layout_package/test_results.py:
12450         * Scripts/webkitpy/layout_tests/port/test.py:
12451         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
12452         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
12453
12454 2011-06-29  Adam Roben  <aroben@apple.com>
12455
12456         Teach TestFailures to detect possibly flaky tests and list them separately
12457
12458         Fixes <http://webkit.org/b/61061> <rdar://problem/9452796> TestFailures page blames
12459         arbitrary revisions for breaking flaky tests
12460
12461         Reviewed by Dan Bates.
12462
12463         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js: Added.
12464         (FlakyLayoutTestDetector): This class identifies flaky tests when given the test results
12465         from various builds (in reverse-chronological order).
12466         (FlakyLayoutTestDetector.prototype.incorporateTestResults): Detects flaky tests. Tests move
12467         monotonically through three states: LastSeenFailing, LastSeenPassing, and PossiblyFlaky.
12468         (FlakyLayoutTestDetector.prototype.flakinessExamples): Finds examples of flakiness for the
12469         given test. Essentially, finds all the transitions from passing to failing (or vice-versa)
12470         and puts them in an array in reverse-chronological order.
12471         (FlakyLayoutTestDetector.prototype.get possiblyFlakyTests): Returns all tests we've detected
12472         to be possibly flaky.
12473
12474         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
12475         (LayoutTestHistoryAnalyzer): Initialize new members.
12476         (LayoutTestHistoryAnalyzer.prototype.start): Now passes the callback an object with two
12477         properties: history and possiblyFlaky. history holds the data this function used to pass to
12478         the callback, while possiblyFlaky lists all tests that might be flaky and examples of their
12479         flakiness. Updated documentation comment to match.
12480         (LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): Now uses a
12481         FlakyLayoutTestDetector to identify possibly flaky tests. Any possibly flaky tests are
12482         removed from the failure history, since when they started failing is no longer meaningful.
12483         We tell our caller to keep calling until all current failures have been explained and we've
12484         gone through 5 builds without any new flaky tests being identified.
12485
12486         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
12487         (sorted): New helper function to return a sorted copy of an array.
12488         (Array.prototype.findLast): New helper function. Like findFirst, but finds the last item
12489         that matches the predicate.
12490
12491         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
12492         (ViewController.prototype._displayBuilder): Updated for change in the object passed to us by
12493         the analyzer. Now puts the list of possibly flaky tests after the failure history.
12494         (ViewController.prototype._domForFailedTest): Moved some code from here...
12495         (ViewController.prototype._domForFailureDiagnosis): ...to here.
12496         (ViewController.prototype._domForPossiblyFlakyTests): New function, builds up a list of
12497         possibly flaky tests and examples of their flakiness and returns it.
12498
12499         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
12500         FlakyLayoutTestDetector.js.
12501
12502 2011-06-29  Eric Seidel  <eric@webkit.org>
12503
12504         Adam says cowboys don't review (or unit test).
12505
12506         * Scripts/webkitpy/tool/commands/download.py:
12507
12508 2011-06-29  Eric Seidel  <eric@webkit.org>
12509
12510         Reviewed by Adam Barth.
12511
12512         Make port/config.py mockable for easier unit testing
12513         https://bugs.webkit.org/show_bug.cgi?id=63661
12514
12515         Fix an exception from the unit tests.
12516
12517         * Scripts/webkitpy/layout_tests/port/config.py:
12518
12519 2011-06-29  Brent Fulgham  <bfulgham@webkit.org>
12520
12521         Unreviewed build correction.
12522
12523         * DumpRenderTree/win/ImageDiffWinCairoCommon.vsprops:
12524         Link against zlib1.dll, not the static zlib.lib.
12525
12526 2011-06-29  Eric Seidel  <eric@webkit.org>
12527
12528         Reviewed by Adam Barth.
12529
12530         Make port/config.py mockable for easier unit testing
12531         https://bugs.webkit.org/show_bug.cgi?id=63661
12532
12533         Right now config.py leaks real system paths for
12534         webkit_base_dir due to using __file__.  This patch
12535         adds a filsystem.py wrapper around __file__ so that
12536         we don't have to jump through hoops in webkit_unittest.py
12537         to avoid leaking local paths to the unit test results.
12538
12539         There are probably many more places where we should now
12540         use filesystem.path_for_module, but starting with just this one.
12541
12542         * Scripts/webkitpy/common/system/filesystem.py:
12543         * Scripts/webkitpy/common/system/filesystem_mock.py:
12544         * Scripts/webkitpy/layout_tests/port/config.py:
12545         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
12546         * Scripts/webkitpy/tool/commands/rebaselineserver_unittest.py:
12547
12548 2011-06-29  Eric Seidel  <eric@webkit.org>
12549
12550         Reviewed by Adam Barth.
12551
12552         Remove duplicate methods in filesystem.py
12553         https://bugs.webkit.org/show_bug.cgi?id=63658
12554
12555         Looks like there was a bad merge at some point.
12556
12557         I also removed a bunch of redundant docstrings.
12558
12559         * Scripts/webkitpy/common/system/filesystem.py:
12560
12561 2011-06-29  Adam Barth  <abarth@webkit.org>
12562
12563         Reviewed by Eric Seidel.
12564
12565         Pass --no-new-test-results to NRWT on build.webkit.org
12566         https://bugs.webkit.org/show_bug.cgi?id=63653
12567
12568         This is in preparation for https://bugs.webkit.org/show_bug.cgi?id=38063.
12569
12570         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
12571
12572 2011-06-29  Xan Lopez  <xlopez@igalia.com>
12573
12574         Reviewed by Adam Barth.
12575
12576         [GTK] overlapping drag&drop tests fail on NRWT
12577         https://bugs.webkit.org/show_bug.cgi?id=57640
12578
12579         Make a GTK test driver that spawns one Xvfb instance per
12580         thread. This avoids bad interactions in DnD tests between threads.
12581
12582         * Scripts/webkitpy/layout_tests/port/gtk.py:
12583
12584 2011-06-29  Sheriff Bot  <webkit.review.bot@gmail.com>
12585
12586         Unreviewed, rolling out r89888.
12587         http://trac.webkit.org/changeset/89888
12588         https://bugs.webkit.org/show_bug.cgi?id=63609
12589
12590          "Seems to be causing merging issues on the dashboard, e.g.
12591         http://test-
12592         results.appspot.com/dashboards/flakiness_dashboard.html#group=%40ToT%20-%20chromium.org&tests=fast%2Fcanvas"
12593         (Requested by aboxhall on #webkit).
12594
12595         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
12596         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
12597         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
12598
12599 2011-06-29  Philippe Normand  <pnormand@igalia.com>
12600
12601         Reviewed by Martin Robinson.
12602
12603         [webkitpy] Executive.running_pids() sometimes doesn't correctly process "ps" output
12604         https://bugs.webkit.org/show_bug.cgi?id=63624
12605
12606         * Scripts/webkitpy/common/system/executive.py: In some cases the
12607         ps output lines can contain one or more leading white-spaces, so
12608         strip it before splitting the pid and command name.
12609
12610 2011-06-29  Noel Gordon  <noel.gordon@gmail.com>
12611
12612         Reviewed by Kent Tamura.
12613
12614         [Chromium] [DRT] ImageDiff should compile in VS2010
12615         https://bugs.webkit.org/show_bug.cgi?id=63614
12616
12617         * DumpRenderTree/chromium/ImageDiff.cpp: Include <iterator> for back_inserter.
12618
12619 2011-06-27  Diego Gonzalez  <diegohcg@webkit.org>
12620
12621         Reviewed by Kenneth Rohde Christiansen.
12622
12623         [Qt] DRT support for setInteractiveFormValidationEnabled
12624         https://bugs.webkit.org/show_bug.cgi?id=63496
12625
12626         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
12627         (WebCore::DumpRenderTree::DumpRenderTree):
12628
12629 2011-06-28  Eric Seidel  <eric@webkit.org>
12630
12631         Reviewed by Adam Barth.
12632
12633         Make new-run-webkit-tests -2 actually use WebKitTestRunner
12634         https://bugs.webkit.org/show_bug.cgi?id=63572
12635
12636         new-run-webkit-tests -2 works with this change.  However a zillion
12637         tests fail since it can't find the right skipped lists for wk2.
12638         The Skipped-list fixing is covered in bug 63501.
12639
12640         * Scripts/webkitpy/layout_tests/port/gtk.py:
12641         * Scripts/webkitpy/layout_tests/port/qt.py:
12642         * Scripts/webkitpy/layout_tests/port/webkit.py:
12643
12644 2011-06-28  Ryosuke Niwa  <rniwa@webkit.org>
12645
12646         Reviewed by Adam Roben.
12647
12648         Add more machines to Chromium Mac Release and Chromium Mac Release (Tests)
12649         https://bugs.webkit.org/show_bug.cgi?id=63500
12650
12651         Added google-mac-3 and google-mac-4.
12652
12653         * BuildSlaveSupport/build.webkit.org-config/config.json:
12654
12655 2011-06-28  Eric Seidel  <eric@webkit.org>
12656
12657         Reviewed by Dirk Pranke.
12658
12659         Remove webkitpy's tiger support now that WebKit no longer supports tiger
12660         https://bugs.webkit.org/show_bug.cgi?id=63568
12661
12662         I don't think our support actually fully worked anyway.  Regardless
12663         tiger is dead now (thankfully)!
12664
12665         I also unwrapped some short lines and added a couple FIXMEs while
12666         touring the code.
12667
12668         No new tests, only removing old ones which tested our Tiger support.
12669
12670         * Scripts/webkitpy/layout_tests/port/base.py:
12671         * Scripts/webkitpy/layout_tests/port/builders.py:
12672         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
12673         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
12674         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
12675         * Scripts/webkitpy/layout_tests/port/mac.py:
12676         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
12677         * Scripts/webkitpy/tool/commands/rebaseline.py:
12678         * Scripts/webkitpy/tool/commands/rebaselineserver_unittest.py:
12679
12680 2011-06-28  Eric Seidel  <eric@webkit.org>
12681
12682         Unreviewed, just fixing a typo.
12683
12684         Fix misspellings of "thead" to "thread".
12685
12686         * Scripts/webkitpy/common/checkout/scm/git.py:
12687         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
12688
12689 2011-06-28  Eric Seidel  <eric@webkit.org>
12690
12691         Reviewed by Tony Chang.
12692
12693         new-run-webkit-tests needs a --webkit-test-runner option
12694         https://bugs.webkit.org/show_bug.cgi?id=63439
12695
12696         This was accidentally fixed in bug 63501 as well.
12697         The root problem here was that the webkit unit tests were not using
12698         enough mocks.  I've fixed them here to use more mocks
12699         which should unbreak windows.
12700
12701         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
12702
12703 2011-06-28  Eric Seidel  <eric@webkit.org>
12704
12705         Reviewed by Tony Chang.
12706
12707         new-run-webkit-tests should not run mthml tests when MHTML support is disabled
12708         https://bugs.webkit.org/show_bug.cgi?id=63549
12709
12710         This way mhtml tests won't show up as "missing" results on all non-chromium platforms.
12711         I happen to go looking for what code was deciding that .mht was a supported
12712         extension, found test_files.py and then removed some redundant doc-strings and
12713         if branching.
12714
12715         * Scripts/webkitpy/layout_tests/port/test_files.py:
12716         * Scripts/webkitpy/layout_tests/port/webkit.py:
12717         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
12718
12719 2011-06-28  Roland Steiner  <rolandsteiner@chromium.org>
12720
12721         Reviewed by Tony Chang.
12722
12723         Fix filesystem_unittest.FileSystemTest test_read_and_write_file()
12724         https://bugs.webkit.org/show_bug.cgi?id=63514
12725
12726         - change write_text_file to write_binary_file
12727         - check that text_path and binary_path are actually existant files
12728         - change cleanup block from 'except' to 'finally'
12729
12730         * Scripts/webkitpy/common/system/filesystem_unittest.py:
12731
12732 2011-06-28  Adam Roben  <aroben@apple.com>
12733
12734         Roll out r89924 and r89925
12735
12736         I didn't mean to commit them!
12737
12738         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
12739         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
12740         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
12741
12742 2011-06-27  Adam Barth  <abarth@webkit.org>
12743
12744         Reviewed by Eric Seidel.
12745
12746         NRWT should wait for ReportCrash
12747         https://bugs.webkit.org/show_bug.cgi?id=63495
12748
12749         When a test crashes on Mac, ReportCrash confuses us into thinking the
12750         test times out.
12751
12752         I'm not sure how to test this in a meaningful way in our current test
12753         harness...
12754
12755         * Scripts/webkitpy/common/system/crashlogs.py:
12756         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
12757         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
12758         * Scripts/webkitpy/layout_tests/port/base.py:
12759         * Scripts/webkitpy/layout_tests/port/mac.py:
12760         * Scripts/webkitpy/layout_tests/port/server_process.py:
12761         * Scripts/webkitpy/tool/commands/queries.py:
12762
12763 2011-06-27  Sheriff Bot  <webkit.review.bot@gmail.com>
12764
12765         Unreviewed, rolling out r89881.
12766         http://trac.webkit.org/changeset/89881
12767         https://bugs.webkit.org/show_bug.cgi?id=63513
12768
12769         Makes Qt massively slow (Requested by abarth on #webkit).
12770
12771         * Scripts/webkitpy/common/system/crashlogs.py:
12772         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
12773         * Scripts/webkitpy/common/system/executive.py:
12774         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
12775         * Scripts/webkitpy/layout_tests/port/base.py:
12776         * Scripts/webkitpy/layout_tests/port/mac.py:
12777         * Scripts/webkitpy/layout_tests/port/server_process.py:
12778         * Scripts/webkitpy/tool/commands/queries.py:
12779
12780 2011-06-27  Alice Boxhall  <aboxhall@chromium.org>
12781
12782         Reviewed by Ojan Vafai.
12783
12784         Convert json_results_generator.py to output version 4 JSON.
12785         https://bugs.webkit.org/show_bug.cgi?id=60869
12786
12787         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
12788         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
12789         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
12790
12791 2011-06-27  Adam Barth  <abarth@webkit.org>
12792
12793         Reviewed by Eric Seidel.
12794
12795         NRWT should wait for ReportCrash
12796         https://bugs.webkit.org/show_bug.cgi?id=63495
12797
12798         When a test crashes on Mac, ReportCrash confuses us into thinking the
12799         test times out.
12800
12801         I'm not sure how to test this in a meaningful way in our current test
12802         harness...
12803
12804         * Scripts/webkitpy/common/system/crashlogs.py:
12805         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
12806         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
12807         * Scripts/webkitpy/layout_tests/port/base.py:
12808         * Scripts/webkitpy/layout_tests/port/mac.py:
12809         * Scripts/webkitpy/layout_tests/port/server_process.py:
12810         * Scripts/webkitpy/tool/commands/queries.py:
12811
12812 2011-06-27  Adam Barth  <abarth@webkit.org>
12813
12814         Reviewed by Eric Seidel.
12815
12816         new-run-webkit-tests should upload crash logs
12817         https://bugs.webkit.org/show_bug.cgi?id=55907
12818
12819         NRWT used to have a different name for crash logs.  I've changed the
12820         name to patch ORWT.
12821
12822         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
12823         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
12824
12825 2011-06-27  Eric Seidel  <eric@webkit.org>
12826
12827         Reviewed by Adam Barth.
12828
12829         new-run-webkit-tests needs a --webkit-test-runner option
12830         https://bugs.webkit.org/show_bug.cgi?id=63439
12831
12832         NRWT doesn't actually know how to run with the WebKitTestRunner yet
12833         but it does have a --webkit-test-runner option and will build WebKitTestRunner correctly.
12834
12835         There is a bunch of other little cleanup in this patch which I added as I
12836         took a tour through all of our hard-coded DumpRenderTree strings.
12837
12838         * Scripts/webkitpy/common/net/layouttestresults.py: Added FIXME.
12839         * Scripts/webkitpy/common/system/outputcapture.py: Fixed spacing to pass PEP8.
12840         * Scripts/webkitpy/layout_tests/layout_package/manager.py: Made the FIXME slightly stronger (bad bug).
12841         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py: test_shell is dead.
12842         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py: Unwrapped silly wrapping.
12843         * Scripts/webkitpy/layout_tests/port/base.py: Removed two dead methods.
12844         * Scripts/webkitpy/layout_tests/port/config.py:
12845          - build_dumprendertree had no business in this class, removed it.
12846          - Exposed _FLAGS_FROM_CONFIGURATIONS through flag_for_configuration()
12847          - Exposed _script_path as script_path() (this probably belongs elsewhere).
12848         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
12849          - Moved these tests to webkit_unittests.
12850         * Scripts/webkitpy/layout_tests/port/webkit.py:
12851          - Now respects the --webkit-test-runner option.
12852          - setup_test_run is empty in base.py too, no need to override it.
12853         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
12854          - Test the new hotness.
12855          - Bad, bad, bad!  The old code was using a real Executive during unit-testing!
12856            I think this code is still hitting disk during the unit tests. :(
12857         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
12858          - Add the option.
12859         * Scripts/webkitpy/tool/bot/queueengine_unittest.py:
12860          - Remove optional args option.
12861         * Scripts/webkitpy/tool/commands/commandtest.py:
12862          - Remove optional args option.
12863         * Scripts/webkitpy/tool/commands/queues_unittest.py:
12864          - Remove optional args option.
12865
12866 2011-06-27  Adam Barth  <abarth@webkit.org>
12867
12868         Reviewed by Dirk Pranke.
12869
12870         new-run-webkit-tests does not report stderr output
12871         https://bugs.webkit.org/show_bug.cgi?id=37739
12872
12873         Almost all the infrastructure appears to be in place to report stderr
12874         in results.html.  We just need to wire up a bool to tell results.html
12875         to create the hyperlink.
12876
12877         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
12878         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
12879         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
12880         * Scripts/webkitpy/layout_tests/layout_package/test_results.py:
12881         * Scripts/webkitpy/layout_tests/port/base.py:
12882
12883 2011-06-27  Adam Barth  <abarth@webkit.org>
12884
12885         Reviewed by Dirk Pranke.
12886
12887         webkitpy unit tests should have more descriptive names than just "Test"
12888         https://bugs.webkit.org/show_bug.cgi?id=63479
12889
12890         Luckily we kept these classes in separate namespaces in the harness so
12891         we were actually running them all even though they had the same name.
12892
12893         * Scripts/webkitpy/common/system/stack_utils_unittest.py:
12894         * Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py:
12895         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py:
12896
12897 2011-06-27  Greg Simon  <gregsimon@chromium.org>
12898
12899         No review necessary.
12900
12901         Add email addr to contributor list
12902         https://bugs.webkit.org/show_bug.cgi?id=63478
12903
12904         * Scripts/webkitpy/common/config/committers.py:
12905
12906 2011-06-27  Sheriff Bot  <webkit.review.bot@gmail.com>
12907
12908         Unreviewed, rolling out r89834.
12909         http://trac.webkit.org/changeset/89834
12910         https://bugs.webkit.org/show_bug.cgi?id=63476
12911
12912         Broke test-webkitpy (Requested by abarth on #webkit).
12913
12914         * Scripts/webkitpy/common/net/layouttestresults.py:
12915         * Scripts/webkitpy/common/system/outputcapture.py:
12916         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
12917         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
12918         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
12919         * Scripts/webkitpy/layout_tests/port/base.py:
12920         * Scripts/webkitpy/layout_tests/port/config.py:
12921         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
12922         * Scripts/webkitpy/layout_tests/port/webkit.py:
12923         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
12924         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
12925         * Scripts/webkitpy/tool/bot/queueengine_unittest.py:
12926         * Scripts/webkitpy/tool/commands/commandtest.py:
12927         * Scripts/webkitpy/tool/commands/queues_unittest.py:
12928
12929 2011-06-27  Jeff Miller  <jeffm@apple.com>
12930
12931         Reviewed by Adam Roben.
12932
12933         Stop copying obsolete WebKit libraries on Windows
12934         https://bugs.webkit.org/show_bug.cgi?id=63474
12935
12936         Don't copy obsolete WebKit library directories and files on Windows, they no longer exist and just generate extraneous warnings. These include the CharacterSets directory, dnssd.dll, and icudt40.dll.
12937
12938         * DumpRenderTree/win/DumpRenderTreePostBuild.cmd:
12939         * MiniBrowser/MiniBrowserPostBuild.cmd:
12940         * WebKitTestRunner/win/WebKitTestRunnerPostBuild.cmd:
12941
12942 2011-06-27  Ryosuke Niwa  <rniwa@webkit.org>
12943
12944         Add Wyatt Carss to the list of non-committer contributors.
12945
12946         * Scripts/webkitpy/common/config/committers.py:
12947
12948 2011-06-27  Adam Barth  <abarth@webkit.org>
12949
12950         Reviewed by Eric Seidel.
12951
12952         webkitpy should understand crash logs
12953         https://bugs.webkit.org/show_bug.cgi?id=63468
12954
12955         We're planning to use this functionality to upload crash logs along
12956         with test results for new-run-webkit-tests.
12957
12958         * Scripts/webkitpy/common/system/crashlog.py: Added.
12959         * Scripts/webkitpy/common/system/crashlog_unittest.py: Added.
12960         * Scripts/webkitpy/common/system/executive.py:
12961         * Scripts/webkitpy/common/system/executive_unittest.py:
12962         * Scripts/webkitpy/common/system/filesystem.py:
12963         * Scripts/webkitpy/common/system/filesystem_mock.py:
12964         * Scripts/webkitpy/tool/commands/queries.py:
12965
12966 2011-06-27  Adam Roben  <aroben@apple.com>
12967
12968         Make LayoutTestResultsLoader cache whether old-run-webkit-tests exited early due to too many
12969         failures
12970
12971         Fixes <http://webkit.org/b/63470> TestFailures page for a particular builder forgets
12972         old-run-webkit-tests exited early after reload
12973
12974         Reviewed by Anders Carlsson.
12975
12976         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
12977         (LayoutTestResultsLoader.prototype.start): Store both the set of failing tests and whether
12978         old-run-webkit-tests exited early due to too many failures in PersistentCache.
12979
12980 2011-06-27  Adam Roben  <aroben@apple.com>
12981
12982         Make TestFailures load build names from build.webkit.org/old-results too
12983
12984         Build results are periodically moved from results to old-results. This change makes those
12985         builds still visible to TestFailures.
12986
12987         Fixes <http://webkit.org/b/63453> TestFailures page doesn't show information for builds that
12988         have been moved to build.webkit.org/old-results
12989
12990         Reviewed by Anders Carlsson.
12991
12992         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
12993         (Builder.prototype.getBuildNames): Extracted code to parse build names from a
12994         build.webkit.org directory listing into a separate function. Instead of omitting .zip files,
12995         we now only include directory entries whose names are parseable as build names (since
12996         old-results sometimes contains other random files/directories from who knows what). We now
12997         fetch both results and old-results (with a FIXME about loading old-results on demand),
12998         extract build names from each, and concatenate the two sets of names.
12999
13000         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBuildbot.js:
13001         (WebKitBuildbot.prototype.parseBuildName): Changed to return null when the build name isn't
13002         parseable, rather than throwing an exception.
13003
13004 2011-06-26  Adam Roben  <aroben@apple.com>
13005
13006         Store analyzed history directly in LayoutTestHistoryAnalyzer instead of in a generic cache
13007
13008         It seems silly to have a whole generic cache just to store a single object.
13009
13010         Fixes <http://webkit.org/b/63407> LayoutTestHistoryAnalyzer has a generic cache but doesn't
13011         need one
13012
13013         Reviewed by Anders Carlsson.
13014
13015         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
13016         (LayoutTestHistoryAnalyzer): Replaced this._cache with this._history.
13017         (LayoutTestHistoryAnalyzer.prototype.start): Ditto.
13018         (LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): Ditto, and removed the
13019         no-longer-needed history argument.
13020
13021 2011-06-26  Adam Roben  <aroben@apple.com>
13022
13023         Move some inappropriate code out of Builder
13024
13025         Fixes <http://webkit.org/b/63406> TestFailure page's Builder class has a bunch of code that
13026         isn't really about a builder
13027
13028         Reviewed by Anders Carlsson.
13029
13030         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
13031         (Builder.prototype.getBuildNames): Renamed from _getBuildNames, since it now needs to be
13032         called by other objects.
13033
13034         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
13035         Added. Code came from Builder.
13036         (LayoutTestHistoryAnalyzer):
13037         (LayoutTestHistoryAnalyzer.prototype.start):
13038         (LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory):
13039
13040         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
13041         Added. Code came from Builder.
13042         (LayoutTestResultsLoader):
13043         (LayoutTestResultsLoader.prototype.start):
13044
13045         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
13046         (ViewController.prototype._displayBuilder): Changed to use LayoutTestHistoryAnalyzer.
13047
13048         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in new
13049         scripts.
13050
13051 2011-05-17  Nat Duca  <nduca@chromium.org>
13052
13053         Reviewed by James Robinson.
13054
13055         [chromium] Implement CCLayerTreeHost and CCLayerTreeHostImpl portions of threaded compositor
13056         https://bugs.webkit.org/show_bug.cgi?id=58408
13057
13058         * DumpRenderTree/chromium/WebViewHost.cpp:
13059         (WebViewHost::paintInvalidatedRegion):
13060
13061 2011-06-27  Eric Seidel  <eric@webkit.org>
13062
13063         Reviewed by Adam Barth.
13064
13065         new-run-webkit-tests needs a --webkit-test-runner option
13066         https://bugs.webkit.org/show_bug.cgi?id=63439
13067
13068         NRWT doesn't actually know how to run with the WebKitTestRunner yet
13069         but it does have a --webkit-test-runner option and will build WebKitTestRunner correctly.
13070
13071         There is a bunch of other little cleanup in this patch which I added as I
13072         took a tour through all of our hard-coded DumpRenderTree strings.
13073
13074         * Scripts/webkitpy/common/net/layouttestresults.py: Added FIXME.
13075         * Scripts/webkitpy/common/system/outputcapture.py: Fixed spacing to pass PEP8.
13076         * Scripts/webkitpy/layout_tests/layout_package/manager.py: Made the FIXME slightly stronger (bad bug).
13077         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py: test_shell is dead.
13078         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py: Unwrapped silly wrapping.
13079         * Scripts/webkitpy/layout_tests/port/base.py: Removed two dead methods.
13080         * Scripts/webkitpy/layout_tests/port/config.py:
13081          - build_dumprendertree had no business in this class, removed it.
13082          - Exposed _FLAGS_FROM_CONFIGURATIONS through flag_for_configuration()
13083          - Exposed _script_path as script_path() (this probably belongs elsewhere).
13084         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
13085          - Moved these tests to webkit_unittests.
13086         * Scripts/webkitpy/layout_tests/port/webkit.py:
13087          - Now respects the --webkit-test-runner option.
13088          - setup_test_run is empty in base.py too, no need to override it.
13089         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
13090          - Test the new hotness.
13091          - Bad, bad, bad!  The old code was using a real Executive during unit-testing!
13092            I think this code is still hitting disk during the unit tests. :(
13093         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
13094          - Add the option.
13095         * Scripts/webkitpy/tool/bot/queueengine_unittest.py:
13096          - Remove optional args option.
13097         * Scripts/webkitpy/tool/commands/commandtest.py:
13098          - Remove optional args option.
13099         * Scripts/webkitpy/tool/commands/queues_unittest.py:
13100          - Remove optional args option.
13101
13102 2011-06-27  Chang Shu  <cshu@webkit.org>
13103
13104         Reviewed by Darin Adler.
13105
13106         [WK2] WebKitTestRunner should reset FrameFlattening flag to disable before each test
13107         https://bugs.webkit.org/show_bug.cgi?id=63457
13108
13109         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
13110         (WTR::InjectedBundle::beginTesting):
13111
13112 2011-06-27  Raphael Kubo da Costa  <kubo@profusion.mobi>
13113
13114         Reviewed by Kent Tamura.
13115
13116         [EFL] DRT: Add an ewk_view specialization for EFL's DumpRenderTree.
13117         It will be plugged into the build system as soon as the rest of the
13118         DRT code is fully upstreamed.
13119         https://bugs.webkit.org/show_bug.cgi?id=63086
13120
13121         * DumpRenderTree/efl/DumpRenderTreeView.cpp: Added.
13122         (onConsoleMessage):
13123         (onJavaScriptAlert):
13124         (onJavaScriptConfirm):
13125         (onJavaScriptPrompt):
13126         (drtViewTiledAdd):
13127         * DumpRenderTree/efl/DumpRenderTreeView.h: Added.
13128
13129 2011-06-27  Leandro Pereira  <leandro@profusion.mobi>
13130
13131         Reviewed by Kent Tamura.
13132
13133         [EFL] Add an EventSender implementation to EFL's DumpRenderTree along
13134         with support files.
13135         Build system changes will be made as soon as the whole DumpRenderTree
13136         is upstreamed.
13137         https://bugs.webkit.org/show_bug.cgi?id=61962
13138
13139         * DumpRenderTree/efl/EventSender.cpp: Added.
13140         (setEvasModifiers):
13141         (translateMouseButtonNumber):
13142         (sendMouseEvent):
13143         (sendClick):
13144         (scheduleAsynchronousClickCallback):
13145         (updateClickCount):
13146         (modifierFromJSValue):
13147         (modifiersFromJSValue):
13148         (mouseDownCallback):
13149         (mouseUpCallback):
13150         (mouseMoveToCallback):
13151         (evasMouseEventFromHorizontalAndVerticalOffsets):
13152         (mouseScrollByCallback):
13153         (continuousMouseScrollByCallback):
13154         (keyPadNameFromJSValue):
13155         (keyNameFromJSValue):
13156         (keyDownCallback):
13157         (scalePageByCallback):
13158         (getClass):
13159         (makeEventSender):
13160         * DumpRenderTree/efl/EventSender.h: Added.
13161
13162 2011-06-27  Eric Seidel  <eric@webkit.org>
13163
13164         Reviewed by Adam Barth.
13165
13166         Remove evil uses of hasattr
13167         https://bugs.webkit.org/show_bug.cgi?id=63430
13168
13169         For some reason these classes believe that they may be called with
13170         various flavors of "option" elements and so carefully check to make
13171         sure that the options element has their option before checking it.
13172
13173         We had a set_option_default method which was never called, so I made it
13174         do what callsites seemed to want it to do and replaced 3 callers
13175         who previously used hasattr manually to use set_option_default instead.
13176
13177         * Scripts/webkitpy/layout_tests/port/base.py:
13178         * Scripts/webkitpy/layout_tests/port/webkit.py:
13179
13180 2011-06-27  Kent Tamura  <tkent@chromium.org>
13181
13182         Reviewed by Hajime Morita.
13183
13184         [Chromium] WebViewHost should be destructed before LayoutTestController, etc.
13185         https://bugs.webkit.org/show_bug.cgi?id=63427
13186
13187         ~WebViewHost() navigates to about:blank, so it calls some
13188         callbacks and LayoutTestController functions.
13189
13190         * DumpRenderTree/chromium/TestShell.h:
13191          Move m_printer upper and move m_webViewHost lower.
13192
13193 2011-06-27  Adam Barth  <abarth@webkit.org>
13194
13195         Reviewed by Eric Seidel.
13196
13197         Move server config files to webkitpy.layout_tests.servers
13198         https://bugs.webkit.org/show_bug.cgi?id=63436
13199
13200         The config files belong with the servers.
13201
13202         * Scripts/webkitpy/layout_tests/port/httpd2.pem: Removed.
13203         * Scripts/webkitpy/layout_tests/port/lighttpd.conf: Removed.
13204         * Scripts/webkitpy/layout_tests/servers/http_server.py:
13205         * Scripts/webkitpy/layout_tests/servers/httpd2.pem: Copied from Tools/Scripts/webkitpy/layout_tests/port/httpd2.pem.
13206         * Scripts/webkitpy/layout_tests/servers/lighttpd.conf: Copied from Tools/Scripts/webkitpy/layout_tests/port/lighttpd.conf.
13207
13208 2011-06-27  Adam Barth  <abarth@webkit.org>
13209
13210         Reviewed by Eric Seidel.
13211
13212         Move websocket_server and http_server to webkitpy.layout_package.servers
13213         https://bugs.webkit.org/show_bug.cgi?id=63434
13214
13215         These classes do not represent ports.  They represent servers.
13216
13217         * Scripts/new-run-webkit-httpd:
13218         * Scripts/new-run-webkit-websocketserver:
13219         * Scripts/webkitpy/layout_tests/port/base.py:
13220         * Scripts/webkitpy/layout_tests/port/chromium.py:
13221         * Scripts/webkitpy/layout_tests/port/http_server.py: Removed.
13222         * Scripts/webkitpy/layout_tests/port/http_server_base.py: Removed.
13223         * Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Removed.
13224         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
13225         * Scripts/webkitpy/layout_tests/port/websocket_server.py: Removed.
13226         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
13227         * Scripts/webkitpy/layout_tests/servers/http_server.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/http_server.py.
13228         * Scripts/webkitpy/layout_tests/servers/http_server_base.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/http_server_base.py.
13229         * Scripts/webkitpy/layout_tests/servers/http_server_integrationtest.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py.
13230         * Scripts/webkitpy/layout_tests/servers/websocket_server.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/websocket_server.py.
13231
13232 2011-06-27  Eric Seidel  <eric@webkit.org>
13233
13234         Reviewed by Adam Barth.
13235
13236         Remove Chromium and ChromiumMac implementations of driver_name now that everything uses DumpRenderTree
13237         https://bugs.webkit.org/show_bug.cgi?id=63428
13238
13239         * Scripts/webkitpy/layout_tests/port/chromium.py:
13240         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
13241
13242 2011-06-27  Adam Barth  <abarth@webkit.org>
13243
13244         Reviewed by Eric Seidel.
13245
13246         Move apache_http_server.py to webkitpy.layout_package.servers
13247         https://bugs.webkit.org/show_bug.cgi?id=63433
13248
13249         The apache_http_server.py is a server, not a port.
13250
13251         * Scripts/webkitpy/layout_tests/port/apache_http_server.py: Removed.
13252         * Scripts/webkitpy/layout_tests/port/base.py:
13253         * Scripts/webkitpy/layout_tests/servers: Added.
13254         * Scripts/webkitpy/layout_tests/servers/__init__.py: Added.
13255         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/apache_http_server.py.
13256
13257 2011-06-27  Adam Barth  <abarth@webkit.org>
13258
13259         Reviewed by Eric Seidel.
13260
13261         Move rebaseline-chromium-webkit-tests.py to webkitpy.to_be_moved
13262         https://bugs.webkit.org/show_bug.cgi?id=63432
13263
13264         This code isn't in the dependency cone of new-run-webkit-tests. It's
13265         unclear what it's final home should be, but this patch moves it out of
13266         layout_tests to simply the package.
13267
13268         * Scripts/rebaseline-chromium-webkit-tests:
13269         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py: Removed.
13270         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py: Removed.
13271         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py: Copied from Tools/Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py.
13272         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py.
13273
13274 2011-06-27  Adam Barth  <abarth@webkit.org>
13275
13276         Reviewed by Eric Seidel.
13277
13278         Move update_webgl_conformance_tests.py to webkitpy.to_be_moved
13279         https://bugs.webkit.org/show_bug.cgi?id=63429
13280
13281         This package as nothing to do with the layout_tests package.
13282
13283         * Scripts/update-webgl-conformance-tests:
13284         * Scripts/webkitpy/layout_tests/update_webgl_conformance_tests.py: Removed.
13285         * Scripts/webkitpy/layout_tests/update_webgl_conformance_tests_unittest.py: Removed.
13286         * Scripts/webkitpy/to_be_moved/update_webgl_conformance_tests.py: Copied from Tools/Scripts/webkitpy/layout_tests/update_webgl_conformance_tests.py.
13287         * Scripts/webkitpy/to_be_moved/update_webgl_conformance_tests_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/update_webgl_conformance_tests_unittest.py.
13288
13289 2011-06-27  Adam Barth  <abarth@webkit.org>
13290
13291         Reviewed by Eric Seidel.
13292
13293         Move deduplicate_tests.py into webkitpy.to_be_moved
13294         https://bugs.webkit.org/show_bug.cgi?id=63426
13295
13296         deduplicate_tests.py doesn't haven anything to do with layout tests.
13297         It's really a separate script.  This patch moves it to the (new)
13298         to_be_moved package so that it doesn't distract us when working on NRWT.
13299
13300         * Scripts/deduplicate-tests:
13301         * Scripts/webkitpy/layout_tests/deduplicate_tests.py: Removed.
13302         * Scripts/webkitpy/layout_tests/deduplicate_tests_unittest.py: Removed.
13303         * Scripts/webkitpy/test/main.py:
13304         * Scripts/webkitpy/to_be_moved: Added.
13305         * Scripts/webkitpy/to_be_moved/__init__.py: Added.
13306         * Scripts/webkitpy/to_be_moved/deduplicate_tests.py: Copied from Tools/Scripts/webkitpy/layout_tests/deduplicate_tests.py.
13307         * Scripts/webkitpy/to_be_moved/deduplicate_tests_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/deduplicate_tests_unittest.py.
13308
13309 2011-06-26  Adam Barth  <abarth@webkit.org>
13310
13311         Rubber-stamped by Eric Seidel.
13312
13313         Remove empty directory.
13314
13315         * Scripts/webkitpy/layout_tests/test_types: Removed.
13316
13317 2011-06-25  Naoki Takano  <honten@chromium.org>
13318
13319         Reviewed by Eric Seidel.
13320
13321         Change my entry from contributors to committers.
13322         https://bugs.webkit.org/show_bug.cgi?id=63043
13323
13324         * Scripts/webkitpy/common/config/committers.py: Change my e-mail from takano.naoki@gmail.com to honten@chromium.org. And move the entry to commiters from contributors.
13325
13326 2011-06-25  Adam Barth  <abarth@webkit.org>
13327
13328         Reviewed by Eric Seidel.
13329
13330         [Chromium] Update WebKit for WEBKIT_FRAME_TO_DOCUMENT_API_MOVE
13331         https://bugs.webkit.org/show_bug.cgi?id=63394
13332
13333         * DumpRenderTree/chromium/LayoutTestController.cpp:
13334         (LayoutTestController::queueLoad):
13335
13336 2011-06-25  Brent Fulgham  <bfulgham@webkit.org>
13337
13338         Unreviewed build correction after r89426.
13339
13340         * WebKitTestRunner/Configurations/WebKitTestRunnerCFLite.vsprops:
13341           Link with Cairo DLL for new functionality.
13342         * WebKitTestRunner/cairo/TestInvocationCairo.cpp: Include compatibility
13343           code for snprintf.
13344
13345 2011-06-25  Dimitri Glazkov  <dglazkov@chromium.org>
13346
13347         Reviewed by Adam Barth.
13348
13349         run-bindings-tests does not work on Chromium Windows.
13350         https://bugs.webkit.org/show_bug.cgi?id=63382
13351
13352         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Removed run-bindings-tests as a build step for cr-win.
13353
13354 2011-06-25  Dan Bernstein  <mitz@apple.com>
13355
13356         Try to fix the Qt and Chromium builds.
13357
13358         * DumpRenderTree/chromium/LayoutTestController.cpp:
13359         (LayoutTestController::setEncodedAudioData):
13360         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
13361
13362 2011-06-24  Dan Bernstein  <mitz@apple.com>
13363
13364         Reviewed by Anders Carlsson.
13365
13366         Add an option to enable legacy rounding hacks
13367         https://bugs.webkit.org/show_bug.cgi?id=63363
13368
13369         Added layoutTestController.allowRoundingHacks() and implemented it
13370         on OS X.
13371
13372         * DumpRenderTree/LayoutTestController.cpp:
13373         (allowRoundingHacksCallback):
13374         (LayoutTestController::staticFunctions):
13375         * DumpRenderTree/LayoutTestController.h:
13376         * DumpRenderTree/chromium/LayoutTestController.cpp:
13377         (LayoutTestController::allowRoundingHacks):
13378         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
13379         (LayoutTestController::allowRoundingHacks):
13380         * DumpRenderTree/mac/DumpRenderTree.mm:
13381         (resetWebViewToConsistentStateBeforeTesting):
13382         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
13383         (LayoutTestController::allowRoundingHacks):
13384         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
13385         (LayoutTestController::allowRoundingHacks):
13386         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
13387         (LayoutTestController::allowRoundingHacks):
13388         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
13389         (LayoutTestController::allowRoundingHacks):
13390
13391 2011-06-24  Adam Roben  <aroben@apple.com>
13392
13393         Include the directory containing the failing tests in titles of bugs filed from TestFailures
13394         when all the tests won't fit
13395
13396         Fixes <http://webkit.org/b/63350> Short bug titles from TestFailures page give no indication
13397         which tests are failing
13398
13399         Reviewed by David Kilzer.
13400
13401         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
13402         (longestCommonPathPrefix): Added. Given a set of paths, returns the longest common prefix
13403         that ends in a path separator.
13404
13405         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
13406         (ViewController.prototype._domForNewAndExistingBugs): When we can't fit all the test names
13407         in the title, first try to include the longest common prefix of the test names, then, if
13408         that's still too long or there's no common prefix, fall back to not including any
13409         information about which tests are failing.
13410
13411 2011-06-24  Adam Roben  <aroben@apple.com>
13412
13413         Include links to Trac in bugs filed from TestFailures
13414
13415         Fixes <http://webkit.org/b/63348> Bugs filed from TestFailures page should include links to
13416         Trac for suspect revisions
13417
13418         Reviewed by David Kilzer.
13419
13420         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
13421         (Trac.prototype.changesetURL): New function, returns the URL for a particular revision in
13422         Trac.
13423
13424         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
13425         (ViewController.prototype._domForNewAndExistingBugs): If there's only one suspect revision,
13426         include a link to Trac for it in the first sentence of the description. Otherwise include a
13427         link to the suspect range after the list of failing tests.
13428
13429 2011-06-24  Nat Duca  <nduca@chromium.org>
13430
13431         Unreviewed, rolling out r89694.
13432         http://trac.webkit.org/changeset/89694
13433         https://bugs.webkit.org/show_bug.cgi?id=58408
13434
13435         Test shell still not ready for animate changes.
13436
13437         * DumpRenderTree/chromium/WebViewHost.cpp:
13438         (WebViewHost::paintInvalidatedRegion):
13439
13440 2011-06-24  Adam Roben  <aroben@apple.com>
13441
13442         Fix typos introduced in r89696
13443
13444         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
13445         (Builder.prototype.getNumberOfFailingTests): Pass the result data's constituent parts to the
13446         callback, since that's what it expects.
13447
13448 2011-06-24  Adam Roben  <aroben@apple.com>
13449
13450         Make TestFailures correctly remember whether old-run-webkit-tests exited early
13451
13452         Fixes <http://webkit.org/b/63342> TestFailures page incorrectly claims test run ran to
13453         completion after reload
13454
13455         Reviewed by David Kilzer.
13456
13457         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
13458         (Builder.prototype.getNumberOfFailingTests): Instead of just storing the number of failures
13459         in the PeristentCache, store an object that contains both the number of failures and whether
13460         old-run-webkit-tests exited early.
13461
13462 2011-06-24  Adam Roben  <aroben@apple.com>
13463
13464         Add links to regression ranges in Trac to the TestFailures page
13465
13466         Fixes <http://webkit.org/b/61060> <rdar://problem/9452153> TestFailures page should provide
13467         links to regression ranges in Trac
13468
13469         Reviewed by David Kilzer.
13470
13471         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js: Added.
13472         (Trac): This new class represents a particular instance of Trac for a single project.
13473         (Trac.prototype.logURL): Returns the URL that can be used to see the log of the given
13474         repository path between the two specified revisions
13475         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
13476         (ViewController): Added new trac argument.
13477         (ViewController.prototype._displayBuilder): Moved code to create the DOM for the
13478         passed/failed builds from here...
13479         (ViewController.prototype._domForRegressionRange): ...to here. Now also includes a link to
13480         Trac if there are multiple suspect revisions.
13481
13482         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
13483         Trac.js and pass a Trac instance for trac.webkit.org to the ViewController.
13484
13485 2011-05-17  Nat Duca  <nduca@chromium.org>
13486
13487         Reviewed by James Robinson.
13488
13489         [chromium] Implement CCLayerTreeHost and CCLayerTreeHostImpl portions of threaded compositor
13490         https://bugs.webkit.org/show_bug.cgi?id=58408
13491
13492         * DumpRenderTree/chromium/WebViewHost.cpp:
13493         (WebViewHost::paintInvalidatedRegion):
13494
13495 2011-06-24  Dominic Cooney   <dominicc@chromium.org>
13496
13497         Reviewed by Dimitri Glazkov.
13498
13499         Convert shadow DOM-related tests to use window.internals
13500         https://bugs.webkit.org/show_bug.cgi?id=61671
13501
13502         Remove ensureShadowRoot, shadowRoot, shadowPseudoId and
13503         removeShadowRoot from layoutTestController; these have been
13504         replaced by equivalents in window.internals instead.
13505
13506         * DumpRenderTree/LayoutTestController.cpp:
13507         (LayoutTestController::staticFunctions):
13508         * DumpRenderTree/LayoutTestController.h:
13509         * DumpRenderTree/chromium/LayoutTestController.cpp:
13510         (LayoutTestController::LayoutTestController):
13511         * DumpRenderTree/chromium/LayoutTestController.h:
13512         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
13513         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
13514         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
13515         * DumpRenderTree/qt/LayoutTestControllerQt.h:
13516         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
13517         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
13518         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
13519         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
13520         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
13521
13522 2011-06-24  Csaba Osztrogonác  <ossy@webkit.org>
13523
13524         Reviewed by Dirk Schulze.
13525
13526         Make run-bindings-tests remove its temporary directories.
13527         https://bugs.webkit.org/show_bug.cgi?id=63320
13528
13529         * Scripts/run-bindings-tests:
13530
13531 2011-06-23  MORITA Hajime  <morrita@google.com>
13532
13533         Reviewed by Kent Tamura.
13534
13535         [Chromium] Invalid write inside WebKit::FrameLoaderClientImpl::dispatchDidClearWindowObjectInWorld
13536         https://bugs.webkit.org/show_bug.cgi?id=63216
13537
13538         Deleted m_webViewHost which looks leaked.
13539
13540         * DumpRenderTree/chromium/TestShell.h:
13541         * DumpRenderTree/chromium/TestShell.cpp:
13542         (TestShell::~TestShell):
13543
13544 2011-06-24  Carlos Garcia Campos  <cgarcia@igalia.com>
13545
13546         Unreviewed. Fix WebKit2 GTK build.
13547
13548         * WebKitTestRunner/GNUmakefile.am:
13549
13550 2011-06-23  Ryuan Choi  <ryuan.choi@samsung.com>
13551
13552         Unreviewed, add myself as committer.
13553
13554         * Scripts/webkitpy/common/config/committers.py:
13555
13556 2011-06-23  Nat Duca  <nduca@chromium.org>
13557
13558         Unreviewed, add myself as committer.
13559
13560         * Scripts/webkitpy/common/config/committers.py:
13561
13562 2011-06-23  Dirk Pranke  <dpranke@chromium.org>
13563
13564         Reviewed by Tony Chang.
13565
13566         nrwt: move http locking code into manager
13567         https://bugs.webkit.org/show_bug.cgi?id=63103
13568
13569         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
13570         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
13571
13572 2011-06-23  Alok Priyadarshi  <alokp@chromium.org>
13573
13574         Reviewed by James Robinson.
13575
13576         [chromium] Add support for running layout tests with accelerated rendering
13577         https://bugs.webkit.org/show_bug.cgi?id=63274
13578
13579         * Scripts/webkitpy/layout_tests/port/chromium.py:
13580         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
13581
13582 2011-06-23  Adam Roben  <aroben@apple.com>
13583
13584         Avoid fetching JSON data when possible on TestFailures page to determine if
13585         old-run-webkit-tests exited early
13586
13587         As of r89610, results.html includes information about whether old-run-webkit-tests exited
13588         early, so we don't need to look at JSON data to get this information for builds newer than
13589         that.
13590
13591         Fixes <http://webkit.org/b/63281> TestFailures page fetches build JSON for every build,
13592         which is slow and limits history to the last build master restart
13593
13594         Reviewed by David Kilzer.
13595
13596         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
13597         (Builder.prototype._getFailingTests): Moved code dealing with results.html into a new
13598         function. If we're dealing with a build that has early-exit information in results.html,
13599         don't bother fetching JSON data and go straight to parsing results.html. Otherwise first use
13600         the JSON data to determine if old-run-webkit-tests exited early, and then parse results.html
13601         (as before).
13602
13603 2011-06-23  Eric Seidel  <eric@webkit.org>
13604
13605         Reviewed by Adam Barth.
13606
13607         sheriffbot can no longer restart
13608         https://bugs.webkit.org/show_bug.cgi?id=63221
13609
13610         Fixed and unit tested.
13611
13612         * Scripts/webkitpy/tool/bot/queueengine.py:
13613         * Scripts/webkitpy/tool/bot/sheriffircbot.py:
13614         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
13615
13616 2011-06-23  Martin Robinson  <mrobinson@igalia.com>
13617
13618         Reviewed by Xan Lopez.
13619
13620         [GTK] [WebKit2] WebKitTestRunner does not properly load TestNetscapePlugin
13621         https://bugs.webkit.org/show_bug.cgi?id=63287
13622
13623         Send the path to the TestNetscapePlugin to the WKTR via an environment variable
13624         so that the plugin loads properly.
13625
13626         * Scripts/old-run-webkit-tests: Pass the appropriate environment variable to WKTR.
13627         * WebKitTestRunner/gtk/TestControllerGtk.cpp:
13628         (WTR::getEnvironmentVariableAsUTF8String): Added this helper which abstracts away some
13629         of the work for getting an environment variable and converting it to UTF-8.
13630         (WTR::TestController::initializeInjectedBundlePath): Use the helper.
13631         (WTR::TestController::initializeTestPluginDirectory): Get the path from an environment variable now.
13632
13633 2011-06-23  Adam Roben  <aroben@apple.com>
13634
13635         Record whether old-run-webkit-tests aborted early in results.html
13636
13637         Fixes <http://webkit.org/b/63280> results.html gives no indication of whether
13638         old-run-webkit-tests stopped running tests early
13639
13640         Reviewed by David Kilzer.
13641
13642         * Scripts/old-run-webkit-tests:
13643         (top level): Switched to heredoc syntax for printing the start of results.html for clarity.
13644         Added a <style> element with some styles for the new message. If we stopped running early,
13645         put the reason why at the top of results.html.
13646         (stopRunningTestsEarlyIfNeeded): Store the message we print in the new
13647         $stoppedRunningEarlyMessage global.
13648
13649 2011-06-23  Adam Roben  <aroben@apple.com>
13650
13651         Cache some of the TestFailures page's data in localStorage
13652
13653         This makes reloading TestFailures much faster. Right now we only store the number of failing
13654         tests and the list of failing tests for each build. We may choose to store more later, but
13655         it's easy to run up against quota limits.
13656
13657         Fixes <http://webkit.org/b/61520> TestFailures page should take advantage of LocalStorage
13658         APIs (or similar) to improve loading performance
13659
13660         Reviewed by David Kilzer.
13661
13662         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
13663         (Builder.prototype.getNumberOfFailingTests): Changed to use the new PersistentCache object.
13664         (Builder.prototype.startFetchingBuildHistory): Changed to pass whether or not we're still
13665         fetching data to the callback.
13666         (Builder.prototype._getFailingTests): Changed to use the new PersistentCache object. We now
13667         store the tests in the cache just before calling the callback. (The previous code in this
13668         function relied on being able to modify the tests object after storing it in the cache and
13669         having the cached version be updated. This worked while it was a non-serialized cache, but
13670         PersistentCache uses serialization.)
13671
13672         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/PersistentCache.js: Added.
13673         (PersistentCache): This object wraps localStorage. It uses JSON to serialize/deserialize
13674         values, and stores the date that each value was initially stored along with it. This is
13675         later used for pruning the cache.
13676         (PersistentCache.contains): Checks whether the key exists in localStorage.
13677         (PersistentCache.get): Fetch the string we stored in localStorage and extract the original
13678         value out of it.
13679         (PersistentCache.set): Serialize the value, add the date to it, and store it in
13680         localStorage. If this fails due to quota limits, empty the whole cache and try again.
13681         (PersistentCache.prune): Delete any cached data that is deemed old enough.
13682         (PersistentCache._addDateToJSONString): Prepend the current date to the string.
13683         (PersistentCache._emptyCache): Delete everything from localStorage.
13684         (PersistentCache._parseDateAndJSONFromString): Split apart the date and the JSON string and
13685         return them.
13686
13687         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
13688         (ViewController.prototype._displayBuilder): Updated for change to callback signature. When
13689         we finish fetching data, prune the PersistentCache. While I was here I also fixed a bug
13690         where we'd never show the new bug link for tests for which we couldn't determine a passing
13691         revision.
13692
13693         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
13694         PersistentCache.js.
13695
13696 2011-06-23  Adam Roben  <aroben@apple.com>
13697
13698         Make finding existing bugs and filing new bugs work on TestFailures even when lots of tests are failing
13699
13700         One bug this fixes is <http://webkit.org/b/61660> New bug links on TestFailures page often
13701         contain titles that are so long they are rejected by Bugzilla
13702
13703         Reviewed by David Kilzer.
13704
13705         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js:
13706         (Bugzilla.prototype.quickSearch): Use fetchResource to POST the search query rather than
13707         using a query string on the URL. If the search query is very long, using a query string can
13708         cause the request to be rejected due to the URL being too long. POSTing avoids this issue.
13709         (Bugzilla.maximumBugTitleLength): Added this constant based on Bugzilla's implementation.
13710
13711         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
13712         (.new-bug-form): Hide the form that we secretly use to file a new bug.
13713
13714         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
13715         (fetchResource): Added. Code came from getResource. If we're using a GET request, add the
13716         query parameters to the URL. Otherwise, send them as the body of the request along with the
13717         appropriate headers.
13718         (getResource): Now just calls through to fetchResource.
13719         (urlEncodedQueryParameters): Added. Moved some code here...
13720         (addQueryParametersToURL): ...from here.
13721
13722         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
13723         (ViewController.prototype._domForNewAndExistingBugs): Shorten the title to just mention the
13724         number of failing tests if mentioning all the tests would make the title too long. Improved
13725         the description for large numbers of failures by listing the tests one-per-line instead of
13726         just having them be comma-separated, which was hard to read. Use a form to file the new bug
13727         instead of a link so that we can POST the form data. (Using a URL with a query string can
13728         result in the URL being too long.) The new bug link now just submits the form.
13729
13730 2011-06-23  Adam Roben  <aroben@apple.com>
13731
13732         Show closed bugs on the TestFailures page in addition to open ones
13733
13734         Fixes <http://webkit.org/b/63194> TestFailures page should show closed bugs, too
13735
13736         Reviewed by David Kilzer.
13737
13738         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js:
13739         (Bugzilla.prototype.quickSearch): Added code to extract the bug's status and store it in the
13740         returned data.
13741         (Bugzilla.isOpenStatus): New function, returns true if the passed-in status indicates that
13742         the associated bug is still open.
13743
13744         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
13745         (addQueryParametersToURL): Add a missing semicolon.
13746         (Node.prototype.appendChildren): New function, like appendChild but takes an array-like
13747         object and appends each of the values stored within.
13748
13749         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
13750         (ViewController.prototype._domForNewAndExistingBugs): Prepend 'ALL' to the query so closed
13751         bugs will be included in the results. Split the returned bugs into two sets: those which are
13752         open, and those which are closed. Put the open bugs at the top level of the list, and the
13753         closed bugs in a second level.
13754
13755 2011-06-23  Adam Roben  <aroben@apple.com>
13756
13757         Don't count new tests as failures on the TestFailures page
13758
13759         Fixes <http://webkit.org/b/63254> TestFailures page calls new tests "failures", even though
13760         they aren't
13761
13762         Reviewed by Anders Carlsson.
13763
13764         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
13765         (Builder.prototype.getNumberOfFailingTests): Don't add the number of new tests to the number
13766         of failures.
13767
13768 2011-06-23  Dirk Pranke  <dpranke@chromium.org>
13769
13770         Reviewed by Eric Seidel.
13771
13772         nrwt: make TestInput objects printable
13773         https://bugs.webkit.org/show_bug.cgi?id=63225
13774
13775         This makes debugging slighlty easier.
13776
13777         * Scripts/webkitpy/layout_tests/layout_package/test_input.py:
13778
13779 2011-06-22  Daniel Bates  <dbates@webkit.org>
13780
13781         Reviewed by Adam Barth.
13782
13783         test-webkitpy --all errors out because scm_unittest.py can't find module checkout
13784         https://bugs.webkit.org/show_bug.cgi?id=62943
13785
13786         Remove "from .checkout import Checkout" from scm_unittest.py since it's included
13787         by Scripts/webkitpy/common/checkout/__init__.py.
13788
13789         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
13790
13791 2011-06-22  Dominic Cooney  <dominicc@chromium.org>
13792
13793         Reviewed by Mark Rowe.
13794
13795         Add window.internals to WebKit2's WebKitTestRunner.
13796         https://bugs.webkit.org/show_bug.cgi?id=61073
13797
13798         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pxbproj:
13799         * WebKitTestRunner/Configurations/Base.xcconfig:
13800         * WebKitTestRunner/Configurations/InjectedBundleCFLite.vsprops:
13801         * WebKitTestRunner/Configurations/InjectedBundleCommon.vsprops:
13802         * WebKitTestRunner/Configurations/InjectedBundleCoreFoundation.vsprops:
13803         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
13804         (WTR::InjectedBundlePage::didClearWindowForFrame):
13805         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
13806         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
13807
13808 2011-06-22  Eric Seidel  <eric@webkit.org>
13809
13810         Reviewed by Adam Barth.
13811
13812         Make sheriff-bot rollout messages a little nicer
13813         https://bugs.webkit.org/show_bug.cgi?id=63107
13814
13815         itertools.chain.from_iterable is new in 2.6,
13816         use itertools.chain(*list) for 2.5 compat.
13817
13818         * Scripts/webkitpy/tool/bot/irc_command.py:
13819
13820 2011-06-22  Eric Seidel  <eric@webkit.org>
13821
13822         Reviewed by Adam Barth.
13823
13824         Make sheriff-bot robust against exceptions from commands
13825         https://bugs.webkit.org/show_bug.cgi?id=63211
13826
13827         sheriff-bot was acting strange this afternoon.
13828         We don't know if this will fix the cause, but
13829         at least it adds some unit tests and catches
13830         one possible cause.
13831
13832         * Scripts/webkitpy/tool/bot/sheriffircbot.py:
13833         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
13834
13835 2011-06-22  Mark Rowe  <mrowe@apple.com>
13836
13837         Reviewed by David Levin.
13838
13839         <http://webkit.org/b/63212> TestWebKitAPI Xcode project has bogus settings since r86287
13840
13841         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Fix the reference to libgtest.a to
13842         be relative to the built products directory rather than using a hard-coded path. Remove the
13843         explicit settings of FRAMEWORK_SEARCH_PATHS, HEADER_SEARCH_PATHS and LIBRARY_SEARCH_PATHS
13844         as they're all unnecessary. If they were necessary they should be set in the .xcconfig file
13845         rather than in the .xcodeproj directly.
13846
13847 2011-06-22  Dirk Pranke  <dpranke@chromium.org>
13848
13849         Reviewed by Ojan Vafai.
13850
13851         new-run-webkit-tests: remove obsolete port.shut_down_http_server method
13852         https://bugs.webkit.org/show_bug.cgi?id=59993
13853
13854         * Scripts/webkitpy/layout_tests/port/base.py:
13855         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
13856         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
13857         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
13858         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
13859         * Scripts/webkitpy/layout_tests/port/gtk.py:
13860         * Scripts/webkitpy/layout_tests/port/mac.py:
13861         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
13862         * Scripts/webkitpy/layout_tests/port/qt.py:
13863         * Scripts/webkitpy/layout_tests/port/win.py:
13864
13865 2011-06-22  Luke Macpherson   <macpherson@chromium.org>
13866
13867         Reviewed by Andreas Kling.
13868
13869         Move macpherson from contributor list to committer list.
13870         https://bugs.webkit.org/show_bug.cgi?id=63179
13871
13872         * Scripts/webkitpy/common/config/committers.py:
13873         Move macpherson from contributor list to committer list.
13874
13875 2011-06-22  David Levin  <levin@chromium.org>
13876
13877         Reviewed by Adam Barth.
13878
13879         check-webkit-style should detect returning (Own|Ref)Ptr instead of the Pass*Ptr version.
13880         https://bugs.webkit.org/show_bug.cgi?id=63204
13881
13882         * Scripts/webkitpy/style/checkers/cpp.py: Added a check for the return value and combined
13883           with similar code for the parameter checking.
13884         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Removed pass_ptr checks from
13885           those done for single lines since they don't make sense in that case (variable decls look like function decls).
13886           Removed some redundant comments (one of which was slightly wrong).
13887           Added checks for the new functionality and minor other test changes.
13888
13889 2011-06-22  Nate Chapin  <japhet@chromium.org>
13890
13891         Reviewed by Adam Barth.
13892
13893         Add libsoup to libWebCoreInternals build.
13894         https://bugs.webkit.org/show_bug.cgi?id=62066
13895
13896         * GNUmakefile.am:
13897
13898 2011-06-22  Dirk Pranke  <dpranke@chromium.org>
13899
13900         Reviewed by Tony Chang.
13901
13902         nrwt: handle missing httpd cleanly
13903         https://bugs.webkit.org/show_bug.cgi?id=62027
13904
13905         We had reverted the change in r89414, so this adds it back in
13906         (modifying check_sys_deps() in port/base.py to check for an
13907         installed web server).
13908
13909         This change then fixes the change in r89414 to stub out the
13910         check_sys_deps() in the test port so that the unit tests run correctly.
13911
13912         * Scripts/webkitpy/layout_tests/port/base.py:
13913         * Scripts/webkitpy/layout_tests/port/test.py:
13914
13915 2011-06-22  David Levin  <levin@chromium.org>
13916
13917         Reviewed by Adam Barth.
13918
13919         check-webkit-style should check for invalid uses of RefPtr/OwnPtr as parameters.
13920         https://bugs.webkit.org/show_bug.cgi?id=63188
13921
13922         * Scripts/webkitpy/style/checkers/cpp.py: Added the check.
13923         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added tests.
13924
13925 2011-06-22  Tony Chang  <tony@chromium.org>
13926
13927         Reviewed by Ojan Vafai.
13928
13929         Add a buildbot for testing ENABLE(CSS3_FLEXBOX)
13930         https://bugs.webkit.org/show_bug.cgi?id=62891
13931
13932         * BuildSlaveSupport/build.webkit.org-config/config.json:
13933
13934 2011-06-22  Dirk Pranke  <dpranke@chromium.org>
13935
13936         Reviewed by Ojan Vafai.
13937
13938         nrwt: don't look for http lock when running the test port
13939         https://bugs.webkit.org/show_bug.cgi?id=63158
13940
13941         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
13942         * Scripts/webkitpy/layout_tests/port/test.py:
13943
13944 2011-06-22  Balazs Kelemen  <kbalazs@webkit.org>
13945
13946         Reviewed by Andreas Kling.
13947
13948         [Qt][WK2] Set up plugin tests
13949         https://bugs.webkit.org/show_bug.cgi?id=63066
13950
13951         * WebKitTestRunner/qt/TestControllerQt.cpp:
13952         (WTR::TestController::initializeTestPluginDirectory):
13953         Use the QTWEBKIT_PLUGIN_PATH environment variable as the path
13954         of the test plugin like DRT does.
13955
13956 2011-06-22  Eric Seidel  <eric@webkit.org>
13957
13958         Reviewed by Ojan Vafai.
13959
13960         Make sheriff-bot rollout messages a little nicer
13961         https://bugs.webkit.org/show_bug.cgi?id=63107
13962
13963         It annoyed me this afternoon that I had to convert sheriff-bots "r12345" revisions
13964         into urls myself.  So I have now fixed its "preparing" message to include a url.
13965
13966         I also figured that I should make the messages mention all of the responsible parties
13967         so that rollouts are never surprises.  If you're in the channel and were involved
13968         in a patch, you will see if someone is using sheriff-bot to rollout a patch.
13969
13970         As part of doing this I also changed (and tested) _parse_args to fail-fast
13971         when given invalid args.
13972
13973         * Scripts/webkitpy/tool/bot/irc_command.py:
13974         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
13975         * Scripts/webkitpy/tool/bot/sheriff.py:
13976         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
13977
13978 2011-06-22  Dirk Pranke  <dpranke@chromium.org>
13979
13980         Reviewed by Tony Chang.
13981
13982         run-webkit-websocketserver fails to stop websocket server
13983         https://bugs.webkit.org/show_bug.cgi?id=63123
13984
13985         There's a ten-character change in new-run-webkit-websocketserver
13986         that fixes the actual bug, and then we update the code in
13987         new-run-webkit-httpd to do the actual stop call, and add more tests
13988         (refactoring the existing test code so that it is shared).
13989
13990         * Scripts/new-run-webkit-httpd:
13991         * Scripts/new-run-webkit-websocketserver:
13992         * Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py:
13993
13994 2011-06-22  Adam Roben  <aroben@apple.com>
13995
13996         Add links to existing bugs related to failing tests on TestFailures page
13997
13998         I changed the layout of the page a little to make it easier to read with all the new
13999         information. Passing/failing revisions have been moved down below the list of tests to be
14000         closer to the existing bugs and the new bug link. And each set of tests and its relevant
14001         information is in a light gray box.
14002
14003         Fixes <http://webkit.org/b/61665> TestFailures page should link to existing bugs when
14004         possible
14005
14006         Reviewed by Darin Adler.
14007
14008         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js: Added.
14009         (Bugzilla): This new class represents a single Bugzilla instance.
14010         (Bugzilla.prototype.quickSearch): Searches Bugzilla using its Quick Search functionality,
14011         passing the resulting bug titles and URLs to the callback when complete. If called multiple
14012         times with the same query before the query returns, caches the callbacks so that only one
14013         query is sent over the wire. When the query completes, all pending callbacks are called.
14014
14015         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
14016         (#failure-history): Reduce the margin/padding on the top-level list a bit.
14017         (#failure-history > li): Put each set of tests in a gray box, and indent most information
14018         inside the box.
14019         (.test-list): Unindent the list of failing tests so it is visually at the top level.
14020         (.new-and-existing-bugs): Reduce the space at the bottom of this area so that the bottom of
14021         each box isn't a big empty space.
14022         (.existing-bugs-list): Use a smaller text size for existing bugs, since their titles can be
14023         quite long.
14024
14025         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
14026         (addQueryParametersToURL): New function extracted from
14027         ViewController.prototype._domForNewAndExistingBugs.
14028
14029         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
14030         (ViewController): Take and store an optional Bugzilla instance.
14031         (ViewController.prototype._displayBuilder): Give the top-level list an id for styling
14032         purposes and move the list of failing tests above all other information. Only show bug
14033         information once we've determined the most-recent passing revision for a set of tests. It's
14034         not that useful to file a new bug before this information has been determined, and searching
14035         for existing bugs before we've figured out which tests started failing at the same time
14036         would end up giving you information about a bunch of unrelated tests.
14037         (ViewController.prototype._domForNewAndExistingBugs): Renamed from _domForNewBugLink. Now
14038         returns a DocumentFragment instead of an HTMLParagraphElement. If we don't have a Bugzilla
14039         instance, just returns an empty DocumentFragment. Starts a search for bugs related to the
14040         failing tests, and adds links to the bugs when the search completes.
14041
14042         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pass a
14043         Bugzilla instance for bugs.webkit.org to the ViewController.
14044
14045 2011-06-22  Carlos Garcia Campos  <cgarcia@igalia.com>
14046
14047         Reviewed by Martin Robinson.
14048
14049         [GTK] Apply current view settings to newly created views in GtkLauncher
14050         https://bugs.webkit.org/show_bug.cgi?id=63142
14051
14052         * GtkLauncher/main.c:
14053         (createWebViewCb):
14054
14055 2011-06-22  Carlos Garcia Campos  <cgarcia@igalia.com>
14056
14057         Reviewed by Martin Robinson.
14058
14059         [GTK] Use GOption to parse main arguments in GtkLauncher
14060         https://bugs.webkit.org/show_bug.cgi?id=63060
14061
14062         Instead of getting the settings from the view, create a new
14063         settings object not associated with any view so that we can parse
14064         the command line arguments before creating the view. It will also
14065         allows to use the same settings for more than one view if we
14066         eventually support passing more than one uri from the command
14067         line.
14068
14069         * GtkLauncher/main.c:
14070         (parseOptionEntryCallback): Make this function static. Also use
14071         g_ascii_strtoll/g_ascii_strtod to parse int/float values instead
14072         of atoi/atof to provide better error messages. And always fill the
14073         error struct when the function fails.
14074         (getOptionEntriesFromWebKitWebSettings): Use an array of
14075         GOptionEntry structs instead of a GArray, since that is what
14076         g_option_group_add_entries() expects and the array size doesn't
14077         change.
14078         (addWebSettingsGroupToContext): Helper function to add the
14079         websettings group to the option context.
14080         (main): Create the option context to parse command line arguments.
14081
14082 2011-06-22  Martin Robinson  <mrobinson@igalia.com>
14083
14084         Reviewed by Adam Roben.
14085
14086         [GTK] Implement pixel dump support for WebKitTestRunner
14087         https://bugs.webkit.org/show_bug.cgi?id=58242
14088
14089         * DumpRenderTree/gtk/GNUmakefile.ImageDiff.am: Added.
14090         * GNUmakefile.am: Separate out the ImageDiff bits of the GNUmakefile.
14091         * Scripts/old-run-webkit-tests: Update to add support for the GTK+ port.
14092         * WebKitTestRunner/GNUmakefile.am: Added Cairo specific files to the build.
14093         * WebKitTestRunner/TestInvocation.cpp: Moved this code to the platform-independent
14094         file so that we can share it across platforms.
14095         (WTR::TestInvocation::compareActualHashToExpectedAndDumpResults):
14096         * WebKitTestRunner/TestInvocation.h: Updated method list.
14097         * WebKitTestRunner/cairo/TestInvocationCairo.cpp: Added Cairo implementation.
14098         * WebKitTestRunner/cg/TestInvocationCG.cpp:
14099         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected): Updated to use new helper.
14100
14101 2011-06-22  Sheriff Bot  <webkit.review.bot@gmail.com>
14102
14103         Unreviewed, rolling out r89414.
14104         http://trac.webkit.org/changeset/89414
14105         https://bugs.webkit.org/show_bug.cgi?id=63130
14106
14107         It broke webkitpy tests (Requested by Ossy on #webkit).
14108
14109         * Scripts/webkitpy/layout_tests/port/base.py:
14110
14111 2011-06-22  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
14112
14113         Reviewed by Dirk Pranke.
14114
14115         [NRWT] Print error message when there is no httpd present on the system
14116         https://bugs.webkit.org/show_bug.cgi?id=62027
14117
14118         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
14119
14120 2011-06-21  Dirk Pranke  <dpranke@chromium.org>
14121
14122         Unreviewed, build fix.
14123         Fix crashes in new-run-webkit-tests resulting from the
14124         change to the http_server logic in r89400. Python 2.5
14125         on Mac 10.5 has some weird error paths.
14126
14127         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
14128         * Scripts/webkitpy/common/system/executive.py:
14129
14130 2011-06-21  MORITA Hajime  <morrita@google.com>
14131
14132         Unreviewed, rolling out r89401 and r89403.
14133         http://trac.webkit.org/changeset/89401
14134         http://trac.webkit.org/changeset/89403
14135         https://bugs.webkit.org/show_bug.cgi?id=62970
14136
14137         Breaks mac build and mistakenly enables the spellcheck API
14138
14139         * Scripts/build-webkit:
14140         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
14141
14142 2011-06-21  Kent Tamura  <tkent@chromium.org>
14143
14144         [Mac] Sort Xcode project files.
14145
14146         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
14147
14148 2011-06-20  MORITA Hajime  <morrita@google.com>
14149
14150         Reviewed by Kent Tamura.
14151
14152         Spellcheck API should be build-able.
14153         https://bugs.webkit.org/show_bug.cgi?id=62970
14154
14155         * Scripts/build-webkit:
14156
14157 2011-06-21  Dirk Pranke  <dpranke@chromium.org>
14158
14159         Reviewed by Tony Chang.
14160
14161         nrwt: attempt #3 at fixing server startup/shutdown
14162         https://bugs.webkit.org/show_bug.cgi?id=62829
14163
14164         Attempt yet again to make starting and stopping servers
14165         reliable. It turns out that apache has races between when the
14166         ctl process exists and when it writes/removes the pid file.
14167         This change accounts for the races.
14168
14169         * Scripts/new-run-webkit-httpd:
14170         * Scripts/webkitpy/common/system/executive.py:
14171         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
14172         * Scripts/webkitpy/layout_tests/port/base.py:
14173         * Scripts/webkitpy/layout_tests/port/http_server.py:
14174         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
14175         * Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Added.
14176         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
14177         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
14178
14179 2011-06-21  Dmitry Lomov  <dslomov@google.com>
14180
14181         Reviewed by Adam Roben.
14182
14183         https://bugs.webkit.org/show_bug.cgi?id=63080
14184         Unit-tests step on test-only bot is broken.
14185         Removing the step until run-api-tests is fixed.
14186
14187         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
14188
14189 2011-06-20  Dmitry Lomov  <dslomov@google.com>
14190
14191         Reviewed by David Levin.
14192
14193         https://bugs.webkit.org/show_bug.cgi?id=63017
14194         "Unit tests" step on bots should provide more information in case of failure.
14195
14196         * BuildSlaveSupport/build.webkit.org-config/master.cfg: added --verbose flag to UnitTests step.
14197
14198 2011-06-20  Dirk Pranke  <dpranke@chromium.org>
14199
14200         Reviewed by Ojan Vafai.
14201
14202         new-run-webkit-tests: order shards by name, not number of tests
14203         https://bugs.webkit.org/show_bug.cgi?id=62753
14204
14205         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
14206
14207 2011-06-20  Dirk Pranke  <dpranke@chromium.org>
14208
14209         Reviewed by Ojan Vafai.
14210
14211         we should only log the test names to the tests_run.txt files during layout_tests
14212         https://bugs.webkit.org/show_bug.cgi?id=62751
14213
14214         * Scripts/old-run-webkit-tests:
14215         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
14216
14217 2011-06-20  Eric Seidel  <eric@webkit.org>
14218
14219         Reviewed by Adam Barth.
14220
14221         commit-queue/ews-bots should run run-bindings-tests for every patch
14222         https://bugs.webkit.org/show_bug.cgi?id=63010
14223
14224         * Scripts/webkitpy/common/config/ports.py:
14225         * Scripts/webkitpy/tool/commands/download_unittest.py:
14226         * Scripts/webkitpy/tool/steps/runtests.py:
14227         * Scripts/webkitpy/tool/steps/runtests_unittest.py:
14228         * Scripts/webkitpy/tool/steps/steps_unittest.py:
14229
14230 2011-06-20  Chang Shu  <cshu@webkit.org>
14231
14232         Reviewed by Adam Barth.
14233
14234         fast/js/sputnik/Conformance/10_Execution_Contexts/10.2_Entering_An_Execution_Context/10.2.2_Eval_Code tests fail when showModalDialog isn't supported
14235         https://bugs.webkit.org/show_bug.cgi?id=53676
14236
14237         * WebKitTestRunner/TestController.cpp:
14238         (WTR::TestController::initialize):
14239
14240 2011-06-20  Eric Seidel  <eric@webkit.org>
14241
14242         Reviewed by Adam Barth.
14243
14244         buildbots should run run-bindings-tests
14245         https://bugs.webkit.org/show_bug.cgi?id=63001
14246
14247         run-bindings-tests takes 2.4 seconds on my machine, so
14248         I expect this will not be noticeable on the bots.
14249
14250         I also removed tiger-specific checks since Tiger is no
14251         longer a supported platform for webkit.
14252
14253         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
14254
14255 2011-06-20  Carlos Garcia Campos  <cgarcia@igalia.com>
14256
14257         Unreviewed. Fix WebKit2 GTK build after r89249.
14258
14259         * GtkLauncher/main.c:
14260         (main):
14261
14262 2011-06-20  Pavel Feldman  <pfeldman@chromium.org>
14263
14264         Reviewed by Yury Semikhatsky.
14265
14266         Web Inspector: remove LayoutTestController::setTimelineProfilingEnabled.
14267         https://bugs.webkit.org/show_bug.cgi?id=62994
14268
14269         * DumpRenderTree/LayoutTestController.cpp:
14270         (closeWebInspectorCallback):
14271         (LayoutTestController::staticFunctions):
14272         * DumpRenderTree/LayoutTestController.h:
14273         * DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
14274         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
14275         * DumpRenderTree/chromium/LayoutTestController.cpp:
14276         (LayoutTestController::LayoutTestController):
14277         * DumpRenderTree/chromium/LayoutTestController.h:
14278         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
14279         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
14280         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
14281         * DumpRenderTree/qt/LayoutTestControllerQt.h:
14282         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
14283         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
14284         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
14285         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
14286         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
14287
14288 2011-06-20  Lukasz Slachciak  <l.slachciak@samsung.com>
14289
14290         Reviewed by Martin Robinson.
14291
14292         [GTK] General mechanism for adjusting WebKitWebSettings in GtkLauncher.
14293         https://bugs.webkit.org/show_bug.cgi?id=55308
14294
14295         * GtkLauncher/main.c: Added general mechanism for adjusting WebKitWebSettings.
14296         (parseOptionEntryCallback): Callback for parsing option entry.
14297         (getOptionEntriesFromWebKitWebSettings): Basing on the WebKitWebSettings properties prepare
14298         list of option entries.
14299         (transformStringToBoolean): Transform function for boolean convertion used by parseOptionEntryCallback.
14300         (transformStringToInt): Transform function for int convertion used by parseOptionEntryCallback.
14301         (transformStringToFloat): Transform function for float convertion used by parseOptionEntryCallback.
14302         (parseAdditionalOptions): Help function parsing additional commandline options.
14303         (main): Added call to parseAdditionalOptions.
14304
14305 2011-06-20  Viatcheslav Ostapenko  <ostapenko.viatcheslav@nokia.com>
14306
14307         Reviewed by Andreas Kling.
14308
14309         [Qt] [WK2] Compile error in ActivateFontsQt.cpp on Mac
14310         https://bugs.webkit.org/show_bug.cgi?id=62869
14311
14312         Add QFont include to fix "incomlete type QFont" error.
14313
14314         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
14315
14316 2011-06-19  Adam Bergkvist  <adam.bergkvist@ericsson.com>
14317
14318         Reviewed by Martin Robinson.
14319
14320         [GTK] Enable Media Stream feature and make it default on
14321         https://bugs.webkit.org/show_bug.cgi?id=60394
14322
14323         * Scripts/build-webkit:
14324
14325 2011-06-19  Sheriff Bot  <webkit.review.bot@gmail.com>
14326
14327         Unreviewed, rolling out r89198.
14328         http://trac.webkit.org/changeset/89198
14329         https://bugs.webkit.org/show_bug.cgi?id=62953
14330
14331         Chromium breakage on "Linux Tests" (Requested by senorblanco
14332         on #webkit).
14333
14334         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
14335         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
14336         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
14337
14338 2011-06-18  Daniel Bates  <dbates@webkit.org>
14339
14340         Reviewed by Eric Seidel.
14341
14342         Git.push_local_commits_to_server() should use passed username and
14343         password when we don't have cached credentials
14344         https://bugs.webkit.org/show_bug.cgi?id=62941
14345
14346         Fixes a issue where Git.push_local_commits_to_server() would always look
14347         for cached credentials regardless of whether a username and password
14348         were passed as arguments. Instead, if given both a username and password
14349         then we shouldn't look for cached credentials (since we were explicitly given
14350         credentials).
14351
14352         * Scripts/webkitpy/common/checkout/scm/git.py:
14353         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py: Added unit tests.
14354
14355 2011-06-18  Daniel Bates  <dbates@rim.com>
14356
14357         Rubber-stamped by Eric Seidel.
14358
14359         Remove run-javascriptcore-tests --skip-build option; Instead add --build/no-build options
14360         https://bugs.webkit.org/show_bug.cgi?id=62767
14361
14362         Part 3 of 3
14363
14364         Remove the command-line option --skip-build now that the master BuildBot has been restarted
14365         following <http://trac.webkit.org/changeset/89138>.
14366
14367         Instead, specify --no-build to get analogous functionality.
14368
14369         * Scripts/run-javascriptcore-tests:
14370
14371 2011-06-18  Dirk Pranke  <dpranke@chromium.org>
14372
14373         Reviewed by Tony Chang.
14374
14375         new-run-webkit-tests: fix test sorting
14376         https://bugs.webkit.org/show_bug.cgi?id=58691
14377
14378         This changes the sorting to use a "natural" sort order such that
14379         foo2.html sorts before foo13.html, and sorts the tests so
14380         that all the tests in a directory are run before any test in a
14381         subdirectory.
14382
14383         This change also adds a split() method to the filesystem since
14384         that is needed to cluster files into a directory otherwise.
14385
14386         * Scripts/webkitpy/common/system/filesystem.py:
14387         * Scripts/webkitpy/common/system/filesystem_mock.py:
14388         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
14389         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
14390
14391 2011-06-18  Evan Martin  <evan@chromium.org>
14392
14393         Reviewed by Darin Adler.
14394
14395         Compiler warning in double->int conversion in TestNetscapePlugIn
14396         https://bugs.webkit.org/show_bug.cgi?id=51554
14397
14398         This code is providing a JS API for removing a property by numeric ID.
14399         We get it from JS as a double, and must convert it to an int before using it.
14400
14401         Test: LayoutTests/plugins/npruntime/remove-property.html
14402
14403         * DumpRenderTree/TestNetscapePlugIn/Tests/NPRuntimeRemoveProperty.cpp:
14404         (NPRuntimeRemoveProperty::TestObject::invoke):
14405
14406 2011-06-18  Patrick Gansterer  <paroga@webkit.org>
14407
14408         Reviewed by Eric Seidel.
14409
14410         Don't exit webkit-patch when running on Win32 python
14411         https://bugs.webkit.org/show_bug.cgi?id=55811
14412
14413         Most of the basic actions work in the meantime,
14414         so let people play with in on Win32 now.
14415
14416         * Scripts/webkit-patch:
14417
14418 2011-06-18  Alice Boxhall  <aboxhall@chromium.org>
14419
14420         Reviewed by Ojan Vafai.
14421
14422         Convert json_results_generator.py to output version 4 JSON.
14423         https://bugs.webkit.org/show_bug.cgi?id=60869
14424
14425         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
14426         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
14427         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
14428
14429 2011-06-18  Berend-Jan Wever  <skylined@chromium.org>
14430
14431         Reviewed by Eric Seidel.
14432
14433         Some names with spaces are incorrectly reported as not containing spaces
14434         while running prepare-ChangeLog
14435         https://bugs.webkit.org/show_bug.cgi?id=55572
14436
14437         * Scripts/VCSUtils.pm: Updated CHANGE_LOG_NAME check for spaces to allow
14438           non-alphanumeric characters in names.
14439
14440 2011-06-18  Michael Saboff  <msaboff@apple.com>
14441
14442         Reviewed by Eric Seidel.
14443
14444         Enhancement: New Script to Format malloc_history output as tree
14445         https://bugs.webkit.org/show_bug.cgi?id=56378
14446
14447         New script to process malloc_history output to tree format.
14448
14449         * Scripts/malloc-tree: Added.
14450
14451 2011-06-17  Dmitry Lomov  <dslomov@google.com>
14452
14453         Reviewed by Daniel Bates.
14454
14455         https://bugs.webkit.org/show_bug.cgi?id=62918
14456         Pass configuration to UnitTests step in master.cfg
14457
14458         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
14459
14460 2011-06-17  Leandro Pereira  <leandro@profusion.mobi>
14461
14462         Reviewed by Eric Seidel.
14463
14464         [EFL] Add Evas-based ImageDiff implementation.
14465         https://bugs.webkit.org/show_bug.cgi?id=61942
14466
14467         Build system changes will be performed as soon as the complete
14468         DumpRenderTree implementation is upstreamed.
14469
14470         * DumpRenderTree/efl/ImageDiff.cpp: Added.
14471         (pixelFromImageData):
14472         (differenceImageFromDifferenceBuffer):
14473         (computeDistanceBetweenPixelComponents):
14474         (calculatePixelDifference):
14475         (calculateDifference):
14476         (getTemporaryFile):
14477         (printImage):
14478         (printImageDifferences):
14479         (resizeEcoreEvasIfNeeded):
14480         (readImageFromStdin):
14481         (parseCommandLineOptions):
14482         (shutdownEfl):
14483         (abortWithErrorMessage):
14484         (initEfl):
14485         (main):
14486
14487 2011-06-17  Tony Chang  <tony@chromium.org>
14488
14489         Reviewed by Dirk Pranke.
14490
14491         [chromium] force rebaseline-chromium-webkit-tests to use the chromium port
14492         https://bugs.webkit.org/show_bug.cgi?id=62907
14493
14494         This fixes a bug where the ImageDiff on the user's machine doesn't
14495         match the logic used on the Chromium buildbots.  This has the downside
14496         of requiring that the Chromium ImageDiff is built, but it turns out
14497         that even with --tolerance 0, the CG ImageDiff will allow small pixel
14498         differences through.
14499
14500         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
14501
14502 2011-06-17  Brent Fulgham  <bfulgham@webkit.org>
14503
14504         [WinCairo] Unreviewed build correction.
14505
14506         * DumpRenderTree/win/DumpRenderTreeCairo.vsprops:
14507         Update WinCairo property sheet to link against the Cairo DLL,
14508         now that we are using a more recent Cairo build.
14509
14510 2011-06-17  Brent Fulgham  <bfulgham@webkit.org>
14511
14512         Reviewed by Brian Weinstein.
14513
14514         Create a simple example of using the COM-based C++ listeners.
14515         A new onclick event is attached (via C++) to the WebKit logo
14516         in the default test pattern displayed on startup.  Clicking on
14517         the logo causes a message box to be displayed.
14518         https://bugs.webkit.org/show_bug.cgi?id=61885
14519
14520         * WinLauncher/DOMDefaultImpl.h: Added.  Stub implementation of
14521           the WebScriptObject and DOMEventListener.
14522         * WinLauncher/WinLauncher.cpp:
14523         (SimpleEventListener::SimpleEventListener): Example implementation
14524           of a simple DOM event listener.
14525         (SimpleEventListener::handleEvent):
14526         (WinLauncherWebHost::didFinishLoadForFrame): Added implementation
14527          to bind a C++ method to the 'onclick' event for the WebKit logo.
14528         (_tWinMain):
14529         * WinLauncher/WinLauncher.h:
14530         * WinLauncher/WinLauncher.vcproj: Add new DOMDefaultImpl.h file.
14531
14532 2011-06-17  Chang Shu  <cshu@webkit.org>
14533
14534         Reviewed by Andreas Kling.
14535
14536         [Qt] [WK2] Support undo/redo in MiniBrowser
14537         https://bugs.webkit.org/show_bug.cgi?id=62809
14538
14539         * MiniBrowser/qt/BrowserWindow.cpp:
14540         (BrowserWindow::BrowserWindow):
14541
14542 2011-06-17  Daniel Bates  <dbates@rim.com>
14543
14544         Rubber-stamped by Eric Seidel.
14545
14546         Remove run-javascriptcore-tests --skip-build option; Instead add --build/no-build options
14547         https://bugs.webkit.org/show_bug.cgi?id=62767
14548
14549         Part 2 of 3
14550
14551         Update the BuildBot master.cfg to use the new --no-build run-javascriptcore-tests
14552         command line option instead of --skip-build.
14553
14554         We'll remove --skip-build from run-javascriptcore-test in a follow up patch once the
14555         master buildbot picks up the change.
14556
14557         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
14558
14559 2011-06-17  Daniel Bates  <dbates@rim.com>
14560
14561         Rubber-stamped by Eric Seidel.
14562
14563         Remove run-javascriptcore-tests --skip-build option; Instead add --build/no-build options
14564         https://bugs.webkit.org/show_bug.cgi?id=62767
14565
14566         Part 1 of 3
14567
14568         Towards renaming the option --skip-build to --no-build. Add --build/--no-build options to
14569         run-javascriptcore-tests. Also, add a remark about these options to the usage message
14570         printed by run-javascriptcore-tests.
14571
14572         We'll remove --skip-build in a follow up patch once the master buildbot picks up the changes.
14573
14574         * Scripts/run-javascriptcore-tests: Additionally, remove unnecessary initialization
14575           of $showHelp.
14576
14577 2011-06-17  Joone Hur  <joone.hur@collabora.co.uk>
14578
14579         Reviewed by Martin Robinson.
14580
14581         [GTK] Replace GdkRectangle by cairo_rectangle_int_t
14582         https://bugs.webkit.org/show_bug.cgi?id=60687
14583
14584         Replace GdkRectangle by cairo_rectangle_int_t.
14585
14586         * DumpRenderTree/gtk/PixelDumpSupportGtk.cpp: Replace GdkRectangle by cairo_rectangle_int_t.
14587         (createBitmapContextFromWebView):
14588         * DumpRenderTree/gtk/TextInputController.cpp: Ditto.
14589         (firstRectForCharacterRangeCallback):
14590
14591 2011-06-16  Dmitry Lomov  <dslomov@google.com>
14592
14593         Reviewed by David Levin.
14594
14595         https://bugs.webkit.org/show_bug.cgi?id=62603
14596         Run run-api-tests on bots in debug mode.
14597         Release mode is disabled due to https://bugs.webkit.org/show_bug.cgi?id=61812.
14598
14599         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
14600
14601 2011-06-16  Leandro Pereira  <leandro@profusion.mobi>
14602
14603         Reviewed by Eric Seidel.
14604
14605         [EFL] DumpRenderTree: Add GCController, PixelDumpSupport and
14606         WorkQueueItem.
14607         They will be plugged into the build system once the EFL's DRT is
14608         fully upstreamed.
14609         https://bugs.webkit.org/show_bug.cgi?id=61958
14610
14611         * DumpRenderTree/efl: Added.
14612         * DumpRenderTree/efl/GCControllerEfl.cpp: Added.
14613         (GCController::collect):
14614         (GCController::collectOnAlternateThread):
14615         (GCController::getJSObjectCount):
14616         * DumpRenderTree/efl/PixelDumpSupportEfl.cpp: Added.
14617         (createBitmapContextFromWebView):
14618         * DumpRenderTree/efl/WorkQueueItemEfl.cpp: Added.
14619         (LoadItem::invoke):
14620         (LoadHTMLStringItem::invoke):
14621         (ReloadItem::invoke):
14622         (ScriptItem::invoke):
14623         (BackForwardItem::invoke):
14624         * DumpRenderTree/efl/WorkQueueItemEfl.h: Added.
14625         (WorkQueueItem::~WorkQueueItem):
14626         (LoadItem::LoadItem):
14627         (LoadHTMLStringItem::LoadHTMLStringItem):
14628         (ScriptItem::ScriptItem):
14629         (LoadingScriptItem::LoadingScriptItem):
14630         (LoadingScriptItem::invoke):
14631         (NonLoadingScriptItem::NonLoadingScriptItem):
14632         (NonLoadingScriptItem::invoke):
14633         (BackForwardItem::BackForwardItem):
14634         (BackItem::BackItem):
14635         (ForwardItem::ForwardItem):
14636
14637 2011-06-16  Kevin Ollivier  <kevino@theolliviers.com>
14638
14639         [wx] Unreviewed build fix, make sure the waf build checks the webKitBranchBuild
14640         preference like the perl scripts do.
14641
14642         * waf/build/build_utils.py:
14643         * waf/build/settings.py:
14644
14645 2011-06-16  Tommy Widenflycht  <tommyw@google.com>
14646
14647         Reviewed by Tony Gentilcore.
14648
14649         MediaStream API: Runtime enabled features
14650         https://bugs.webkit.org/show_bug.cgi?id=62798
14651
14652         Add the WebRuntimeFeatures API for Media Streams and enable the feature by default in DumpRenderTree.
14653
14654         * DumpRenderTree/chromium/TestShell.cpp:
14655         (TestShell::TestShell):
14656
14657 2011-06-16  Torsten Kurbad  <gentoo@tk-webart.de>
14658
14659         Reviewed by Eric Seidel.
14660
14661         Add support for sys.platform == 'linux3' by relaxing
14662         conditional statements and/or extending lookup dicts.
14663         This way, Linux kernels >=3.0-rc1 get supported in the
14664         build process.
14665         https://bugs.webkit.org/show_bug.cgi?id=62730
14666
14667         * Scripts/webkitpy/common/system/executive.py:
14668         * Scripts/webkitpy/common/system/file_lock.py:
14669         * Scripts/webkitpy/common/system/path_unittest.py:
14670         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
14671         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
14672         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
14673         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
14674         * Scripts/webkitpy/layout_tests/port/factory.py:
14675         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
14676         * Scripts/webkitpy/layout_tests/port/qt.py:
14677         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
14678
14679 2011-06-16  Sheriff Bot  <webkit.review.bot@gmail.com>
14680
14681         Unreviewed, rolling out r88796.
14682         http://trac.webkit.org/changeset/88796
14683         https://bugs.webkit.org/show_bug.cgi?id=62790
14684
14685         It made fast/dom/nodesFromRect-basic.html time out on Qt,
14686         64-bit, debug mode (Requested by Ossy on #webkit).
14687
14688         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
14689         (LayoutTestController::nodesFromRect):
14690         * DumpRenderTree/qt/LayoutTestControllerQt.h:
14691         * DumpRenderTree/qt/PlainTextControllerQt.cpp:
14692         (PlainTextController::plainText):
14693         * DumpRenderTree/qt/PlainTextControllerQt.h:
14694
14695 2011-06-15  Dirk Pranke  <dpranke@chromium.org>
14696
14697         Unreviewed, build fix.
14698
14699         Revert the patches in r88995 and r88999; the http servers don't
14700         seem to be shutting down reliably.
14701
14702         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
14703         * Scripts/webkitpy/layout_tests/port/base.py:
14704         * Scripts/webkitpy/layout_tests/port/http_server.py:
14705         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
14706         * Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Removed.
14707         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
14708         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
14709
14710 2011-06-15  Dirk Pranke  <dpranke@chromium.org>
14711
14712         Unreviewed, build fix.
14713
14714         Add some debugging messages to triage problems introduced in
14715         r88995.
14716
14717         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
14718
14719 2011-06-15  Dirk Pranke  <dpranke@chromium.org>
14720
14721         Reviewed by Tony Chang.
14722
14723         nrwt: should clean up stale server processes from a previous run
14724         https://bugs.webkit.org/show_bug.cgi?id=62615
14725
14726         This change fixes cases where NRWT might start a server and then
14727         exit, leaving the server still running. Now NRWT should detect
14728         this case and clean up on the next run, without shutting down
14729         any processes it didn't start.
14730
14731         This change also fixes a bug in executive.kill_process on UNIX
14732         where zombies weren't being handled properly.
14733
14734         * Scripts/webkitpy/common/system/executive.py:
14735         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
14736         * Scripts/webkitpy/layout_tests/port/base.py:
14737         * Scripts/webkitpy/layout_tests/port/http_server.py:
14738         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
14739         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
14740         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
14741
14742 2011-06-15  Dirk Pranke  <dpranke@chromium.org>
14743
14744         Reviewed by Tony Chang.
14745
14746         nrwt: fix http, websocket server startup, shutdown
14747         https://bugs.webkit.org/show_bug.cgi?id=62180
14748
14749         This change refactors all of the server-related code to
14750         have consistent semantics. It pushes the start()/stop()
14751         logic into the base class and adds four much smaller routines
14752         for specialization:
14753         _prepare_config() - for optionally writing config files etc.
14754             to disk prior to starting the server
14755         _remove_stale_log_files()
14756         _spawn_process() - the actual launch of the subprocess
14757         _cleanup_after_stop() - and custom cleanup code
14758
14759         Prior to this change the three server implementations were
14760         fairly inconsistent, leading to lots of weirdness on the bots
14761         with servers being left around, etc.
14762
14763         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
14764         * Scripts/webkitpy/layout_tests/port/http_server.py:
14765         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
14766         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
14767
14768 2011-06-15  Dirk Pranke  <dpranke@chromium.org>
14769
14770         Reviewed by Tony Chang.
14771
14772         webkitpy: clean up code prior to functional changes for server startup/shutdown
14773         https://bugs.webkit.org/show_bug.cgi?id=62256
14774
14775         This patch does a bunch of minor cleanup of the code, but there
14776         should be no functional changes except that all of the modules
14777         will now use the same exception type. Other changes include
14778         removing unused code paths and functions, pushing 'name' and
14779         'pid_file' into the base class of the server objects, and using
14780         the Filesystem object instead of the codecs, tempfile, and
14781         shutils modules.
14782
14783         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
14784         * Scripts/webkitpy/layout_tests/port/http_server.py:
14785         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
14786         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
14787
14788 2011-06-15  Dirk Pranke  <dpranke@chromium.org>
14789
14790         Reviewed by Tony Chang.
14791
14792         webkitpy: add integration tests for new-run-webkit-httpd, stop calling shut_down_http_server
14793         https://bugs.webkit.org/show_bug.cgi?id=62251
14794
14795         shut_down_http_server() was a total hack that was only used by
14796         new-run-webkit-httpd, so I've moved the code there and switched
14797         to using executive.kill_process() for the common case. The
14798         method itself will be removed in the patch on bug 59993.
14799
14800         * Scripts/new-run-webkit-httpd:
14801         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
14802         * Scripts/webkitpy/layout_tests/port/http_server.py:
14803         * Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Added.
14804         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
14805
14806 2011-06-15  Dirk Pranke  <dpranke@chromium.org>
14807
14808         Unreviewed, build fix.
14809
14810         Fix test-webkitpy failure introduced in r88986.
14811
14812         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
14813
14814 2011-06-15  Dirk Pranke  <dpranke@chromium.org>
14815
14816         Reviewed by Eric Seidel.
14817
14818         NRWT: doesn't support webarchives, has stale skipping logic.
14819         https://bugs.webkit.org/show_bug.cgi?id=58293
14820
14821         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
14822         * Scripts/webkitpy/layout_tests/port/base.py:
14823         * Scripts/webkitpy/layout_tests/port/webkit.py:
14824
14825 2011-06-15  Jeremy Orlow  <jorlow@webkit.org>
14826
14827         Change my email to @webkit.org
14828
14829         * Scripts/webkitpy/common/config/committers.py:
14830
14831 2011-06-15  Nikolas Zimmermann  <nzimmermann@rim.com>
14832
14833         Reviewed by David Hyatt.
14834
14835         CSS 2.1: Support replaced elements with relative intrinsic sizes
14836         https://bugs.webkit.org/show_bug.cgi?id=15849
14837
14838         CSS 2.1 failure: html4/replaced-intrinsic-ratio-001
14839         https://bugs.webkit.org/show_bug.cgi?id=53099
14840
14841         Recognize *.htm tests in the old & new DRT implementation, needed for the vanilla css2.1 html4 tests.
14842
14843         * Scripts/old-run-webkit-tests:
14844         * Scripts/webkitpy/layout_tests/port/test_files.py:
14845
14846 2011-06-15  Carlos Garcia Campos  <cgarcia@igalia.com>
14847
14848         Reviewed by Martin Robinson.
14849
14850         [GTK] Add a statusbar to MiniBrowser
14851         https://bugs.webkit.org/show_bug.cgi?id=62634
14852
14853         It shows the url of the current hovered link.
14854
14855         * MiniBrowser/gtk/BrowserWindow.c:
14856         (browserWindowConstructed): Create the GtkSatusbar.
14857         (mouseDidMoveOverElement): Update statusbar text.
14858         (browserWindowUIClientInit): Add implementation for
14859         mouseDidMoveOverElement().
14860         * MiniBrowser/gtk/GNUmakefile.am: Add new files to compilation.
14861         * MiniBrowser/gtk/WebBundle/WebBundleMain.c: Added.
14862         (mouseDidMoveOverElement): Pass the url of the current hovered
14863         link to the UI process.
14864         (didCreatePage): Set the UI client adding an implementation for
14865         mouseDidMoveOverElement().
14866         (WKBundleInitialize):
14867         * MiniBrowser/gtk/main.c:
14868         (createWKContextWithInjectedBundle): Create a global context with
14869         the injected bundle.
14870         (loadURI): Use the global context.
14871         (main):
14872
14873 2011-06-14  Lucas Forschler  <lforschler@apple.com>
14874
14875         Reviewed by Stephanie Lewis.
14876
14877         https://bugs.webkit.org/show_bug.cgi?id=62495
14878         Bug 62495 - combine windows and mac kill-old-processes script
14879         Add a platform agnostic script.  Remove the recently added mac flavor.
14880         Note: Leaving the windows one alone, since it is currently used by the WebKit bots.
14881
14882         * BuildSlaveSupport/kill-old-processes: Added.
14883         * BuildSlaveSupport/mac: Removed.
14884         * BuildSlaveSupport/mac/kill-old-processes: Removed.
14885
14886 2011-06-14  Dirk Pranke  <dpranke@chromium.org>
14887
14888         Reviewed by Tony Chang.
14889
14890         nrwt: handle worker exceptions cleanly
14891         https://bugs.webkit.org/show_bug.cgi?id=62614
14892
14893         This change modifiers new-run-webkit-tests to handle exceptions
14894         raised by worker threads better, by:
14895         - capturing the worker's stack and logging it in the manager
14896         - propagating the worker's exception in the caller correctly
14897         - attempting to cancel the workers and clean up even when
14898           we get an unexpected exception
14899
14900         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
14901         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
14902         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
14903         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
14904
14905 2011-06-14  Qi Zhang  <qi.2.zhang@nokia.com>
14906
14907         Reviewed by Laszlo Gombos.
14908
14909         [Qt] fast/dom/HTMLScriptElement/nested-execution.html failed
14910         https://bugs.webkit.org/show_bug.cgi?id=62227
14911
14912         QtWebkit does not yet support different CacheModels. This change will
14913         expose setCacheModel() with a stub implementation, which is enough to pass the LayoutTest.
14914
14915         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
14916         (LayoutTestController::setCacheModel):
14917         * DumpRenderTree/qt/LayoutTestControllerQt.h:
14918
14919 2011-06-14  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
14920
14921         Reviewed by Andreas Kling.
14922
14923         [Qt] tst_QWebFrame::overloadedSlots() fails
14924         https://bugs.webkit.org/show_bug.cgi?id=37319
14925
14926         Since we don't implictly convert 'document' object to QWebElement
14927         in metacalls anymore, change the controller to expect QVariantMap instead.
14928
14929         The method plainText() was updated to use QVariantMap as well to let the bridge
14930         do the conversion directly for us.
14931
14932         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
14933         (LayoutTestController::nodesFromRect):
14934         * DumpRenderTree/qt/LayoutTestControllerQt.h:
14935         * DumpRenderTree/qt/PlainTextControllerQt.cpp:
14936         (PlainTextController::plainText):
14937         * DumpRenderTree/qt/PlainTextControllerQt.h:
14938
14939 2011-06-14  Andras Becsi  <abecsi@webkit.org>
14940
14941         Reviewed by Csaba Osztrogonác.
14942
14943         Remove dead code in DumpRenderTree/TestNetscapePlugIn/main.cpp
14944         https://bugs.webkit.org/show_bug.cgi?id=62630
14945
14946         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
14947         (NPP_New): remove dead variable 'forceCarbon'
14948
14949 2011-06-13  Hayato Ito  <hayato@chromium.org>
14950
14951         Reviewed by Ojan Vafai.
14952
14953         [NRWT] Fix a WebKitDriver so that it returns an actual hash value if DRT returns that.
14954         https://bugs.webkit.org/show_bug.cgi?id=59188
14955
14956         A current WebKitDriver doesn't try to return an actual hash value if DRT does not return an image.
14957         WebKitDriver should always return an actual hash value if DRT returns that.
14958
14959         No tests since I'll add sample reftests to make sure mismatch reftests work later,
14960         which should be in a separate patch.
14961
14962         * Scripts/webkitpy/layout_tests/port/webkit.py:
14963
14964 2011-06-13  Kent Tamura  <tkent@chromium.org>
14965
14966         [Chromium-win] Fix two wrong assertions exposed by r88757.
14967
14968         * DumpRenderTree/chromium/WebThemeEngineDRTWin.cpp:
14969         (WebThemeEngineDRTWin::paintTrackbar):
14970
14971 2011-06-13  Alexey Proskuryakov  <ap@apple.com>
14972
14973         Added an updated e-mail address for Matt Lilek, to make Bugzilla autocomplete
14974         and commit queue happy.
14975
14976         * Scripts/webkitpy/common/config/committers.py:
14977
14978 2011-06-13  Chris Rogers  <crogers@google.com>
14979
14980         Unreviewed build fix.
14981
14982         Build fix to chromium DRT until we can rebaseline some tests
14983         https://bugs.webkit.org/show_bug.cgi?id=62608
14984
14985         * DumpRenderTree/chromium/TestShell.cpp:
14986         (TestShell::TestShell):
14987
14988 2011-06-13  Chris Rogers  <crogers@google.com>
14989
14990         Reviewed by Dirk Pranke.
14991
14992         Add Web Audio support to chromium DRT
14993         https://bugs.webkit.org/show_bug.cgi?id=62306
14994
14995         * DumpRenderTree/chromium/LayoutTestController.cpp:
14996         (LayoutTestController::LayoutTestController):
14997         (LayoutTestController::reset):
14998         (LayoutTestController::setEncodedAudioData):
14999         * DumpRenderTree/chromium/LayoutTestController.h:
15000         (LayoutTestController::encodedAudioData):
15001         (LayoutTestController::shouldDumpAsAudio):
15002         (LayoutTestController::setShouldDumpAsAudio):
15003         * DumpRenderTree/chromium/TestEventPrinter.cpp:
15004         (DRTPrinter::handleAudioHeader):
15005         (TestShellPrinter::handleAudioHeader):
15006         * DumpRenderTree/chromium/TestEventPrinter.h:
15007         * DumpRenderTree/chromium/TestShell.cpp:
15008         (TestShell::TestShell):
15009         (TestShell::dump):
15010
15011 2011-06-13  Jaehun Lim  <ljaehun.lim@samsung.com>
15012
15013         Reviewed by Eric Seidel.
15014
15015         [EFL] Add load error handler to EWebLauncher
15016         https://bugs.webkit.org/show_bug.cgi?id=62347
15017
15018         Add a handler function of "load,error" to EWebLauncher.
15019         EWebLauncher now displays a simple error page on load errors.
15020
15021         * EWebLauncher/main.c:
15022         (on_load_error):
15023         (browserCreate):
15024
15025 2011-06-13  Martin Robinson  <mrobinson@igalia.com>
15026
15027         Reviewed by Eric Seidel.
15028
15029         [GTK] [Qt] Eliminate duplicate TestNetscapePlugin implementation
15030         https://bugs.webkit.org/show_bug.cgi?id=62385
15031
15032         Remove duplicate TestNetscapePlugIn implementation. Having two copies
15033         of this code makes keeping it in sync much more difficult. The files are
15034         almost identical and this change ports the X11 changes to main.cpp.
15035
15036         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
15037         (NPP_New): Do not set the plugin as windowless always. Doing this breaks
15038         some results on Qt and this seems to be unnecessary for the moment.
15039         (keyEventToChar): Added this helper which converts a X11 keycode into a char.
15040         (handleEventX11): Use the handler to properly convert the keycode. Do not print
15041         adjustCursorEvent output. The tests currently do not trigger this for Cocoa and thus
15042         it should not be in the results.
15043         (NPP_GetValue): Properly handle NPPVpluginNeedsXEmbed and make it clearer why the
15044         XP_UNIX block is first by guarding against using a null instance.
15045         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro: Update source list add
15046         the XP_UNIX define for Linux.
15047         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp: Removed.
15048         * GNUmakefile.am: Update source list.
15049
15050 2011-06-13  Sheriff Bot  <webkit.review.bot@gmail.com>
15051
15052         Unreviewed, rolling out r88671.
15053         http://trac.webkit.org/changeset/88671
15054         https://bugs.webkit.org/show_bug.cgi?id=62591
15055
15056         "didn't work" (Requested by dpranke on #webkit).
15057
15058         * Scripts/new-run-webkit-httpd:
15059         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
15060         * Scripts/webkitpy/layout_tests/port/http_server.py:
15061         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
15062         * Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Removed.
15063         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
15064         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
15065
15066 2011-06-13  Nate Chapin  <japhet@chromium.org>
15067
15068         Reviewed by Darin Fisher.
15069
15070         New test method on TestNetscapePlugin for https://bugs.webkit.org/show_bug.cgi?id=61482.
15071
15072         * DumpRenderTree/TestNetscapePlugIn/PluginObject.cpp:
15073
15074 2011-06-13  Dirk Pranke  <dpranke@chromium.org>
15075
15076         Reviewed by Tony Chang.
15077
15078         webkitpy: add integration tests for new-run-webkit-httpd, stop calling shut_down_http_server
15079         https://bugs.webkit.org/show_bug.cgi?id=62251
15080
15081         shut_down_http_server() was a total hack that was only used by
15082         new-run-webkit-httpd, so I've moved the code there and switched
15083         to using executive.kill_process() for the common case. The
15084         method itself will be removed in the patch on bug 59993.
15085
15086         * Scripts/new-run-webkit-httpd:
15087         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
15088         * Scripts/webkitpy/layout_tests/port/http_server.py:
15089         * Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Added.
15090         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
15091
15092 2011-06-13  Tony Chang  <tony@chromium.org>
15093
15094         Reviewed by Dimitri Glazkov.
15095
15096         rename ENABLE_NEW_FLEXBOX to ENABLE_CSS3_FLEXBOX
15097         https://bugs.webkit.org/show_bug.cgi?id=62578
15098
15099         * Scripts/build-webkit:
15100
15101 2011-06-13  Tony Chang  <tony@chromium.org>
15102
15103         Reviewed by Adam Barth.
15104
15105         rename ENABLE_FLEXBOX to ENABLE_NEW_FLEXBOX
15106         https://bugs.webkit.org/show_bug.cgi?id=62545
15107
15108         * Scripts/build-webkit:
15109
15110 2011-06-13  Carlos Garcia Campos  <cgarcia@igalia.com>
15111
15112         Reviewed by Martin Robinson.
15113
15114         [GTK] Export an API similar to WebKit1
15115         https://bugs.webkit.org/show_bug.cgi?id=57820
15116
15117         Build GtkLauncher2 for WebKit2. It's actually the same GtkLauncher
15118         code, but linking to libwebkit2gtk.
15119
15120         * GNUmakefile.am:
15121         * GtkLauncher/main.c:
15122         (createBrowser):
15123         (main):
15124
15125 2011-06-12  MORITA Hajime  <morrita@google.com>
15126
15127         Unreviewed, rolling out r88625.
15128         http://trac.webkit.org/changeset/88625
15129         https://bugs.webkit.org/show_bug.cgi?id=61073
15130
15131         Breaks SL Webkit2 Tests
15132
15133         * WebKitTestRunner/Configurations/InjectedBundleCFLite.vsprops:
15134         * WebKitTestRunner/Configurations/InjectedBundleCommon.vsprops:
15135         * WebKitTestRunner/Configurations/InjectedBundleCoreFoundation.vsprops:
15136         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
15137         (WTR::InjectedBundlePage::didClearWindowForFrame):
15138         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
15139
15140 2011-06-12  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
15141
15142         Reviewed by Antonio Gomes.
15143
15144         [Qt] LayoutTestController needs setTextDirection implementation
15145         https://bugs.webkit.org/show_bug.cgi?id=62442
15146
15147         Implements LayoutController.setTextDirection.
15148
15149         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
15150         (LayoutTestController::setTextDirection):
15151         * DumpRenderTree/qt/LayoutTestControllerQt.h:
15152
15153 2011-06-12  Dominic Cooney  <dominicc@chromium.org>
15154
15155         Reviewed by Hajime Morita.
15156
15157         Add window.internals to WebKit2's WebKitTestRunner.
15158         https://bugs.webkit.org/show_bug.cgi?id=61073
15159
15160         Test: fast/harness/internals-object.html
15161
15162         * WebKitTestRunner/Configurations/InjectedBundleCFLite.vsprops:
15163         * WebKitTestRunner/Configurations/InjectedBundleCommon.vsprops:
15164         * WebKitTestRunner/Configurations/InjectedBundleCoreFoundation.vsprops:
15165         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
15166         (WTR::InjectedBundlePage::didClearWindowForFrame): initialize window.internals
15167         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
15168
15169 2011-06-12  Sheriff Bot  <webkit.review.bot@gmail.com>
15170
15171         Unreviewed, rolling out r88616.
15172         http://trac.webkit.org/changeset/88616
15173         https://bugs.webkit.org/show_bug.cgi?id=62517
15174
15175         It broke editing/text-iterator/findString.html (Requested by
15176         tonikitoo on #webkit).
15177
15178         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
15179         * DumpRenderTree/qt/LayoutTestControllerQt.h:
15180
15181 2011-06-12  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
15182
15183         Reviewed by Antonio Gomes.
15184
15185         [Qt] LayoutTestController needs setTextDirection implementation
15186         https://bugs.webkit.org/show_bug.cgi?id=62442
15187
15188         Implements LayoutController.setTextDirection.
15189
15190         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
15191         (LayoutTestController::setTextDirection):
15192         * DumpRenderTree/qt/LayoutTestControllerQt.h:
15193
15194 2011-06-10  Ryosuke Niwa  <rniwa@webkit.org>
15195
15196         Reviewed by Ojan Vafai.
15197
15198         new-run-webkit-tests: 'images' links on results.html don't work
15199         https://bugs.webkit.org/show_bug.cgi?id=62194
15200
15201         Use relative path when generating images-diffs.html
15202
15203         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
15204         Added _output_testname.
15205
15206 2011-06-10  Lucas Forschler  <lforschler@apple.com>
15207
15208         Reviewed by Stephanie Lewis.
15209
15210         Script to kill old processes on Mac.
15211         This matches the windows version.
15212         https://bugs.webkit.org/show_bug.cgi?id=62479
15213
15214         * BuildSlaveSupport/mac: Added.
15215         * BuildSlaveSupport/mac/kill-old-processes: Added.
15216
15217 2011-06-10  Adam Barth  <abarth@webkit.org>
15218
15219         Reviewed by James Robinson.
15220
15221         When sheriffbot reports failures on IRC instead of saying "(and more..)" it should say the number of tests that are failing
15222         https://bugs.webkit.org/show_bug.cgi?id=62489
15223
15224         * Scripts/webkitpy/tool/bot/sheriff.py:
15225         * Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
15226
15227 2011-06-10  Tony Chang  <tony@chromium.org>
15228
15229         Reviewed by Ojan Vafai.
15230
15231         add a compile guard ENABLE(FLEXBOX)
15232         https://bugs.webkit.org/show_bug.cgi?id=62049
15233
15234         * Scripts/build-webkit:
15235
15236 2011-06-10  Mark Rowe  <mrowe@apple.com>
15237
15238         Reviewed by Dimitri Glazkov.
15239
15240         <rdar://problem/9562114> Fix DumpRenderTree build for production configuration.
15241
15242         Ensure that DumpRenderTree can find libWebCoreTestSupport.dylib and the associated
15243         header file.
15244
15245         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
15246         * DumpRenderTree/mac/Configurations/Base.xcconfig:
15247
15248 2011-06-09  Csaba Osztrogonác  <ossy@webkit.org>
15249
15250         Unreviewed rolling out r88471, because it broke plugin tests on Qt.
15251
15252         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
15253         (handleEventX11):
15254         (NPP_GetValue):
15255         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
15256         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp: Added.
15257         (webkit_test_plugin_new_instance):
15258         (webkit_test_plugin_destroy_instance):
15259         (webkit_test_plugin_set_window):
15260         (executeScript):
15261         (webkit_test_plugin_new_stream):
15262         (webkit_test_plugin_destroy_stream):
15263         (webkit_test_plugin_stream_as_file):
15264         (webkit_test_plugin_write_ready):
15265         (webkit_test_plugin_write):
15266         (webkit_test_plugin_print):
15267         (keyEventToChar):
15268         (webkit_test_plugin_handle_event):
15269         (webkit_test_plugin_url_notify):
15270         (webkit_test_plugin_get_value):
15271         (webkit_test_plugin_set_value):
15272         (NP_GetMIMEDescription):
15273         (NP_Initialize):
15274         (NP_Shutdown):
15275         (NP_GetValue):
15276         * GNUmakefile.am:
15277
15278 2011-06-09  Martin Robinson  <mrobinson@igalia.com>
15279
15280         Reviewed by Eric Seidel.
15281
15282         [GTK] Implement the next bit of the TextInputController (hasMarkedText and markedRange)
15283         https://bugs.webkit.org/show_bug.cgi?id=55603
15284
15285         Add callbacks to support hasMarkedText and markedRange and also clean up existing
15286         TextInputController support.
15287
15288         * DumpRenderTree/gtk/TextInputController.cpp:
15289         (setMarkedTextCallback): Accept parameters in terms of start and length, instead
15290         of start and end. This matches the other ports. Do not use GLib style early returns.
15291         (hasMarkedTextCallback): Added.
15292         (markedRangeCallback): Added.
15293         (insertTextCallback): Cleanup as described above.
15294         (unmarkTextCallback): Cleanup as described above.
15295         (firstRectForCharacterRangeCallback): Cleanup as described above.
15296         (selectedRangeCallback): Cleanup as described above.
15297
15298 2011-06-09  Dirk Pranke  <dpranke@chromium.org>
15299
15300         Reviewed by Tony Chang.
15301
15302         nrwt: support webaudio in chromium driver
15303         https://bugs.webkit.org/show_bug.cgi?id=62226
15304
15305         * Scripts/webkitpy/layout_tests/port/chromium.py:
15306
15307 2011-06-09  Dan Bernstein  <mitz@apple.com>
15308
15309         Reviewed by Anders Carlsson.
15310
15311         Added a test for Vector::reverse()
15312         https://bugs.webkit.org/show_bug.cgi?id=62393
15313
15314         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
15315         * TestWebKitAPI/Tests/WTF/VectorReverse.cpp: Added.
15316         (TestWebKitAPI::TEST):
15317         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
15318
15319 2011-06-09  Martin Robinson  <mrobinson@igalia.com>
15320
15321         Reviewed by Andreas Kling.
15322
15323         [GTK] [Qt] Eliminate duplicate TestNetscapePlugin implementation
15324         https://bugs.webkit.org/show_bug.cgi?id=62385
15325
15326         Remove duplicate TestNetscapePlugIn implementation. Having two copies
15327         of this code makes keeping it in sync much more difficult. The files are
15328         almost identical and this change ports the X11 changes to main.cpp.
15329
15330         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
15331         (keyEventToChar): Added this helper which converts a X11 keycode into a char.
15332         (handleEventX11): Use the handler to properly convert the keycode. Do not print
15333         adjustCursorEvent output. The tests currently do not trigger this for Cocoa and thus
15334         it should not be in the results.
15335         (NPP_GetValue): Properly handle NPPVpluginNeedsXEmbed.
15336         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro: Update source list.
15337         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp: Removed.
15338         * GNUmakefile.am: Update source list.
15339
15340 2011-06-09  Noel Gordon  <noel.gordon@gmail.com>
15341
15342         Reviewed by Tony Chang.
15343
15344         Test plugin should support event logging on the windows port.
15345         https://bugs.webkit.org/show_bug.cgi?id=61721
15346
15347         The WebKit win port interprets a 0 return from plugin NPP_HandleEvent()
15348         calls to mean the plugin handled the event.
15349
15350         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
15351         (handleEventWin): Return 0 for each event on windows.
15352
15353 2011-06-09  Chang Shu  <cshu@webkit.org>
15354
15355         Reviewed by Andreas Kling.
15356
15357         [Qt] [WK2] Many editing tests failed on missing shouldBeginEditingInDOMRange:range printout
15358         https://bugs.webkit.org/show_bug.cgi?id=62381
15359
15360         Activate QGraphicsScene and setFocus on QGraphicsItem to make sure the view has the focus.
15361         QWKPagePrivate::isViewFocused() will check it later.
15362
15363         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
15364         (WTR::PlatformWebView::PlatformWebView):
15365
15366 2011-06-09  Robert Hogan  <robert@webkit.org>
15367
15368         Reviewed by Andreas Kling.
15369
15370         Teach Qt about window.internals
15371         https://bugs.webkit.org/show_bug.cgi?id=61074
15372
15373         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
15374         (WebCore::DumpRenderTree::initJSObjects):
15375
15376 2011-06-09  Eric Seidel  <eric@webkit.org>
15377
15378         Reviewed by Adam Barth.
15379
15380         Disable running layout tests on Mac EWS bots (and fix cr-linux bots to run them again)
15381         https://bugs.webkit.org/show_bug.cgi?id=62362
15382
15383         The mac bots are failing inspector tests.  We don't know why yet.
15384         So for now, turning run-webkit-tests off for the mac-ews.
15385
15386         Also noticed that due to the wrong variable name the cr-linux
15387         bots weren't running tests!
15388
15389         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
15390
15391 2011-06-09  Zoltan Herczeg  <zherczeg@inf.u-szeged.hu>
15392
15393         Reviewed by Csaba Osztrogonac.
15394
15395         [Qt][WK2] Canvas tests make css2.1 ones failing
15396         https://bugs.webkit.org/show_bug.cgi?id=53427
15397
15398         Adding a workaround for this issue by mimicing the behaviour of DumpRenderTree.
15399         Since the Qt port should eventually go to QRawFont, this issue will be resolved
15400         in the future. Both clearMemoryCaches() and removeAllApplicationFonts() are
15401         needed for the workaround.
15402
15403         * WebKitTestRunner/InjectedBundle/qt/LayoutTestControllerQt.cpp:
15404         (WTR::LayoutTestController::platformInitialize):
15405
15406 2011-06-09  Mario Sanchez Prada  <msanchez@igalia.com>
15407
15408         Reviewed by Chris Fleizach.
15409
15410         [GTK] Provide a way in DRT to check the platform name
15411         https://bugs.webkit.org/show_bug.cgi?id=56855
15412
15413         Patch for adding 'platformName' static property to DRT's
15414         LayoutController and provide platform specific implementations.
15415
15416         In principle, this feature will be only of interest for writing
15417         accessibility tests, so that's why it's only implemented in the
15418         Mac, Gtk and Win platforms.
15419
15420         * DumpRenderTree/LayoutTestController.h: New method platformName().
15421         * DumpRenderTree/LayoutTestController.cpp:
15422         (getPlatformNameCallback): New callback for 'platformName'.
15423         (LayoutTestController::staticValues): New static property.
15424
15425         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
15426         (LayoutTestController::platformName): Implemented.
15427
15428         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
15429         (LayoutTestController::platformName): Implemented.
15430
15431         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
15432         (LayoutTestController::platformName): Implemented.
15433
15434 2011-06-08  Mike Lawther  <mikelawther@chromium.org>
15435
15436         Reviewed by James Robinson.
15437
15438         Add mikelawther to committers.py
15439         https://bugs.webkit.org/show_bug.cgi?id=62354
15440
15441         * Scripts/webkitpy/common/config/committers.py:
15442
15443 2011-06-08  Kent Tamura  <tkent@chromium.org>
15444
15445         [Chromium] Fix a typo.
15446
15447         * DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
15448         (WebThemeControlDRTWin::draw):
15449
15450 2011-06-08  Kent Tamura  <tkent@chromium.org>
15451
15452         Reviewed by Hajime Morita.
15453
15454         [Chromium] Support vertical sliders in Windows DRT
15455         https://bugs.webkit.org/show_bug.cgi?id=62352
15456
15457         * DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
15458         (WebThemeControlDRTWin::draw):
15459           Paint VerticalSliderThumbType and VerticalSliderTrackType.
15460         * DumpRenderTree/chromium/WebThemeControlDRTWin.h:
15461           Add VerticalSliderTrackType and VerticalSliderThumbType.
15462         * DumpRenderTree/chromium/WebThemeEngineDRTWin.cpp:
15463         (WebThemeEngineDRTWin::paintTrackbar):
15464           Convert Windows identifiers to the internal identifiers for vertical sliders.
15465
15466 2011-06-08  Ojan Vafai  <ojan@chromium.org>
15467
15468         Reviewed by Tony Chang.
15469
15470         add a short-flag for --experimental-fully-parallel
15471         https://bugs.webkit.org/show_bug.cgi?id=62321
15472
15473         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
15474
15475 2011-06-08  Ryosuke Niwa  <rniwa@webkit.org>
15476
15477         Reviewed by Tony Chang.
15478
15479         new-run-webkit-tests: don't turn the tree red when there are new passes
15480         https://bugs.webkit.org/show_bug.cgi?id=62311
15481
15482         Look for "new passes" instead of "passed unexpectedly" in evaluateCommand
15483         to match commandComplete.
15484
15485         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
15486
15487 2011-06-08  Tom Hudson  <tomhudson@google.com>
15488
15489         Reviewed by Mihai Parparita.
15490
15491         Allow sorting in RebaselineServer based on 'metric' field in unexpected_results.json
15492         https://bugs.webkit.org/show_bug.cgi?id=60964
15493
15494         * Scripts/webkitpy/tool/commands/data/rebaselineserver/index.html:
15495         Add 'Sort tests by metric' link.
15496         * Scripts/webkitpy/tool/commands/data/rebaselineserver/main.js:
15497         (disableSorting): Activate 'Sort tests by metric' link.
15498         (enableSorting): Deactivate 'Sort tests by metric' link.
15499         (selectDirectory): Call enableSorting()/disableSorting() depending
15500         on currently selected failure type, and sort tests if requested.
15501         * Scripts/webkitpy/tool/commands/data/rebaselineserver/main.css:
15502         New .disabled-control class for deactivated links.
15503
15504 2011-06-08  Tom Sepez  <tsepez@chromium.org>
15505
15506         Reviewed by Adam Barth.
15507
15508         Add web permissions callbacks for mixed content blocking.
15509         https://bugs.webkit.org/show_bug.cgi?id=61946
15510
15511         * DumpRenderTree/chromium/LayoutTestController.cpp:
15512         (LayoutTestController::setAllowDisplayOfInsecureContent):
15513         (LayoutTestController::setAllowRunningOfInsecureContent):
15514         (LayoutTestController::overridePreference):
15515         * DumpRenderTree/chromium/WebPermissions.cpp:
15516         (WebPermissions::allowDisplayingInsecureContent):
15517         (WebPermissions::allowRunningInsecureContent):
15518         (WebPermissions::setDisplayingInsecureContentAllowed):
15519         (WebPermissions::setRunningInsecureContentAllowed):
15520         (WebPermissions::reset):
15521         * DumpRenderTree/chromium/WebPermissions.h:
15522
15523 2011-06-08  Dirk Pranke  <dpranke@chromium.org>
15524
15525         Reviewed by Tony Chang.
15526
15527         webkitpy: fix typo causing us to skip the port unit tests in chromium_linux
15528         https://bugs.webkit.org/show_bug.cgi?id=62259
15529
15530         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
15531
15532 2011-06-08  Ryosuke Niwa  <rniwa@webkit.org>
15533
15534         Reviewed by Tony Chang.
15535
15536         new-run-webkit-tests: master.cfg errors when nrwt successfully runs test
15537         https://bugs.webkit.org/show_bug.cgi?id=62303
15538
15539         Access self.incorrectLayoutLines instead of incorrectLayoutLines in evaluateCommand.
15540
15541         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
15542
15543 2011-06-08  Greg Simon  <gregsimon@chromium.org>
15544
15545         Reviewed by Dimitri Glazkov.
15546
15547         Control Indexeddb backends from LayoutTestController
15548         https://bugs.webkit.org/show_bug.cgi?id=61000
15549
15550         * DumpRenderTree/chromium/LayoutTestController.cpp:
15551         (LayoutTestController::LayoutTestController):
15552         (LayoutTestController::setOverrideIndexedDBBackingStore):
15553         (LayoutTestController::clearAllDatabases):
15554         * DumpRenderTree/chromium/LayoutTestController.h:
15555
15556 2011-06-08  Andreas Kling  <kling@webkit.org>
15557
15558         Add Pierre Rossi to contributors list for EZ-CC.
15559
15560         * Scripts/webkitpy/common/config/committers.py:
15561
15562 2011-06-08  Ilya Sherman  <isherman@chromium.org>
15563
15564         Reviewed by Andreas Kling.
15565
15566         Add Ilya Sherman to list of committers
15567         https://bugs.webkit.org/show_bug.cgi?id=62270
15568
15569         * Scripts/webkitpy/common/config/committers.py:
15570
15571 2011-06-07  Ryosuke Niwa  <rniwa@webkit.org>
15572
15573         Reviewed by Tony Chang.
15574
15575         new-run-webkit-tests: Bot master should print useful information on waterfall/console for nrwt
15576         https://bugs.webkit.org/show_bug.cgi?id=62178
15577
15578         Added commandComplete and evaluateCommand for NewRunWebKitTests class
15579         to parse new-run-webkit-tests' results.
15580
15581         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
15582
15583 2011-06-07  Dirk Pranke  <dpranke@chromium.org>
15584
15585         Reviewed by Eric Seidel.
15586
15587         webkitpy: add a popen() call to executive
15588         https://bugs.webkit.org/show_bug.cgi?id=62179
15589
15590         This change adds a popen() wrapper call to the Executive object.
15591         This will allow other webkitpy scripts that currently call
15592         subprocess directly to use Executive, instead.
15593
15594         No additional tests are necessary (the existing tests cover the code).
15595
15596         * Scripts/webkitpy/common/system/executive.py:
15597
15598 2011-06-07  Sam Weinig  <sam@webkit.org>
15599
15600         Reviewed by Anders Carlsson.
15601
15602         Remove repeated IPC traffic from SetRenderTreeSize message
15603         https://bugs.webkit.org/show_bug.cgi?id=62244
15604
15605         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
15606         (WTR::InjectedBundlePage::InjectedBundlePage):
15607         Add stub for didLayoutForFrame.
15608
15609 2011-06-07  Dirk Pranke  <dpranke@chromium.org>
15610
15611         Reviewed by Tony Chang.
15612
15613         rebaseline-chromium-webkit-tests: fix baselining order for linux, linux_x86_64
15614         https://bugs.webkit.org/show_bug.cgi?id=62241
15615
15616         * Scripts/webkitpy/layout_tests/port/chromium.py:
15617
15618 2011-06-07  Adam Barth  <abarth@webkit.org>
15619
15620         Rubber-stamped by Eric Seidel.
15621
15622         Enable tests on the mac-ews
15623
15624         When we bring the mac-ews back up, we're going to try running tests!
15625
15626         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
15627
15628 2011-06-07  Dmitry Lomov  <dslomov@google.com>
15629
15630         Reviewed by David Levin.
15631
15632         https://bugs.webkit.org/show_bug.cgi?id=62215
15633         Allow comparisons with 0 in LIKELY and UNLIKELY macros.
15634
15635         * Scripts/webkitpy/style/checkers/cpp.py:
15636         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
15637
15638 2011-06-07  Nico Weber  <thakis@chromium.org>
15639
15640         Reviewed by Dimitri Glazkov.
15641
15642         [chromium] -Wdelete-non-virtual-dtor pass for DumpRenderTree
15643         https://bugs.webkit.org/show_bug.cgi?id=62210
15644
15645         The change to TestEventPrinter fixes a latent bug, because
15646         objects are deleted through the TestEventPrinter type, but none
15647         of the subclasses have destructors or non-POD members.
15648
15649         The changes to NotificationPresenter and WebViewHost do _not_ to fix a
15650         real bug, they just make clang's -Wdelete-non-virtual-dtor happy. As
15651         discussed at http://codereview.chromium.org/7094005/, we prefer making
15652         leaf class destructors virtual over making the leaf classes final.
15653
15654         * DumpRenderTree/chromium/NotificationPresenter.cpp:
15655         (NotificationPresenter::~NotificationPresenter):
15656         * DumpRenderTree/chromium/NotificationPresenter.h:
15657         * DumpRenderTree/chromium/TestEventPrinter.cpp:
15658         (TestEventPrinter::~TestEventPrinter):
15659         * DumpRenderTree/chromium/TestEventPrinter.h:
15660         * DumpRenderTree/chromium/WebViewHost.h:
15661
15662 2011-06-06  Ryosuke Niwa  <rniwa@webkit.org>
15663
15664         Reviewed by Dirk Pranke.
15665
15666         [chromium] Chromium bots on build.webkit.org don't upload results.html
15667         https://bugs.webkit.org/show_bug.cgi?id=61961
15668
15669         Don't fix the path on chromium port erroneously.
15670
15671         * BuildSlaveSupport/test-result-archive:
15672
15673 2011-06-06  Martin Robinson  <mrobinson@igalia.com>
15674
15675         Fix the GTK+ build by ensuring that the autogenerated sources
15676         necessary for libWebCoreInternals are recorded in a predeclared
15677         variable before assigning to BUILT_SOURCES.
15678
15679         * GNUmakefile.am: Predeclare libwebcoreinternals_built_sources.
15680
15681 2011-06-06  Yong Li  <yoli@rim.com>
15682
15683         Reviewed by Eric Seidel.
15684
15685         https://bugs.webkit.org/show_bug.cgi?id=62097
15686         Fix the problem that integral bitfield check incorrectly barked at ?: expressions.
15687
15688         * Scripts/webkitpy/style/checkers/cpp.py:
15689         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
15690
15691 2011-06-06  Xan Lopez  <xlopez@igalia.com>
15692
15693         Reviewed by Martin Robinson.
15694
15695         [GTK] Enable feature/symbol detection in NWRT/GTK
15696         https://bugs.webkit.org/show_bug.cgi?id=62136
15697
15698         Disable feature detection through DRT, since we don't support it,
15699         and set the libwebcore patch correctly so that symbol detection
15700         through 'nm' works.
15701
15702         * Scripts/webkitpy/layout_tests/port/gtk.py:
15703
15704 2011-06-06  Shishir Agrawal  <shishir@chromium.org>
15705
15706         Reviewed by Tony Gentilcore.
15707
15708         Add a "prerender" state for page visibility.
15709         https://bugs.webkit.org/show_bug.cgi?id=62062
15710
15711         * DumpRenderTree/chromium/LayoutTestController.cpp:
15712         (LayoutTestController::setPageVisibility):
15713
15714 2011-06-06  Csaba Osztrogonác  <ossy@webkit.org>
15715
15716         [Qt] Implement download feature for QtTestBrowser
15717         https://bugs.webkit.org/show_bug.cgi?id=61865
15718
15719         Buildfix for --minimal and CONFIG+=qt_minimal build after r88161.
15720
15721         * QtTestBrowser/launcherwindow.cpp:
15722         (LauncherWindow::LauncherWindow):
15723         * QtTestBrowser/launcherwindow.h:
15724
15725 2011-06-06  Barát Tibor  <Barat.Tibor@stud.u-szeged.hu>
15726
15727         Reviewed by Csaba Osztrogonác.
15728
15729         [Qt] Implement download feature for QtTestBrowser
15730         https://bugs.webkit.org/show_bug.cgi?id=61865
15731
15732         * QtTestBrowser/launcherwindow.cpp:
15733         (LauncherWindow::LauncherWindow):
15734         (LauncherWindow::downloadRequest):
15735         (LauncherWindow::fileDownloadFinished):
15736         * QtTestBrowser/launcherwindow.h:
15737
15738 2011-06-04  Dominic Cooney  <dominicc@chromium.org>
15739
15740         Reviewed by Dimitri Glazkov.
15741
15742         Make DumpRenderTree initialize window.internals.
15743         https://bugs.webkit.org/show_bug.cgi?id=61076
15744
15745         * DumpRenderTree/win/DumpRenderTreeCommon.vsprops: link WebCoreTestSupport.
15746         * DumpRenderTree/win/FrameLoadDelegate.cpp:
15747         (FrameLoadDelegate::didClearWindowObjectForFrameInStandardWorld):
15748
15749 2011-06-06  Alexandru Chiculita  <achicu@adobe.com>
15750
15751         Reviewed by Kent Tamura.
15752
15753         Add ENABLE_CSS_EXCLUSIONS support for build-webkit script
15754         https://bugs.webkit.org/show_bug.cgi?id=61628
15755
15756         * Scripts/build-webkit:
15757
15758 2011-06-06  Mihnea Ovidenie  <mihnea@adobe.com>
15759
15760         Reviewed by Kent Tamura.
15761
15762         Add ENABLE(CSS_REGIONS) guard for CSS Regions support
15763         https://bugs.webkit.org/show_bug.cgi?id=61631
15764
15765         * Scripts/build-webkit:
15766
15767 2011-06-06  Carlos Garcia Campos  <cgarcia@igalia.com>
15768
15769         Unreviewed. Fix the GTK+ build.
15770
15771         * GNUmakefile.am: Use javascriptcore_cflags when building libWebCoreInternals.
15772
15773 2011-06-05  Alexey Proskuryakov  <ap@apple.com>
15774
15775         Add an e-mail address for Cameron McCormack that matches Bugzilla (so that auto-completion
15776         works properly).
15777
15778         * Scripts/webkitpy/common/config/committers.py:
15779
15780 2011-06-04  Kevin Ollivier  <kevino@theolliviers.com>
15781
15782         [wx] Unreviewed build fix. Update git branch build dir handling to match what other
15783         WebKit scripts expect, so other scripts work properly with wx.
15784
15785         * waf/build/build_utils.py:
15786         * waf/build/settings.py:
15787
15788 2011-06-04  Kevin Ollivier  <kevino@theolliviers.com>
15789
15790         [wx] Unreviewed build fix. Restore the PPC build and allow users to specify architectures
15791         to build on Mac.
15792
15793         * waf/build/settings.py:
15794         * wx/install-unix-extras:
15795
15796 2011-06-04  Kevin Ollivier  <kevino@theolliviers.com>
15797
15798         [wx] Unreviewed build fixes for recent trunk changes.
15799
15800         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
15801         (LayoutTestController::setTextDirection):
15802         * waf/build/settings.py:
15803
15804 2011-06-04  Martin Robinson  <mrobinson@igalia.com>
15805
15806         Fix the GTK+ build.
15807
15808         * GNUmakefile.am: Add some missing newline escape.
15809
15810 2011-06-03  Martin Robinson  <mrobinson@igalia.com>
15811
15812         Try to fix the GTK+ bulid.
15813
15814         * GNUmakefile.am: Add JSDOMWrapper.cpp to the libWebCoreInternals
15815         source list.
15816
15817 2011-06-03  Dirk Pranke  <dpranke@chromium.org>
15818
15819         Reviewed by Adam Barth.
15820
15821         nrwt: should skip chromium test expectation overrides on upstream bots
15822         https://bugs.webkit.org/show_bug.cgi?id=61888
15823
15824         * Scripts/webkitpy/layout_tests/port/chromium.py:
15825         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
15826
15827 2011-06-03  Dirk Pranke  <dpranke@chromium.org>
15828
15829         Reviewed by Ojan Vafai.
15830
15831         new-run-webkit-tests: minor logging, code cleanup
15832         https://bugs.webkit.org/show_bug.cgi?id=62055
15833
15834         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
15835         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
15836
15837 2011-06-03  Martin Robinson  <mrobinson@igalia.com>
15838
15839         Try to fix the build failure on the GTK+ 32-bit Release bot.
15840
15841         * GNUmakefile.am: Add JSInternal.{cpp,h} to the BUILT_SOURCES list.
15842
15843 2011-05-31  Martin Robinson  <mrobinson@igalia.com>
15844
15845         Reviewed by Ryosuke Niwa.
15846
15847         [GTK] Support smart replace for the pasteboard
15848         https://bugs.webkit.org/show_bug.cgi?id=61734
15849
15850         Call into the DumpRenderTreeSupportGtk hooks to toggle smart replace.
15851         This is required to produce consistent test results.
15852
15853         * DumpRenderTree/gtk/DumpRenderTree.cpp:
15854         (resetDefaultsToConsistentValues): Set the smart replace setting to on by default.
15855         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
15856         (LayoutTestController::setSmartInsertDeleteEnabled): Added an implementation of this
15857         methods which calls into DumpRenderTreeSupportGtk.
15858
15859 2011-06-03  Martin Robinson  <mrobinson@igalia.com>
15860
15861         Reviewed by Dimitri Glazkov.
15862
15863         Teach Gtk build about window.internals
15864         https://bugs.webkit.org/show_bug.cgi?id=61071
15865
15866         * DumpRenderTree/gtk/DumpRenderTree.cpp:
15867         (webViewWindowObjectCleared): Inject the new internals object when the
15868         window object clears.
15869         * GNUmakefile.am: Add new source files and includes necessary for window.internals.
15870         Also do a little bit of cleanup.
15871
15872 2011-06-03  Dirk Pranke  <dpranke@chromium.org>
15873
15874         Reviewed by Tony Chang.
15875
15876         Worker may not be stopped after LT.
15877         https://bugs.webkit.org/show_bug.cgi?id=61847
15878
15879         Re-submit a variant of the first fix for this bug. It's possible
15880         that actually calling cleanup() in the worker threads was taking
15881         too long and we triggered the assertions in the manager thread.
15882
15883         I've bumped up the timeout in the manager and added better
15884         diagnostics. I also cleaned up the cleanup() code a bit because
15885         there was some unnecessary if-checks.
15886
15887         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
15888         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
15889
15890 2011-06-03  Dirk Pranke  <dpranke@chromium.org>
15891
15892         Reviewed by Tony Chang.
15893
15894         nrwt: fix duplicate logging regression
15895         https://bugs.webkit.org/show_bug.cgi?id=61981
15896
15897         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
15898
15899 2011-06-03  Robert Kroeger  <rjkroege@chromium.org>
15900
15901         Reviewed by Adam Barth.
15902
15903         Modified addTouchPoint to correctly re-use touchPoint id values starting
15904         from the lowest not-un-use touchPoint id as is done by the embedder
15905         APIs. (Fixes BUGCR36415)
15906
15907         https://bugs.webkit.org/show_bug.cgi?id=62016
15908
15909         * DumpRenderTree/chromium/EventSender.cpp:
15910         (EventSender::addTouchPoint):
15911
15912 2011-06-03  Cary Clark  <caryclark@google.com>
15913
15914         Reviewed by Eric Seidel.
15915
15916         Draw DRT scroll bar thumb with CG when WebKit renders with Skia
15917         https://bugs.webkit.org/show_bug.cgi?id=61534
15918
15919         Skia on Mac uses Skia to render WebKit, and CG
15920         to render UI elements. The CG elements need a
15921         transcribed graphics context that preserves the
15922         canvas matrix, and the canvas clip.
15923
15924         The SkiaBitLocker utility class sets up a CGContext
15925         from the SkCanvas, locks the bitmap's bits, and
15926         releases the lock when the class goes out of scope.
15927
15928         The SkiaBitLocker implementation is described by
15929         this patch:
15930         http://codereview.chromium.org/7031006/
15931
15932         When WebKit uses Skia for rendering, the WebCanvas
15933         is an SkCanvas. Create an equivalent CGContext to
15934         draw the scrollbar thumb.
15935
15936         As the Chrome Mac port does not yet use Skia, this
15937         patch has no effect and requires no tests.
15938
15939         * DumpRenderTree/chromium/WebThemeEngineDRTMac.mm:
15940         (WebThemeEngineDRTMac::paintHIThemeScrollbarThumb):
15941
15942 2011-06-03  Andras Becsi  <abecsi@webkit.org>
15943
15944         Unreviewed build fix.
15945
15946         Fix typo.
15947
15948         * Scripts/old-run-webkit-tests:
15949
15950 2011-06-03  Andras Becsi  <abecsi@webkit.org>
15951
15952         Reviewed by Csaba Osztrogonác.
15953
15954         Print error message when there is no httpd present on the system
15955         https://bugs.webkit.org/show_bug.cgi?id=61939
15956
15957         * Scripts/old-run-webkit-tests: Propose to run with --no-httpd.
15958         * Scripts/webkitperl/httpd.pm:
15959
15960 2011-06-03  Mikhail Naganov  <mnaganov@chromium.org>
15961
15962         Reviewed by Yury Semikhatsky.
15963
15964         Web Inspector: [Chromium] cpu-profiler-profiling layout test is flaky on Linux Debug.
15965         https://bugs.webkit.org/show_bug.cgi?id=61533
15966
15967         Skip cpu-profiler-profiling in debug, implement a fast headless alternative.
15968
15969         * DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
15970         (DRTDevToolsAgent::setJavaScriptProfilingEnabled):
15971         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
15972         * DumpRenderTree/chromium/LayoutTestController.cpp:
15973         (LayoutTestController::LayoutTestController):
15974         (LayoutTestController::setJavaScriptProfilingEnabled):
15975         * DumpRenderTree/chromium/LayoutTestController.h:
15976         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
15977         (LayoutTestController::setJavaScriptProfilingEnabled):
15978         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
15979         (WTR::LayoutTestController::setJavaScriptProfilingEnabled):
15980         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
15981
15982 2011-06-02  Dirk Pranke  <dpranke@chromium.org>
15983
15984         Reviewed by Tony Chang.
15985
15986         NRWT needs a way to log to a file without including backspaces without needing --verbose
15987         https://bugs.webkit.org/show_bug.cgi?id=60328
15988
15989         This change modifies NRWT so that if it can tell if it is
15990         writing to a terminal, it will overwrite any messages printed
15991         in the 'one-line-progress' and 'updates' print options, and if
15992         not (or if --verbose was specified) the messages will be printed
15993         without backspaces.
15994
15995         If the overwriting updates are disabled, one-line-progress will
15996         only print every 10 seconds instead of on every update. This
15997         ensures that we still get updates once in a while while the
15998         tests are running, even if we're not in --verbose mode. Note
15999         that one-line-progress will now still be printed in --verbose,
16000         since it's useful as long as you're not flooded with the
16001         updates.
16002
16003         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
16004         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
16005         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
16006
16007 2011-06-02  Martin Robinson  <mrobinson@igalia.com>
16008
16009         Reviewed by Gustavo Noronha Silva.
16010
16011         [GTK] autogen.sh is run twice for each buld on the bots
16012         https://bugs.webkit.org/show_bug.cgi?id=61951
16013
16014         * Scripts/webkitdirs.pm: Only check the previous autotools arguments
16015         for the WebKit project. This prevents build-jsc runs from forcing subsequent
16016         build-webkit runs to re-rerun autogen.sh.
16017
16018 2011-06-02  Sheriff Bot  <webkit.review.bot@gmail.com>
16019
16020         Unreviewed, rolling out r87946.
16021         http://trac.webkit.org/changeset/87946
16022         https://bugs.webkit.org/show_bug.cgi?id=61965
16023
16024         Cause NRWT to break (Requested by abarth|gardener on #webkit).
16025
16026         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
16027
16028 2011-06-02  Dimitri Glazkov  <dglazkov@chromium.org>
16029
16030         Reviewed by Darin Adler.
16031
16032         Add build logistics and plumbing for window.internals object.
16033         https://bugs.webkit.org/show_bug.cgi?id=60313
16034
16035         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added linking new WebCoreTestSupport library.
16036         * DumpRenderTree/chromium/TestShell.cpp:
16037         (TestShell::bindJSObjectsToWindow): Added injection code.
16038         * DumpRenderTree/mac/FrameLoadDelegate.mm:
16039         (-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]): Ditto.
16040
16041 2011-06-02  Hao Zheng  <zhenghao@chromium.org>
16042
16043         Reviewed by Dirk Pranke.
16044
16045         Call cleanup() explicitly at the end of worker.run().
16046         https://bugs.webkit.org/show_bug.cgi?id=61847
16047
16048         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
16049
16050 2011-06-02  Dimitri Glazkov  <dglazkov@chromium.org>
16051
16052         Unreviewed, rolling out r87926.
16053         http://trac.webkit.org/changeset/87926
16054         https://bugs.webkit.org/show_bug.cgi?id=60313
16055
16056         Fails to find WebCoreTestSupport.dylib on bots.
16057
16058         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
16059         * DumpRenderTree/chromium/TestShell.cpp:
16060         (TestShell::bindJSObjectsToWindow):
16061         * DumpRenderTree/mac/FrameLoadDelegate.mm:
16062         (-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]):
16063
16064 2011-05-21  Dimitri Glazkov  <dglazkov@chromium.org>
16065
16066         Reviewed by Darin Adler.
16067
16068         Add build logistics and plumbing for window.internals object.
16069         https://bugs.webkit.org/show_bug.cgi?id=60313
16070
16071         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added linking new WebCoreTestSupport library.
16072         * DumpRenderTree/chromium/TestShell.cpp:
16073         (TestShell::bindJSObjectsToWindow): Added injection code.
16074         * DumpRenderTree/mac/FrameLoadDelegate.mm:
16075         (-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]): Ditto.
16076
16077 2011-06-02  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
16078
16079         Reviewed by Eric Seidel.
16080
16081         Adjust efl coding style to more efl directories
16082         https://bugs.webkit.org/show_bug.cgi?id=61903
16083
16084         * Scripts/webkitpy/style/checker.py: Adjust efl coding style to efl directories.
16085
16086 2011-06-01  Dirk Pranke  <dpranke@chromium.org>
16087
16088         Reviewed by Tony Chang.
16089
16090         old-run-webkit-tests: add support for audio files
16091         https://bugs.webkit.org/show_bug.cgi?id=57992
16092
16093         * Scripts/old-run-webkit-tests:
16094
16095 2011-06-01  Jochen Eisinger  <jochen@chromium.org>
16096
16097         Reviewed by Adam Barth.
16098
16099         Implement layoutTestController.setImagesAllowed for controlling the respective behavior of the WebPermissionClient
16100         https://bugs.webkit.org/show_bug.cgi?id=34314
16101
16102         * DumpRenderTree/DumpRenderTree.gypi:
16103         * DumpRenderTree/chromium/LayoutTestController.cpp:
16104         (LayoutTestController::LayoutTestController):
16105         (LayoutTestController::setImagesAllowed):
16106         (LayoutTestController::setStorageAllowed):
16107         * DumpRenderTree/chromium/LayoutTestController.h:
16108         * DumpRenderTree/chromium/WebPermissions.cpp: Copied from Tools/DumpRenderTree/chromium/WebPermissions.h.
16109         (WebPermissions::WebPermissions):
16110         (WebPermissions::~WebPermissions):
16111         (WebPermissions::allowImages):
16112         (WebPermissions::allowStorage):
16113         (WebPermissions::setImagesAllowed):
16114         (WebPermissions::setStorageAllowed):
16115         (WebPermissions::reset):
16116         * DumpRenderTree/chromium/WebPermissions.h:
16117
16118 2011-06-01  Noel Gordon  <noel.gordon@gmail.com>
16119
16120         Reviewed by Tony Chang.
16121
16122         Test plugin should support event logging on the windows port.
16123         https://bugs.webkit.org/show_bug.cgi?id=61721
16124
16125         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
16126         (handleEventWin): Well, handle them.
16127
16128 2011-06-01  Daniel Cheng  <dcheng@chromium.org>
16129
16130         Reviewed by Tony Chang.
16131
16132         [chromium] Use correct file name for dragging out images.
16133         https://bugs.webkit.org/show_bug.cgi?id=24887
16134
16135         * DumpRenderTree/chromium/EventSender.cpp:
16136         (EventSender::EventSender):
16137         (EventSender::dumpFilenameBeingDragged):
16138         * DumpRenderTree/chromium/EventSender.h:
16139
16140 2011-06-01  Dirk Pranke  <dpranke@chromium.org>
16141
16142         Reviewed by Tony Chang.
16143
16144         nrwt: make webaudio work w/o content-length header
16145         https://bugs.webkit.org/show_bug.cgi?id=61819
16146
16147         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
16148         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
16149         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
16150         * Scripts/webkitpy/layout_tests/port/webkit.py:
16151
16152 2011-06-01  Viatcheslav Ostapenko  <ostapenko.viatcheslav@nokia.com>
16153
16154         Reviewed by Andreas Kling.
16155
16156         [Qt][WK2] Add QGLWidget viewport support to MiniBrowser
16157         https://bugs.webkit.org/show_bug.cgi?id=61802
16158
16159         Add command line parameter and menu item to MiniBrowser application
16160         enabling use of QGLWidget for browser viewport.
16161
16162         * MiniBrowser/qt/BrowserWindow.cpp:
16163         (BrowserWindow::BrowserWindow):
16164         (BrowserWindow::toggleGLViewport):
16165         * MiniBrowser/qt/BrowserWindow.h:
16166         * MiniBrowser/qt/MiniBrowserApplication.cpp:
16167         (MiniBrowserApplication::handleUserOptions):
16168         * MiniBrowser/qt/MiniBrowserApplication.h:
16169         (WindowOptions::WindowOptions):
16170
16171 2011-06-01  Martin Robinson  <mrobinson@igalia.com>
16172
16173         Reviewed by Xan Lopez.
16174
16175         Print out autogen arguments when they differ from the last set, to help
16176         with debugging build issues.
16177
16178         * Scripts/webkitdirs.pm: Print out previous and current build arguments
16179         when they differ.
16180
16181 2011-06-01  Carlos Garcia Campos  <cgarcia@igalia.com>
16182
16183         Reviewed by Martin Robinson.
16184
16185         [GTK] run-launcher should run MiniBrowser in WebKit2
16186         https://bugs.webkit.org/show_bug.cgi?id=61858
16187
16188         * Scripts/run-launcher: Check whether -2 command line option has
16189         been passed and run MiniBrowser instead of GtkLauncher.
16190
16191 2011-05-31  Hao Zheng  <zhenghao@chromium.org>
16192
16193         Reviewed by Dirk Pranke.
16194
16195         Fix exception in verbose printing mode.
16196         https://bugs.webkit.org/show_bug.cgi?id=61723
16197
16198         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
16199
16200 2011-05-31  Yong Li  <yoli@rim.com>
16201
16202         Reviewed by Eric Seidel.
16203
16204         https://bugs.webkit.org/show_bug.cgi?id=54807
16205         We have been assuming plain bitfields (like "int a : 31") are always signed integers.
16206         However some compilers can treat them as unsigned. For example, RVCT 4.0 states plain
16207         bitfields (declared without either signed or unsigned qualifiers) are treats as unsigned.
16208         http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dui0348c/Babjddhe.html
16209         Although we can use "--signed-bitfields" flag to make RVCT 4.0 behave as most other compilers,
16210         always using "signed"/"unsigned" qualifier to declare integral type bitfields is still a good
16211         rule we should have in order to make our code independent from compilers and compiler flags.
16212
16213         * Scripts/webkitpy/style/checkers/cpp.py:
16214
16215 2011-05-31  Hironori Bono  <hbono@chromium.org>
16216
16217         Reviewed by Eric Seidel.
16218
16219         Add a JavaScript function layoutTestController.setTextDirection().
16220         https://bugs.webkit.org/show_bug.cgi?id=50952
16221
16222         This change adds a new JavaScript function setTextDirection() to the
16223         LayoutTestController class of Safari and Chrome to write layout tests
16224         that changes the text direction of an editable element.
16225
16226         * DumpRenderTree/LayoutTestController.cpp:
16227         (setTextDirectionCallback):
16228         (LayoutTestController::staticFunctions):
16229         * DumpRenderTree/LayoutTestController.h:
16230         * DumpRenderTree/chromium/LayoutTestController.cpp:
16231         (LayoutTestController::LayoutTestController):
16232         (LayoutTestController::setTextDirection):
16233         * DumpRenderTree/chromium/LayoutTestController.h:
16234         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
16235         (LayoutTestController::setTextDirection):
16236         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
16237         (LayoutTestController::setTextDirection):
16238         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
16239         (LayoutTestController::setTextDirection):
16240
16241 2011-05-31  Ademar de Souza Reis Jr  <ademar.reis@openbossa.org>
16242
16243         Reviewed by Simon Hausmann.
16244
16245         [Qt] User-agent list of Qt test browsers is very outdated
16246         https://bugs.webkit.org/show_bug.cgi?id=61531
16247
16248         The user can edit the user-agent manually, so there's no need to have
16249         a super complete and up to date list, but it's nice to have it
16250         minimally updated.
16251
16252         * MiniBrowser/qt/useragentlist.txt:
16253         * QtTestBrowser/useragentlist.txt:
16254
16255 2011-03-30  Martin Robinson  <mrobinson@igalia.com>
16256
16257         Reviewed by Adam Roben.
16258
16259         [GTK] [WebKit2] Implement a basic WebKitTestRunner
16260         https://bugs.webkit.org/show_bug.cgi?id=57068
16261
16262         Add an implementation of WebKitTestRunner for GTK+.
16263
16264         * Scripts/build-webkittestrunner: Added knowledge of GTK+ TestRunner.
16265         * Scripts/old-run-webkit-tests: Ditto.
16266         * Scripts/run-launcher: Ditto.
16267         * Scripts/webkitdirs.pm: Ditto.
16268         * WebKitTestRunner/GNUmakefile.am: Added.
16269         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
16270         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp: Added.
16271         * WebKitTestRunner/InjectedBundle/gtk/InjectedBundleGtk.cpp: Copied from Source/WebKit2/WebProcess/InjectedBundle/gtk/InjectedBundleGtk.cpp.
16272         * WebKitTestRunner/InjectedBundle/gtk/LayoutTestControllerGtk.cpp: Copied from Source/WebKit2/WebProcess/InjectedBundle/gtk/InjectedBundleGtk.cpp.
16273         * WebKitTestRunner/PlatformWebView.h:
16274         * WebKitTestRunner/gtk/PlatformWebViewGtk.cpp: Added.
16275         * WebKitTestRunner/gtk/TestControllerGtk.cpp: Added.
16276         * WebKitTestRunner/gtk/TestInvocationGtk.cpp: Copied from Source/WebKit2/WebProcess/InjectedBundle/gtk/InjectedBundleGtk.cpp.
16277         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
16278         * WebKitTestRunner/gtk/main.cpp: Copied from Source/WebKit2/WebProcess/InjectedBundle/gtk/InjectedBundleGtk.cpp.
16279         (main): Added.
16280
16281 2011-05-31  Dirk Pranke  <dpranke@chromium.org>
16282
16283         Reviewed by Tony Chang.
16284
16285         new-run-webkit-tests: doesn't wait for children if it gets a ctrl-c
16286         https://bugs.webkit.org/show_bug.cgi?id=60241
16287
16288         * Scripts/new-run-webkit-tests:
16289
16290 2011-05-31  Eric Seidel  <eric@webkit.org>
16291
16292         Reviewed by David Kilzer.
16293
16294         webkitpy should create zips with zip -9
16295         https://bugs.webkit.org/show_bug.cgi?id=61789
16296
16297         Dave Kilzer suggested we add this to make zips uploaded to bugzilla smaller.
16298
16299         * Scripts/webkitpy/common/system/workspace.py:
16300         * Scripts/webkitpy/common/system/workspace_unittest.py:
16301
16302 2011-05-31  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
16303
16304         Reviewed by Csaba Osztrogonác.
16305
16306         [Qt] Implement find feature for QtTestBrowser
16307         https://bugs.webkit.org/show_bug.cgi?id=61425
16308
16309         Buildfix for --minimal and CONFIG+=qt_minimal build.
16310
16311         * QtTestBrowser/launcherwindow.cpp:
16312         (LauncherWindow::LauncherWindow):
16313         (LauncherWindow::createChrome):
16314         * QtTestBrowser/launcherwindow.h:
16315
16316 2011-05-31  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
16317
16318         Reviewed by Csaba Osztrogonác.
16319
16320         [Qt] Implement find feature for QtTestBrowser
16321         https://bugs.webkit.org/show_bug.cgi?id=61425
16322
16323         * QtTestBrowser/launcherwindow.cpp:
16324         (LauncherWindow::LauncherWindow):
16325         (LauncherWindow::createChrome):
16326         (LauncherWindow::showFindBar):
16327         (LauncherWindow::find):
16328         * QtTestBrowser/launcherwindow.h:
16329
16330 2011-05-30  Carlos Garcia Campos  <cgarcia@igalia.com>
16331
16332         Reviewed by Martin Robinson.
16333
16334         [GTK] Implement JavaScript dialogs in MiniBrowser
16335         https://bugs.webkit.org/show_bug.cgi?id=61725
16336
16337         * MiniBrowser/gtk/BrowserWindow.c:
16338         (WKStringGetCString): Helper function to get a UTF-8 c string from
16339         a WKStringRef.
16340         (WKURLGetCString): Helper function to get a UTF-8 c string from a
16341         WKURLRef.
16342         (browserWindowUpdateURL): Update to use WKURLGetCString.
16343         (didReceiveTitleForFrame): Update to use WKStringGetCString.
16344         (createMessageDialog): Create a GtkMessageDialog.
16345         (runJavaScriptAlert):
16346         (runJavaScriptConfirm):
16347         (runJavaScriptPrompt):
16348         (browserWindowUIClientInit):
16349
16350 2011-05-29  Yuta Kitamura  <yutak@chromium.org>
16351
16352         Reviewed by Kent Tamura.
16353
16354         WebSocket closing handshake
16355         https://bugs.webkit.org/show_bug.cgi?id=35721
16356
16357         * Scripts/webkitpy/thirdparty/__init__.py:
16358         Pull in pywebsocket 0.6b1. We need to update pywebsocket
16359         to get the right behavior of closing handshake.
16360
16361 2011-05-29  Daniel Bates  <dbates@rim.com>
16362
16363         Reviewed by David Kilzer.
16364
16365         REGRESSION (r86515): svn-apply ignores diffs that omit line count in chunk range
16366         https://bugs.webkit.org/show_bug.cgi?id=61162
16367
16368         Fixes an issue where svn-apply may ignore a diff that contains a chunk range line
16369         that omits a line count. In particular, the chunk range regular expression does
16370         not match a chunk range line that omits a line count. GNU diff(1) will omit the
16371         line count in the chunk range if the line count is exactly 1. For example, appending
16372         a new line to the end of an existing file F that contains exactly one line of text will
16373         be represented in a diff with a chunk range line that omits the line count for F.
16374
16375         * Scripts/VCSUtils.pm:
16376           (parseChunkRange): Added.
16377         * Scripts/webkitperl/VCSUtils_unittest/parseChunkRange.pl: Added.
16378         * Scripts/webkitperl/VCSUtils_unittest/parseDiff.pl:
16379           - Added unit test "Git: Append new line to the end of an existing file".
16380
16381 2011-05-28  Adam Barth  <abarth@webkit.org>
16382
16383         Reviewed by Eric Seidel.
16384
16385         EWS builds patches that fail to build twice, which seems useless and slows down the bots
16386         https://bugs.webkit.org/show_bug.cgi?id=55585
16387
16388         This patch switches all the early warning system bots over to the new
16389         PatchAnalysisTask-based infrastructure.  This patch makes these bots
16390         more efficient (in the case where patches fail to build) and paves the
16391         way for running tests on these bots!
16392
16393         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
16394         * Scripts/webkitpy/tool/bot/earlywarningsystemtask.py:
16395         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
16396         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
16397         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
16398         * Scripts/webkitpy/tool/commands/queues.py:
16399
16400 2011-05-28  Kenichi Ishibashi  <bashi@chromium.org>
16401
16402         Reviewed by Kent Tamura.
16403
16404         Add bashi as a committer.
16405
16406         https://bugs.webkit.org/show_bug.cgi?id=61685
16407
16408         * Scripts/webkitpy/common/config/committers.py:
16409
16410 2011-05-27  Adam Barth  <abarth@webkit.org>
16411
16412         Reviewed by Eric Seidel.
16413
16414         When checking whether the tree is red, the EWS posts a link to the wrong log
16415         https://bugs.webkit.org/show_bug.cgi?id=61072
16416
16417         We need to cache the original script error because that contains the
16418         failure log we want to upload.  If we don't cache that script error,
16419         self._script_error will get overwritten when we sanity check the clean
16420         tree (and it also has test failures).
16421
16422         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
16423             - This testing approach is slightly goofy.  We'd like to use
16424               assertRaisesRegexp, but that's not available until Python 2.7.
16425         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
16426
16427 2011-05-27  Jochen Eisinger  <jochen@chromium.org>
16428
16429         Reviewed by Adam Barth.
16430
16431         Add layoutTestController.setStorageAllowed() to control whether access
16432         to the localStorage API is enabled via the WebPermissionClient
16433         https://bugs.webkit.org/show_bug.cgi?id=61581
16434
16435         * DumpRenderTree/chromium/LayoutTestController.cpp:
16436         (LayoutTestController::LayoutTestController):
16437         (LayoutTestController::setStorageAllowed):
16438         * DumpRenderTree/chromium/LayoutTestController.h:
16439         * DumpRenderTree/chromium/TestShell.cpp:
16440         (TestShell::TestShell):
16441         (TestShell::createNewWindow):
16442         * DumpRenderTree/chromium/TestShell.h:
16443         * DumpRenderTree/chromium/WebPermissions.h: Added.
16444         (WebPermissions::WebPermissions):
16445         (WebPermissions::allowStorage):
16446         (WebPermissions::setStorageAllowed):
16447         (WebPermissions::reset):
16448
16449 2011-05-27  Dirk Pranke  <dpranke@chromium.org>
16450
16451         Reviewed by Eric Seidel.
16452
16453         NRWT: clean up metered_stream code in preparation for 'nooverwriting' patch
16454         https://bugs.webkit.org/show_bug.cgi?id=60326
16455
16456         This patch removes a lot of the complexity from the
16457         metered_stream implementation that was unnecessary since there
16458         was only one caller and the logic could be coordinated better.
16459
16460         There should be no functional changes in this patch, just code
16461         getting deleted and cleaned up.
16462
16463         * Scripts/webkitpy/layout_tests/layout_package/metered_stream.py:
16464         * Scripts/webkitpy/layout_tests/layout_package/metered_stream_unittest.py:
16465         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
16466         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
16467
16468 2011-05-27  Dirk Pranke  <dpranke@chromium.org>
16469
16470         Reviewed by Eric Seidel.
16471
16472         NRWT: minor cleanup in printing module
16473         https://bugs.webkit.org/show_bug.cgi?id=60329
16474
16475         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
16476
16477 2011-05-27  Steve Lacey  <sjl@chromium.org>
16478
16479         Reviewed by David Levin.
16480
16481         Add sjl@chromium.org as a committer.
16482
16483         https://bugs.webkit.org/show_bug.cgi?id=61672
16484
16485         * Scripts/webkitpy/common/config/committers.py:
16486
16487 2011-05-27  Dirk Pranke  <dpranke@chromium.org>
16488
16489         Reviewed by Adam Barth.
16490
16491         NRWT: remove --print detailed-progress
16492         https://bugs.webkit.org/show_bug.cgi?id=60324
16493
16494         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
16495         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
16496         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
16497         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
16498
16499 2011-05-27  Dirk Pranke  <dpranke@chromium.org>
16500
16501         Reviewed by Ojan Vafai.
16502
16503         NRWT: debug messages from the workers are being logged twice
16504         https://bugs.webkit.org/show_bug.cgi?id=60428
16505
16506         It looks like when the workers are run in separate processes
16507         we end up getting two copies of every log message they print.
16508         This has to do with the multiprocessing module on UNIX cloning
16509         the log configuration in a way I wasn't expecting, and so two
16510         log handlers end up getting registered.
16511
16512         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
16513         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
16514         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
16515         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
16516
16517 2011-05-27  Adam Roben  <aroben@apple.com>
16518
16519         Always decode tester names from location.hash
16520
16521         When navigating directly to a tester-specific URL (e.g.,
16522         <http://build.webkit.org/TestFailures/#/Windows%207%20Release%20(Tests)>), you could end up
16523         with a URL-encoded tester name in the page, and then the encoded name could make its way
16524         into Bugzilla bugs. Bad!
16525
16526         Reviewed by David Kilzer.
16527
16528         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
16529         (ViewController.prototype.parseHash): Pass the tester name we got from location.hash through
16530         decodeURIComponent before trying to use it.
16531
16532 2011-05-27  Adam Roben  <aroben@apple.com>
16533
16534         Remove testing code that snuck into r87528
16535
16536         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
16537         (ViewController.prototype._domForNewBugLink): Remove a line I was using for some testing.
16538
16539 2011-05-27  Adam Roben  <aroben@apple.com>
16540
16541         Add links to TestFailures to easily file bugs about failing tests
16542
16543         Fixes <http://webkit.org/b/61636> <rdar://problem/9514886> TestFailures page should have
16544         links to file bugs for failing tests
16545
16546         Reviewed by David Kilzer.
16547
16548         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
16549         (Builder.prototype.resultsPageURL): New helper function to return the URL to results.html
16550         for a particular build.
16551         (Builder.prototype._getFailingTests): Use resultsPageURL.
16552
16553         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
16554         (ViewController.prototype._displayBuilder): Add links to file bugs about each set of
16555         failures.
16556         (ViewController.prototype._domForBuildName): Use resultsPageURL.
16557         (ViewController.prototype._domForNewBugLink): Added. Creates an element that represents the
16558         new bug link. The bug has a title, description, and URL already filled in for you!
16559
16560 2011-05-27  Sheriff Bot  <webkit.review.bot@gmail.com>
16561
16562         Unreviewed, rolling out r87464.
16563         http://trac.webkit.org/changeset/87464
16564         https://bugs.webkit.org/show_bug.cgi?id=61643
16565
16566         client-close/server-close tests are flaky on Chromium Mac and
16567         Windows (Requested by aklein on #webkit).
16568
16569         * Scripts/webkitpy/thirdparty/__init__.py:
16570
16571 2011-05-27  Alexey Proskuryakov  <ap@apple.com>
16572
16573         Added Douglas Davidson to contributors list for easier CC'ing.
16574
16575         * Scripts/webkitpy/common/config/committers.py:
16576
16577 2011-05-26  Mihai Parparita  <mihaip@chromium.org>
16578
16579         Reviewed by Adam Barth.
16580
16581         Fix worldID and destinationDomain argument names
16582         https://bugs.webkit.org/show_bug.cgi?id=61571
16583
16584         As part of working on r87423, I noticed a couple of inconsistencies in
16585         argument names:
16586         - We would use worldId in .h files but worldID in .cpp files.
16587           Standardize on the latter
16588         - SecurityOrigin::addOriginAccessWhitelistEntry would take a
16589           destinationDomains argument, even though the actual parameter was for
16590           a single domain (renamed to be singular).
16591
16592         * DumpRenderTree/LayoutTestController.h:
16593         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
16594         (LayoutTestController::evaluateScriptInIsolatedWorld):
16595
16596 2011-05-26  Yuta Kitamura  <yutak@chromium.org>
16597
16598         Reviewed by Kent Tamura.
16599
16600         WebSocket closing handshake
16601         https://bugs.webkit.org/show_bug.cgi?id=35721
16602
16603         * Scripts/webkitpy/thirdparty/__init__.py:
16604         Pull in pywebsocket 0.6b1. We need to update pywebsocket
16605         to get the right behavior of closing handshake.
16606
16607 2011-05-26  Qi Zhang  <qi.2.zhang@nokia.com>
16608
16609         Reviewed by Andreas Kling.
16610
16611         WebKitTestRunner needs layoutTestController.setAuthorAndUserStylesEnabled
16612         https://bugs.webkit.org/show_bug.cgi?id=42676
16613
16614         Implemented layoutTestController.setAuthorAndUserStylesEnabled for WebKitTestRunner.
16615
16616         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
16617         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
16618         (WTR::InjectedBundle::beginTesting):
16619         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
16620         (WTR::LayoutTestController::setAuthorAndUserStylesEnabled):
16621         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
16622
16623 2011-05-26  Ademar de Souza Reis Jr.  <ademar.reis@openbossa.org>
16624
16625         Reviewed by Andreas Kling.
16626
16627         [Qt] QtTestBrowser "Switch User Agent" dialog is broken
16628         https://bugs.webkit.org/show_bug.cgi?id=61527
16629
16630         The setMaximumHeight() call is unecessary and breaks the dialog at
16631         least on my setup.
16632
16633         * QtTestBrowser/launcherwindow.cpp:
16634         (LauncherWindow::showUserAgentDialog):
16635
16636 2011-05-02  Robert Hogan  <robert@webkit.org>
16637
16638         Reviewed by Adam Roben.
16639
16640         [Qt] Allow popup windows from plugins when initiated by a user gesture
16641
16642         https://bugs.webkit.org/show_bug.cgi?id=41292
16643
16644         Support parameters introduced in plugin-initiate-popup-window.html
16645
16646         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
16647         (webkit_test_plugin_new_instance):
16648         (webkit_test_plugin_handle_event):
16649
16650 2011-05-26  Mihai Parparita  <mihaip@chromium.org>
16651
16652         Reviewed by Adam Barth.
16653
16654         Support cross-origin XMLHttpRequest in isolated worlds
16655         https://bugs.webkit.org/show_bug.cgi?id=59843
16656
16657         Expose setIsolatedWorldSecurityOrigin in LayoutTestController.
16658
16659         * DumpRenderTree/chromium/LayoutTestController.cpp:
16660         (LayoutTestController::LayoutTestController):
16661         (LayoutTestController::setIsolatedWorldSecurityOrigin):
16662         * DumpRenderTree/chromium/LayoutTestController.h:
16663
16664 2011-05-25  Brian Weinstein  <bweinstein@apple.com>
16665
16666         Reviewed by Adam Roben.
16667
16668         WebKit2: Status bar, toolbar, and menu bar checks should be in the injected bundle
16669         https://bugs.webkit.org/show_bug.cgi?id=61474
16670         <rdar://problem/9468337>
16671
16672         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
16673         (WTR::InjectedBundlePage::InjectedBundlePage): Add empty entries in the WKBundlePageUIClient.
16674
16675 2011-05-26  Andreas Kling  <kling@webkit.org>
16676
16677         Unreviewed, correct address to webkit-committers mailing list.
16678
16679         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
16680
16681 2011-05-25  Adam Roben  <aroben@apple.com>
16682
16683         Don't assume that builds finish in the order they were started
16684
16685         Fixes <http://webkit.org/b/61464> TestFailures page shows failures from a not-most-recent
16686         build when one slave is taking a long time to finish a build
16687
16688         Reviewed by Antti Koivisto.
16689
16690         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
16691         (Builder.prototype.getMostRecentCompletedBuildNumber): Don't assume that all builds that are
16692         newer than the oldest in-progress build are also themselves in progress. Just check whether
16693         each build is actually in progress (as signified by its presence in the currentBuilds set).
16694
16695 2011-05-25  Adam Roben  <aroben@apple.com>
16696
16697         Correctly detect cases where only a single test failed/crashed/timed out
16698
16699         Fixes <http://webkit.org/b/61463> TestFailures page doesn't show testers with only a single
16700         failing test
16701
16702         Reviewed by Darin Adler.
16703
16704         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
16705         (Builder.prototype.getNumberOfFailingTests): Don't require "test cases" to be plural. It
16706         will be singular when only a single test fails/crashes/times out.
16707
16708         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
16709         (ViewController.prototype._displayTesters): Only pluralize "test" when more than one test
16710         failed/crashed/timed out.
16711
16712 2011-05-26  Tom Hudson  <tomhudson@google.com>
16713
16714         Reviewed by Tony Chang.
16715
16716         Add flags to Chromium ImageDiff to write image comparison metrics on
16717         stdout.
16718         https://bugs.webkit.org/show_bug.cgi?id=60569
16719
16720         * DumpRenderTree/chromium/ImageDiff.cpp:
16721         (Image::createFromStdin): Fix spelling.
16722         (maxOf3): New function to do 3-way maximum.
16723         (getRedComponent): Extract red component from ABGR packed encoding.
16724         (getGreenComponent): Extract green component from ABGR packed encoding.
16725         (getBlueComponent): Extract blue component from ABGR packed encoding.
16726         (weightedPercentageDifferent): Determine % of pixels different in two
16727         images multiplied by intensity difference of each pixel.
16728         (printHelp): Document new arguments.
16729         (compareImages): Parameterize.
16730         (untestedCompareImages): Parameterize.
16731         (diffImages): Parameterize; write results on stdout if requested.
16732         (main): Parse new arguments.
16733
16734 2011-05-25  Qi Zhang  <qi.2.zhang@nokia.com>
16735
16736         Reviewed by Simon Hausmann.
16737
16738         WebKitTestRunner needs layoutTestController.setPrivateBrowsingEnabled
16739         https://bugs.webkit.org/show_bug.cgi?id=42697
16740
16741         Implemented layoutTestController.setPrivateBrowsingEnabled for WebKitTestRunner.
16742
16743         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
16744         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
16745         (WTR::InjectedBundle::beginTesting):
16746         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
16747         (WTR::LayoutTestController::setPrivateBrowsingEnabled):
16748         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
16749
16750 2011-05-25  Jon Honeycutt  <jhoneycutt@apple.com>
16751
16752         REGRESSION (WebKit2): Crash in Flash on USA Today photo gallery
16753         https://bugs.webkit.org/show_bug.cgi?id=61428
16754         <rdar://problem/9457006>
16755
16756         Reviewed by Adam Roben.
16757
16758         The crash occurs when Flash posts a message to a window that it
16759         creates, and in processing the message, it calls NPN_Evaluate to
16760         evaluate JavaScript that removes the plug-in from the page. Flash then
16761         crashes when we return to Flash code.
16762
16763         This test emulates that behavior.
16764
16765         * DumpRenderTree/TestNetscapePlugIn/win/CallJSThatDestroysPlugin.cpp: Added.
16766         (CallJSThatDestroysPlugin::CallJSThatDestroysPlugin):
16767         Initialize member vars.
16768         (CallJSThatDestroysPlugin::~CallJSThatDestroysPlugin):
16769         Remove our custom property from the message window, and destroy it.
16770         (CallJSThatDestroysPlugin::NPP_Destroy):
16771         Set m_isDestroyed, log that the plug-in was destroyed, and notify the
16772         layout test controller that we're done.
16773         (wndProc):
16774         Get the PluginTest object, and call its runTest() function.
16775         (CallJSThatDestroysPlugin::NPP_New):
16776         Setup the test: register a class for the message-only window, create
16777         it, and post a message to it to run the test.
16778         (CallJSThatDestroysPlugin::runTest):
16779         Execute JS that removes the plug-in from the page, and if we're not
16780         destroyed, log a success message.
16781
16782         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
16783         Add new test to project.
16784
16785 2011-05-25  Tony Chang  <tony@chromium.org>
16786
16787         Reviewed by Adam Barth.
16788
16789         webkitpy's NetworkTransaction shouldn't use mechanize
16790         https://bugs.webkit.org/show_bug.cgi?id=61450
16791
16792         mechanize.HTTPError is just an alias for urllib2.HTTPError, so
16793         use urllib2 directly.  Re-add NetworkTransaction to
16794         test_results_uploader.py (reverting r87124).
16795
16796         * Scripts/webkitpy/common/net/networktransaction.py:
16797         * Scripts/webkitpy/layout_tests/layout_package/test_results_uploader.py:
16798
16799 2011-05-25  Kevin Ollivier  <kevino@theolliviers.com>
16800
16801         [Chromium] Unreviewed build fix. Do not define WTF/JS_EXPORT_PRIVATE to JS_EXPORTDATA
16802         until the port begins using those export macros.
16803
16804         * DumpRenderTree/chromium/config.h:
16805
16806 2011-05-25  Sheriff Bot  <webkit.review.bot@gmail.com>
16807
16808         Unreviewed, rolling out r87257.
16809         http://trac.webkit.org/changeset/87257
16810         https://bugs.webkit.org/show_bug.cgi?id=61457
16811
16812         This broke layout tests, see bug 61431. (Requested by
16813         hwennborg on #webkit).
16814
16815         * DumpRenderTree/chromium/LayoutTestController.cpp:
16816         (LayoutTestController::LayoutTestController):
16817         (LayoutTestController::clearAllDatabases):
16818         * DumpRenderTree/chromium/LayoutTestController.h:
16819
16820 2011-05-25  Qi Zhang  <qi.2.zhang@nokia.com>
16821
16822         Reviewed by Andreas Kling.
16823
16824         [Qt] fast/css/disabled-author-styles.html failed
16825         https://bugs.webkit.org/show_bug.cgi?id=61438
16826
16827         Implemented API setAuthorAndUserStylesEnabled in LayoutTestControllerQt.
16828
16829         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
16830         (WebCore::WebPage::resetSettings):
16831         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
16832         (LayoutTestController::setAuthorAndUserStylesEnabled):
16833         * DumpRenderTree/qt/LayoutTestControllerQt.h:
16834
16835 2011-05-25  Adam Roben  <aroben@apple.com>
16836
16837         Identify, rather than skip, builds where run-webkit-tests exited early due to too many failures
16838
16839         Fixes <http://webkit.org/b/61441> TestFailures page should show when run-webkit-tests
16840         started exiting early due to too many crashes, timeouts, or failures
16841
16842         Reviewed by David Kilzer.
16843
16844         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
16845         (Builder.prototype.getNumberOfFailingTests): Rather than returning -1 ("unknown") when
16846         run-webkit-tests exits early due to too many failures, pass that information in another
16847         argument to the callback.
16848         (Builder.prototype.startFetchingBuildHistory): Updated the documentation comment to reflect
16849         the new structure of the object passed to the callback.
16850         (Builder.prototype._getFailingTests): Updated to expect a tooManyFailures boolean from
16851         getNumberOfFailingTests and to pass that along to our own callbacks.
16852         (Builder.prototype._incorporateBuildHistory): Updated to expect a tooManyFailures boolean
16853         from _getFailingTests and to store that value in the history object.
16854
16855         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
16856         (.info): Added.
16857
16858         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
16859         (ViewController.prototype._displayBuilder): Updated for change in structure to the history
16860         object and to add a note when run-webkit-tests exited early due to too many failures.
16861         (ViewController.prototype._displayTesters): Renamed testersAndFailureCounts to
16862         latestBuildInfos since each entry now contains more than just the tester and failure count.
16863         Now displays a message for testers where the latest build exited early due to too many
16864         failures. Updated to expect a tooManyFailures boolean from getNumberOfFailingTests and to
16865         store that value in latestBuildInfos.
16866
16867 2011-05-24  Keishi Hattori  <keishi@webkit.org>
16868
16869         Reviewed by Kent Tamura.
16870
16871         Disable textfield implementation of <input type=color>. Add INPUT_COLOR feature flag. Add input color sanitizer.
16872         https://bugs.webkit.org/show_bug.cgi?id=61273
16873
16874         * Scripts/build-webkit: Added INPUT_COLOR feature flag.
16875
16876 2011-05-24  Greg Simon  <gregsimon@chromium.org>
16877
16878         Reviewed by Dimitri Glazkov.
16879
16880         Control Indexeddb backends from LayoutTestController
16881         https://bugs.webkit.org/show_bug.cgi?id=61000
16882
16883         * DumpRenderTree/chromium/LayoutTestController.cpp:
16884         (LayoutTestController::LayoutTestController):
16885         (LayoutTestController::setOverrideIndexedDBBackingStore):
16886         (LayoutTestController::clearAllDatabases):
16887         * DumpRenderTree/chromium/LayoutTestController.h:
16888
16889 2011-05-24  Robert Hogan  <robert@webkit.org>
16890
16891         Reviewed by Ryosuke Niwa.
16892
16893         [Qt] Fix fast/events/selectstart-by-single-click-with-shift.html
16894         https://bugs.webkit.org/show_bug.cgi?id=61393
16895
16896         * DumpRenderTree/qt/EventSenderQt.h:
16897         * DumpRenderTree/qt/EventSenderQt.cpp:
16898         (getModifiers): static function for interpreting modifiers
16899         (EventSender::mouseDown): Respect keyboard modifiers
16900         (EventSender::keyDown): use new static function
16901         (EventSender::sendOrQueueEvent): whitespace fix
16902
16903 2011-05-24  Jay Civelli  <jcivelli@chromium.org>
16904
16905         Reviewed by Adam Barth.
16906
16907         Adding MHTML reading support.
16908         https://bugs.webkit.org/show_bug.cgi?id=7168
16909
16910         * Scripts/build-webkit:
16911         * Scripts/old-run-webkit-tests:
16912         * Scripts/webkitperl/features.pm:
16913         * Scripts/webkitpy/layout_tests/port/test_files.py:
16914         * Scripts/webkitpy/layout_tests/port/webkit.py:
16915
16916 2011-05-24  Stephen White  <senorblanco@chromium.org>
16917
16918         Reviewed by Kenneth Russell.
16919
16920         Remove most style checks on Source/ThirdParty/glu, since it doesn't
16921         follow WebKit style.
16922         https://bugs.webkit.org/show_bug.cgi?id=61365
16923
16924         * Scripts/webkitpy/style/checker.py:
16925
16926 2011-05-24  Adam Roben  <aroben@apple.com>
16927
16928         Don't allow forcing builds through the build.webkit.org web interface
16929
16930         Someone has been forcing bogus builds.
16931
16932         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
16933
16934 2011-05-24  Qi Zhang  <qi.2.zhang@nokia.com>
16935
16936         Reviewed by Adam Roben.
16937
16938         WebKitTestRunner needs layoutTestController.setJavaScriptCanAccessClipboard
16939         https://bugs.webkit.org/show_bug.cgi?id=42669
16940
16941         Fixes the regression issue caused by patch 94265. The patch didn't intialize the setting before the
16942         layouttest starting, it will make the test cases after editing/execCommand/clipboard-access.html failed.
16943         Because that case changed the setting.
16944
16945         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
16946         (WTR::InjectedBundle::beginTesting):
16947
16948 2011-05-24  Csaba Osztrogonác  <ossy@webkit.org>
16949
16950         Reviewed by Kenneth Rohde Christiansen.
16951
16952         [ORWT]Make --skipped=only option work with parameters with LayoutTests/ prefix
16953         https://bugs.webkit.org/show_bug.cgi?id=61052
16954
16955         * Scripts/old-run-webkit-tests:
16956
16957 2011-05-24  Adam Roben  <aroben@apple.com>
16958
16959         Make TestFailures show how many tests are failing on each tester, and omit testers with no failures
16960
16961         Fixes <http://webkit.org/b/61063> <rdar://problem/9460533> TestFailures page shows testers
16962         that don't have any failing tests, which isn't useful
16963
16964         Reviewed by David Kilzer.
16965
16966         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js:
16967         (Buildbot.prototype.getTesters): Renamed from getTesterNames. Now returns Builder objects
16968         instead of name strings.
16969         (Buildbot.prototype._buildersForNames): Added. Helper function to convert an array of
16970         builder names into an array of builders.
16971
16972         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
16973         (Builder.prototype.getMostRecentCompletedBuildNumber): Added. Returns the build number of
16974         the most recently completed build, or -1 if there is no such build.
16975         (Builder.prototype.getNumberOfFailingTests): Added. Returns the number of tests that failed
16976         in the given build, or -1 if the number could not be determined. Some of this code came from
16977         _getFailingTests.
16978         (Builder.prototype._getBuildJSON): Added. Code came from _getFailingTests.
16979         (Builder.prototype._getFailingTests): Changed to use new _getBuildJSON and
16980         getNumberOfFailingTests functions.
16981
16982         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
16983         (ViewController.prototype._displayTesters): Get the current number of test failures for each
16984         tester and show it in the list. Omit testers that have no failures at all. We keep the
16985         testers and failure counts in an array and sort it before displaying the current data, as
16986         the order in which data will be fetched is unpredictable.
16987
16988 2011-05-23  Tony Chang  <tony@chromium.org>
16989
16990         Reviewed by Ojan Vafai.
16991
16992         [nrwt] don't use NetworkTranslation when uploading results
16993         https://bugs.webkit.org/show_bug.cgi?id=61315
16994
16995         This avoids autoinstalling mechanize.  It didn't matter anyway
16996         because test_results_uploader.py wasn't using mechanize to make the
16997         request so NetworkTransaction::run wasn't doing anything extra
16998         (it only matters if it catches mechanize.HTTPError).
16999
17000         * Scripts/webkitpy/layout_tests/layout_package/test_results_uploader.py:
17001
17002 2011-05-23  Tony Chang  <tony@chromium.org>
17003
17004         Reviewed by Dirk Pranke.
17005
17006         [chromium] fix the rebaseline tool to work with Linux GPU failures
17007         https://bugs.webkit.org/show_bug.cgi?id=61312
17008
17009         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
17010
17011 2011-05-23  Darin Adler  <darin@apple.com>
17012
17013         Ignore generated .pyc files in a directory full of python code.
17014
17015         * Scripts/webkitpy/common/checkout/scm: Added property svn:ignore.
17016
17017 2011-05-23  Adam Barth  <abarth@webkit.org>
17018
17019         Reviewed by Eric Seidel.
17020
17021         sheriffbot should give more details about the failures in IRC
17022         https://bugs.webkit.org/show_bug.cgi?id=61233
17023
17024         With this patch, sheriffbot will annonce the set of failing tests,
17025         which might help folks triage the problem.
17026
17027         * Scripts/webkitpy/tool/commands/sheriffbot.py:
17028         * Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
17029
17030 2011-05-23  Sheriff Bot  <webkit.review.bot@gmail.com>
17031
17032         Unreviewed, rolling out r87078.
17033         http://trac.webkit.org/changeset/87078
17034         https://bugs.webkit.org/show_bug.cgi?id=61294
17035
17036         Patch contains errors and seems likely to break a bunch of
17037         bots (Requested by abarth on #webkit).
17038
17039         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
17040         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
17041         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
17042
17043 2011-05-23  Alice Boxhall  <aboxhall@chromium.org>
17044
17045         Reviewed by Ojan Vafai.
17046
17047         Convert json_results_generator.py to output version 4 JSON.
17048         https://bugs.webkit.org/show_bug.cgi?id=60869
17049
17050         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
17051         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
17052         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
17053
17054 2011-05-23  Patrick Gansterer  <paroga@webkit.org>
17055
17056         Reviewed by Adam Barth.
17057
17058         Move generation of shell_command from port to executive
17059         https://bugs.webkit.org/show_bug.cgi?id=61251
17060
17061         After move it's possible to use the code in other places too.
17062
17063         * Scripts/webkitpy/common/config/ports.py:
17064         * Scripts/webkitpy/common/system/executive.py:
17065
17066 2011-05-23  Qi Zhang  <qi.2.zhang@nokia.com>
17067
17068         Reviewed by Darin Adler.
17069
17070         WebKitTestRunner needs layoutTestController.setJavaScriptCanAccessClipboard
17071         https://bugs.webkit.org/show_bug.cgi?id=42669
17072
17073         Implement setJavaScriptCanAccessClipboard in WebKitTestRunner.
17074
17075         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
17076         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
17077         (WTR::LayoutTestController::setJavaScriptCanAccessClipboard):
17078         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
17079
17080 2011-05-22  Robert Hogan  <robert@webkit.org>
17081
17082         Reviewed by Kenneth Rohde Christiansen.
17083
17084         Fix policyDelegate in Qt DRT
17085         https://bugs.webkit.org/show_bug.cgi?id=61247
17086
17087         Use the 'policy delegate' implemented in FrameLoaderClient
17088         for layout tests. The partial implementation in DumpRenderTreeQt
17089         is redundant, so remove it.
17090
17091         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
17092         (WebCore::WebPage::acceptNavigationRequest):
17093         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
17094         (LayoutTestController::waitForPolicyDelegate):
17095
17096 2011-05-22  Patrick Gansterer  <paroga@webkit.org>
17097
17098         Reviewed by Darin Adler.
17099
17100         Use double quotes for filename arguments
17101         https://bugs.webkit.org/show_bug.cgi?id=61250
17102
17103         Windows does not support single quotes on the command line,
17104         so replace them with double quotes.
17105
17106         * Scripts/prepare-ChangeLog:
17107
17108 2011-05-20  Simon Fraser  <simon.fraser@apple.com>
17109
17110         Reviewed by Sam Weinig.
17111
17112         WebKitTestRunner needs layoutTestController.pauseTransitionAtTimeOnElementWithId
17113         https://bugs.webkit.org/show_bug.cgi?id=42550
17114
17115         Implement pauseTransitionAtTimeOnElementWithId in WebKitTestRunner.
17116
17117         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
17118         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
17119         Copy the code for pausing animations, FIXME comments and all.
17120         (WTR::LayoutTestController::pauseTransitionAtTimeOnElementWithId):
17121         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
17122
17123 2011-05-20  Alok Priyadarshi  <alokp@chromium.org>
17124
17125         Reviewed by Stephen White.
17126
17127         Adding myself to the committers list.
17128
17129         * Scripts/webkitpy/common/config/committers.py:
17130
17131 2011-05-19  Adam Roben  <aroben@apple.com>
17132
17133         Test that the WebKit2 UI process doesn't crash when starting a download
17134
17135         Test for <http://webkit.org/b/61142> <rdar://problem/9471680> REGRESSION (r86812): Crash
17136         (preceded by assertion) in fastMalloc when downloading a file
17137
17138         Reviewed by Darin Adler.
17139
17140         * TestWebKitAPI/Tests/WebKit2/18-characters.html: Added.
17141
17142         * TestWebKitAPI/Tests/WebKit2/DownloadDecideDestinationCrash.cpp: Added.
17143         (TestWebKitAPI::decidePolicyForNavigationAction): Start a download.
17144         (TestWebKitAPI::decideDestinationWithSuggestedFilename): Record that the download was
17145         started, cancel the download, and return a bogus string.
17146
17147         (TestWebKitAPI::setContextDownloadClient):
17148         (TestWebKitAPI::setPagePolicyClient):
17149         Simple helper functions.
17150
17151         (TestWebKitAPI::TEST): Load 18-characters.html, which should trigger a download thanks to
17152         our policy client, and run until we know that the download was started. If we haven't
17153         crashed, we win!
17154
17155         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
17156         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
17157         * TestWebKitAPI/win/copy-resources.cmd:
17158         Added new files.
17159
17160 2011-05-20  Kent Tamura  <tkent@chromium.org>
17161
17162         Reviewed by Ryosuke Niwa.
17163
17164         Fix style errors in DumpRenderTree/chromium/.
17165         https://bugs.webkit.org/show_bug.cgi?id=61172
17166
17167         * DumpRenderTree/chromium/CppBoundClass.cpp:
17168         (CppBoundClass::getAsCppVariant):
17169         (CppBoundClass::bindToJavascript):
17170         * DumpRenderTree/chromium/CppBoundClass.h:
17171         (CppBoundClass::GetterCallback::~GetterCallback):
17172         (CppBoundClass::CppBoundClass):
17173         (CppBoundClass::Callback::~Callback):
17174         (CppBoundClass::MemberCallback::MemberCallback):
17175         (CppBoundClass::MemberCallback::~MemberCallback):
17176         (CppBoundClass::MemberGetterCallback::MemberGetterCallback):
17177         (CppBoundClass::MemberGetterCallback::~MemberGetterCallback):
17178         * DumpRenderTree/chromium/CppVariant.h:
17179         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
17180         (DRTDevToolsAgent::~DRTDevToolsAgent):
17181         * DumpRenderTree/chromium/DRTDevToolsClient.cpp:
17182         (DRTDevToolsClient::sendFrontendLoaded):
17183         * DumpRenderTree/chromium/DRTDevToolsClient.h:
17184         * DumpRenderTree/chromium/DumpRenderTree.cpp:
17185         (main):
17186         * DumpRenderTree/chromium/EventSender.cpp:
17187         (SavedEvent::SavedEvent):
17188         (EventSender::EventSender):
17189         (EventSender::keyDown):
17190         * DumpRenderTree/chromium/EventSender.h:
17191         * DumpRenderTree/chromium/ImageDiff.cpp:
17192         (Image::Image):
17193         * DumpRenderTree/chromium/LayoutTestController.cpp:
17194         (LayoutTestController::LayoutTestController):
17195         (LayoutTestController::WorkQueue::reset):
17196         (WorkItemBackForward::WorkItemBackForward):
17197         (WorkItemLoadingScript::WorkItemLoadingScript):
17198         (WorkItemNonLoadingScript::WorkItemNonLoadingScript):
17199         (WorkItemLoad::WorkItemLoad):
17200         (WorkItemLoadHTMLString::WorkItemLoadHTMLString):
17201         (LayoutTestController::pathToLocalResource):
17202         * DumpRenderTree/chromium/LayoutTestController.h:
17203         (LayoutTestController::WorkItem::~WorkItem):
17204         (LayoutTestController::WorkQueue::WorkQueue):
17205         * DumpRenderTree/chromium/MockSpellCheck.cpp:
17206         (MockSpellCheck::MockSpellCheck):
17207         (MockSpellCheck::~MockSpellCheck):
17208         * DumpRenderTree/chromium/MockSpellCheck.h:
17209         * DumpRenderTree/chromium/NotificationPresenter.h:
17210         (NotificationPresenter::NotificationPresenter):
17211         * DumpRenderTree/chromium/Task.cpp:
17212         (WebTask::WebTask):
17213         * DumpRenderTree/chromium/Task.h:
17214         (TaskList::TaskList):
17215         * DumpRenderTree/chromium/TestEventPrinter.cpp:
17216         (DRTPrinter::DRTPrinter):
17217         (TestShellPrinter::TestShellPrinter):
17218         * DumpRenderTree/chromium/TestNavigationController.cpp:
17219         (TestNavigationEntry::TestNavigationEntry):
17220         (TestNavigationEntry::~TestNavigationEntry):
17221         (TestNavigationController::TestNavigationController):
17222         (TestNavigationController::reload):
17223         (TestNavigationController::loadEntry):
17224         (TestNavigationController::didNavigateToEntry):
17225         * DumpRenderTree/chromium/TestNavigationController.h:
17226         (TestShellExtraData::TestShellExtraData):
17227         * DumpRenderTree/chromium/TestShell.cpp:
17228         (dumpHistoryItem):
17229         * DumpRenderTree/chromium/TestShell.h:
17230         (TestParams::TestParams):
17231         * DumpRenderTree/chromium/TestShellWin.cpp:
17232         (TestShell::waitTestFinished):
17233         * DumpRenderTree/chromium/TestWebWorker.h:
17234         (TestWebWorker::startWorkerContext):
17235         (TestWebWorker::terminateWorkerContext):
17236         (TestWebWorker::postMessageToWorkerContext):
17237         (TestWebWorker::clientDestroyed):
17238         (TestWebWorker::postMessageToWorkerObject):
17239         (TestWebWorker::postExceptionToWorkerObject):
17240         (TestWebWorker::postConsoleMessageToWorkerObject):
17241         (TestWebWorker::confirmMessageFromWorkerObject):
17242         (TestWebWorker::reportPendingActivity):
17243         (TestWebWorker::workerContextClosed):
17244         (TestWebWorker::~TestWebWorker):
17245         * DumpRenderTree/chromium/TextInputController.cpp:
17246         * DumpRenderTree/chromium/WebThemeControlDRTWin.h:
17247         * DumpRenderTree/chromium/WebThemeEngineDRTMac.h:
17248         * DumpRenderTree/chromium/WebThemeEngineDRTWin.cpp:
17249         (WebThemeEngineDRTWin::paintButton):
17250         (WebThemeEngineDRTWin::paintMenuList):
17251         (WebThemeEngineDRTWin::paintTrackbar):
17252         * DumpRenderTree/chromium/WebThemeEngineDRTWin.h:
17253         (WebThemeEngineDRTWin::WebThemeEngineDRTWin):
17254         * DumpRenderTree/chromium/WebViewHost.cpp:
17255         (WebViewHost::navigate):
17256         (WebViewHost::updateSessionHistory):
17257         (WebViewHost::paintInvalidatedRegion):
17258         * DumpRenderTree/chromium/WebViewHost.h:
17259
17260 2011-05-19  Dmitry Lomov  <dslomov@google.com>
17261
17262         Reviewed by Adam Roben.
17263
17264         Detect hangs in run-api-tests
17265         https://bugs.webkit.org/show_bug.cgi?id=48043
17266
17267         * Scripts/run-api-tests: Added test timeouts
17268
17269 2011-05-19  Dmitry Lomov  <dslomov@google.com>
17270
17271         Reviewed by Adam Roben.
17272
17273         run-api-tests should run one test per process
17274         https://bugs.webkit.org/show_bug.cgi?id=61088
17275
17276         * Scripts/run-api-tests: Resurrecting the previous revison of this file, with fixes to system call under Windows,
17277         return code, and parsing GTest output format.
17278
17279 2011-05-15  Robert Hogan  <robert@webkit.org>
17280
17281         Reviewed by Antonio Gomes.
17282
17283         [Qt][GTK] plugins/get-url-with-javascript-url.html fails
17284         https://bugs.webkit.org/show_bug.cgi?id=60834
17285         Fix unix test plugin for plugins/get-url-with-javascript-url.html
17286
17287         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
17288         (webkit_test_plugin_new_stream):
17289         (webkit_test_plugin_write_ready):
17290         (webkit_test_plugin_write):
17291
17292 2011-05-19  Sheriff Bot  <webkit.review.bot@gmail.com>
17293
17294         Unreviewed, rolling out r86869, r86873, r86875, and r86877.
17295         http://trac.webkit.org/changeset/86869
17296         http://trac.webkit.org/changeset/86873
17297         http://trac.webkit.org/changeset/86875
17298         http://trac.webkit.org/changeset/86877
17299         https://bugs.webkit.org/show_bug.cgi?id=61139
17300
17301         broke builds and debug DRT (Requested by rniwa on #webkit).
17302
17303         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
17304         * DumpRenderTree/chromium/TestShell.cpp:
17305         (TestShell::bindJSObjectsToWindow):
17306         * DumpRenderTree/mac/FrameLoadDelegate.mm:
17307         (-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]):
17308
17309 2011-05-18  Dimitri Glazkov  <dglazkov@chromium.org>
17310
17311         Reviewed by Darin Adler.
17312
17313         Add build logistics and plumbing for window.internals object.
17314         https://bugs.webkit.org/show_bug.cgi?id=60313
17315
17316         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added linking new WebCoreTestSupport library.
17317         * DumpRenderTree/chromium/TestShell.cpp:
17318         (TestShell::bindJSObjectsToWindow): Added injection code.
17319         * DumpRenderTree/mac/FrameLoadDelegate.mm:
17320         (-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]): Ditto.
17321
17322 2011-05-19  Zsolt Fehér  <h490267@stud.u-szeged.hu>
17323
17324         Reviewed by Csaba Osztrogonác.
17325
17326         [Qt] Implement eventSender.scalePageBy
17327         https://bugs.webkit.org/show_bug.cgi?id=60015
17328
17329         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
17330         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
17331         * DumpRenderTree/qt/EventSenderQt.cpp:
17332         (EventSender::scalePageBy):
17333         * DumpRenderTree/qt/EventSenderQt.h:
17334
17335 2011-05-19  Kenichi Ishibashi  <bashi@chromium.org>
17336
17337         Reviewed by Kent Tamura.
17338
17339         [Chromium] IME candidate window appears wrong position in an iframe
17340         https://bugs.webkit.org/show_bug.cgi?id=61023
17341
17342         Call the focused frame's firstRectForCharacterRange() instead of the
17343         main frame so that DRT behaves as the same as Chromium.
17344
17345         * DumpRenderTree/chromium/TextInputController.cpp:
17346         (TextInputController::firstRectForCharacterRange):
17347
17348 2011-05-18  Adam Roben  <aroben@apple.com>
17349
17350         Add a new page to build.webkit.org to help find when tests started failing
17351
17352         The page is accessible at <http://build.webkit.org/TestFailures/>. It is pretty minimalist
17353         right now, but already shows some useful information. It's somewhat similar to webkit-patch
17354         failure-reason and sheriffbot, and perhaps can be combined with them eventually. It's a
17355         little more convenient than either of them, though, because it's all done in the browser
17356         (and thus it's easy to go directly to the relevant test results).
17357
17358         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js: Added.
17359         (Buildbot): This class represents a Buildbot server.
17360         (Buildbot.prototype.buildURL): Returns the URL for the summary page for a particular build.
17361         (Buildbot.prototype.builderNamed): Returns a Builder with the given name.
17362         (Buildbot.prototype.getTesterNames): Fetches the names of all testers and passes them to the
17363         callback.
17364         (Buildbot.prototype.parseBuildName): Breaks up a build name into its constituent parts. Must
17365         be implemented by a derived class that understands this server's build naming scheme.
17366         (Buildbot.prototype.resultsDirectoryURL): Returns the URL for the results directory for a
17367         particular build.
17368
17369         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js: Added.
17370         (Builder): This class represents one builder on the buildbot.
17371         (Builder.prototype.buildURL): Returns the URL for the summary page for a particular build.
17372         (Builder.prototype.failureDiagnosisTextAndURL): Returns data that provides a little more
17373         information about a particular test failure.
17374         (Builder.prototype.startFetchingBuildHistory): Periodically calls the callback with
17375         information about when tests started failing.
17376         (Builder.prototype.resultsDirectoryURL): Returns the URL for the results directory for a
17377         particular build.
17378         (Builder.prototype._getBuildNames): Fetches the names of all builds and passes them to the
17379         callback.
17380         (Builder.prototype._getFailingTests): Fetches the results.html page for the given build and
17381         extracts all the failing tests listed in it, passing them to the callback.
17382         (Builder.prototype._incorporateBuildHistory): Gets the failing tests for the specified
17383         build, merges them into the build history, and calls the callback telling it whether the
17384         next build should be fetched to provide more information.
17385
17386         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
17387         Added. Just some simple styles.
17388
17389         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js: Added.
17390         (createDefinitionList): Takes an array of pairs and turns them into a DL element.
17391         (getResource): Wrapper around XMLHttpRequest.
17392         (Array.prototype.findFirst): Finds the first element matching the given predicate and
17393         returns it.
17394         (Array.prototype.last): Returns the last element of the array.
17395
17396         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js: Added.
17397         (ViewController): This class contains the main logic for displaying the page.
17398         (ViewController.loaded): Just calls through to parseHash.
17399         (ViewController.parseHash): Either starts analyzing failures on a particular builder, or
17400         shows the list of all testers so one can be chosen. This function is called when the page
17401         loads and whenever we get a hashchange event.
17402         (ViewController._displayBuilder): Asks the builder to fetch build history, and displays it
17403         as it is fetched. The display ends up grouping tests by when they started failing.
17404         (ViewController._displayTesters): Gets the list of testers and displays it.
17405
17406         (ViewController._domForBuildName):
17407         (ViewController._domForFailedTest):
17408         Helper functions to create descriptions and links for a particular build or failed test.
17409
17410         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBuildbot.js: Added.
17411         (WebKitBuildbot): Calls up to the base class constructor with the correct base URL.
17412         (WebKitBuildbot.prototype.parseBuildName): Parses a build.webkit.org-style build name.
17413
17414         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Added. Just
17415         loads all the files and sets up a ViewController, which does the rest.
17416
17417         * BuildSlaveSupport/build.webkit.org-config/templates/root.html: Added a link to the new
17418         page.
17419
17420 2011-05-18  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
17421
17422         Reviewed by Csaba Osztrogonác.
17423
17424         [Qt] Implement layoutTestController.setValueForUser()
17425         https://bugs.webkit.org/show_bug.cgi?id=60956
17426
17427         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
17428         (LayoutTestController::setValueForUser):
17429         * DumpRenderTree/qt/LayoutTestControllerQt.h:
17430
17431 2011-05-16  Robert Hogan  <robert@webkit.org>
17432
17433         Reviewed by Kenneth Rohde Christiansen.
17434
17435         plugins/invalidate_rect.html fails on linux ports
17436
17437         - Make the unix test netscape plugin recognize the onPaintEvent
17438           and windowedPlugin parameters.
17439
17440         https://bugs.webkit.org/show_bug.cgi?id=54051
17441
17442         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
17443         (handleEventX11):
17444         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
17445         (LayoutTestController::displayInvalidatedRegion):
17446         * DumpRenderTree/qt/LayoutTestControllerQt.h:
17447         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
17448         (webkit_test_plugin_new_instance):
17449         (webkit_test_plugin_handle_event):
17450
17451 2011-05-17  Tony Chang  <tony@chromium.org>
17452
17453         Reviewed by Ojan Vafai.
17454
17455         [chromium] move Lucid 64 bit results into LayoutTests/platform/chromium-linux
17456         https://bugs.webkit.org/show_bug.cgi?id=60895
17457
17458         Update the tools to handle the move.  Lucid 32 results now go in
17459         chromium-linux-x86 and the default platform on Linux is now x86_64.
17460
17461         * Scripts/webkitpy/layout_tests/deduplicate_tests.py: Default to x86_64
17462         * Scripts/webkitpy/layout_tests/deduplicate_tests_unittest.py:
17463         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py: Default to Lucid
17464         * Scripts/webkitpy/layout_tests/port/base.py: ditto
17465         * Scripts/webkitpy/layout_tests/port/chromium_linux.py: Update directory fallback and default to Lucid 64
17466         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
17467         * Scripts/webkitpy/layout_tests/port/test.py: Update tests to default to x86_64
17468         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py: Update bot names.
17469         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py: Whitespace cleanup.
17470
17471 2011-05-17  Adam Roben  <aroben@apple.com>
17472
17473         Make run-api-tests work on Windows when there are spaces in the path
17474
17475         Fixes <http://webkit.org/b/60954> REGRESSION (r86511): run-api-tests fails if there are
17476         spaces in the path to TestWebKitAPI.exe
17477
17478         Reviewed by David Levin.
17479
17480         * Scripts/run-api-tests:
17481         (runTestTool): Use the "direct object" form of system() to avoid having the path to
17482         TestWebKitAPI.exe be split by the shell.
17483
17484 2011-05-16  Ian Henderson  <ianh@apple.com>
17485
17486         Reviewed by Joseph Pecoraro.
17487
17488         Page::goToItem doesn't work while loading is deferred
17489         https://bugs.webkit.org/show_bug.cgi?id=60412
17490
17491         Add setDefersLoading and goBack methods to LayoutTestController.  We
17492         need to use goBack() instead of history.back() because the latter goes
17493         through NavigationScheduler, hence doesn't exhibit the bug.
17494
17495         * DumpRenderTree/LayoutTestController.cpp:
17496         (goBackCallback):
17497         (setDefersLoadingCallback):
17498         (LayoutTestController::staticFunctions):
17499         * DumpRenderTree/LayoutTestController.h:
17500         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
17501         (LayoutTestController::goBack):
17502         (LayoutTestController::setDefersLoading):
17503         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
17504         (LayoutTestController::goBack):
17505         (LayoutTestController::setDefersLoading):
17506         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
17507         (LayoutTestController::goBack):
17508         (LayoutTestController::setDefersLoading):
17509         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
17510         (LayoutTestController::goBack):
17511         (LayoutTestController::setDefersLoading):
17512
17513 2011-05-16  Sam Weinig  <sam@webkit.org>
17514
17515         Reviewed by Anders Carlsson.
17516
17517         TestWebKitAPI should build with clang if it can
17518         https://bugs.webkit.org/show_bug.cgi?id=60918
17519
17520         * TestWebKitAPI/Configurations/CompilerVersion.xcconfig:
17521         Update CompilerVersion.xcconfig to match others.
17522
17523         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
17524         Add CompilerVersion.xcconfig to the project.
17525
17526         * TestWebKitAPI/Tests/WebKit2/FailedLoad.cpp:
17527         (TestWebKitAPI::didFailProvisionalLoadWithErrorForFrame):
17528         * TestWebKitAPI/Tests/WebKit2/PageLoadDidChangeLocationWithinPageForFrame.cpp:
17529         (TestWebKitAPI::didSameDocumentNavigationForFrame):
17530         Add some casts to quiet warnings from clang.
17531
17532 2011-05-16  Dirk Pranke  <dpranke@chromium.org>
17533
17534         Reviewed by David Levin.
17535
17536         add dpranke as a reviewer
17537         https://bugs.webkit.org/show_bug.cgi?id=60919
17538
17539         * Scripts/webkitpy/common/config/committers.py:
17540
17541 2011-05-16  Robert Hogan  <robert@webkit.org>
17542
17543         Reviewed by Anders Carlsson.
17544
17545         [Gtk] plugins/get-url-notify-with-url-that-fails-to-load.html crashes on buildbot
17546
17547         https://bugs.webkit.org/show_bug.cgi?id=60838
17548
17549         The unix test plugin needs to call the test's NPP_URLNotify
17550         rather than the browsers.
17551
17552         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
17553         (webkit_test_plugin_url_notify):
17554
17555 2011-05-16  Brent Fulgham  <bfulgham@webkit.org>
17556
17557         Rubber-stamped by Adam Roben.
17558
17559         Adding myself as a reviewer.
17560         * Scripts/webkitpy/common/config/committers.py:
17561
17562 2011-05-16  Sam Weinig  <sam@webkit.org>
17563
17564         Fix typo pointed out by Dave Levin.
17565
17566         * TestWebKitAPI/PlatformUtilities.h:
17567         (TestWebKitAPI::Util::assertWKStringEqual):
17568         Strig -> String.
17569
17570 2011-05-16  Sam Weinig  <sam@webkit.org>
17571
17572         Reviewed by David Levin.
17573
17574         Convert api tester over to using gtest expectations directly
17575         https://bugs.webkit.org/show_bug.cgi?id=60862
17576
17577         * TestWebKitAPI/PlatformUtilities.cpp:
17578         (TestWebKitAPI::Util::toSTD):
17579         * TestWebKitAPI/PlatformUtilities.h:
17580         (TestWebKitAPI::Util::assertWKStrigEqual):
17581         Add convenience macro to compare WK2 strings. Add some overloads of toSTD to make
17582         the implementation of the macro simpler.
17583
17584         * TestWebKitAPI/Test.h:
17585         Remove TEST_ASSERT forwarder.
17586
17587         [Test changes elided]
17588
17589 2011-05-16  David Kilzer  <ddkilzer@apple.com>
17590
17591         <http://webkit.org/b/60913> C++ exceptions should not be enabled when building with llvm-gcc-4.2
17592         <rdar://problem/9446430>
17593
17594         Reviewed by Mark Rowe.
17595
17596         * MiniBrowser/Configurations/Base.xcconfig: Fixed typo.
17597         * TestWebKitAPI/Configurations/Base.xcconfig: Ditto.
17598         * WebKitTestRunner/Configurations/Base.xcconfig: Ditto.
17599
17600 2011-05-16  Jon Lee  <jonlee@apple.com>
17601
17602         Reviewed by Simon Fraser.
17603
17604         Can't horizontally scroll iframes and overflow because wheel events are always accepted
17605         https://bugs.webkit.org/show_bug.cgi?id=60779
17606
17607         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
17608         (WTR::InjectedBundlePage::InjectedBundlePage): Set new default method to nil.
17609
17610 2011-05-16  Ademar de Souza Reis Jr.  <ademar.reis@openbossa.org>
17611
17612         Reviewed by Csaba Osztrogonác.
17613
17614         [Qt] Tools.pro misses include(common.pri)
17615         https://bugs.webkit.org/show_bug.cgi?id=60883
17616
17617         It includes features.pri, which depends on common.pri (the build is
17618         not broken today with the default options, but this is the right
17619         thing to do).
17620
17621         * Tools.pro:
17622
17623 2011-05-15  Dominic Cooney  <dominicc@chromium.org>
17624
17625         Unreviewed: adding myself to committers.py.
17626
17627         * Scripts/webkitpy/common/config/committers.py:
17628
17629 2011-05-15  Daniel Bates  <dbates@webkit.org>
17630
17631         Reviewed by Chris Jerdonek.
17632
17633         svn-apply can't handle Git diff that only changes executable bit
17634         https://bugs.webkit.org/show_bug.cgi?id=60848
17635
17636         Fixes an issue where svn-{apply, unapply} fail to handle a Git diff for a file
17637         that only has an executable bit change.
17638
17639         * Scripts/VCSUtils.pm:
17640           - Modified regex $chunkRangeRegEx to match chunk range with ending text, say: @@ -2,6 +2,18 @@ foo().
17641         (parseDiff): Modified to count the number of text chunks in the diff. This should also help towards
17642         fixing <https://bugs.webkit.org/show_bug.cgi?id=29684>.
17643         * Scripts/svn-apply:
17644         (patch): Only use standard patch tool if the patch is non-empty and for a text file (i.e. not a binary patch).
17645         * Scripts/svn-unapply:
17646         (patch): Ditto.
17647         * Scripts/webkitperl/VCSUtils_unittest/parseDiff.pl:
17648           - Updated expected results for test cases to account for numTextChunks.
17649           - Fixed malformed chunk range in test cases: "Git: simple", and "Git: Git diff followed by SVN diff".
17650           - Added test case "Git: file that only has an executable bit change".
17651         * Scripts/webkitperl/VCSUtils_unittest/parseDiffWithMockFiles.pl:
17652           - Updated expected results for test cases to account for numTextChunks.
17653
17654 2011-05-15  Sam Weinig  <sam@webkit.org>
17655
17656         Fix windows build.
17657
17658         * TestWebKitAPI/win/main.cpp:
17659         (main):
17660
17661 2011-05-15  Sam Weinig  <sam@webkit.org>
17662
17663         Reviewed by David Levin.
17664
17665         Make API tester more gtest friendly
17666         https://bugs.webkit.org/show_bug.cgi?id=60859
17667
17668         * Scripts/run-api-tests:
17669         Numerous changes:
17670         - Remove logic to run tests one at a time.
17671         - Remove logic to build up a list of tests, instead just pass --gtest_list_tests to
17672           the test runner and allow gtest to print it for us.
17673         - Forward stdout from the test tool to stdout, even in non-verbose runs.
17674         - Run test tool from the current working directory instead of the root webkit
17675           directory.
17676
17677         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
17678         Move plist to be with other bundle sources.
17679
17680         * TestWebKitAPI/TestsController.cpp:
17681         (TestWebKitAPI::TestsController::TestsController):
17682         (TestWebKitAPI::TestsController::run):
17683         * TestWebKitAPI/TestsController.h:
17684         * TestWebKitAPI/mac/main.mm:
17685         Simplify TestController down to a single run function. We should consider removing the class
17686         entirely at some point.
17687
17688 2011-05-15  Sheriff Bot  <webkit.review.bot@gmail.com>
17689
17690         Unreviewed, rolling out r86504.
17691         http://trac.webkit.org/changeset/86504
17692         https://bugs.webkit.org/show_bug.cgi?id=60853
17693
17694         "Broke Qt EventSender in editing/selection tests" (Requested
17695         by mwenge on #webkit).
17696
17697         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
17698         (WebCore::WebPage::WebPage):
17699         (WebCore::DumpRenderTree::DumpRenderTree):
17700         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
17701         (WebCore::DumpRenderTree::initJSObjects):
17702         * DumpRenderTree/qt/DumpRenderTreeQt.h:
17703         (WebCore::DumpRenderTree::eventSender):
17704         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
17705         (webkit_test_plugin_new_instance):
17706         (webkit_test_plugin_set_window):
17707         (webkit_test_plugin_handle_event):
17708
17709 2011-05-12  Robert Hogan  <robert@webkit.org>
17710
17711         Reviewed by Benjamin Poulain.
17712
17713         [Qt] fix http/tests/plugins/plugin-document-has-focus.html
17714
17715         The support for this test added to the unix test plugin here
17716         may allow other platforms to pass it (nearly everyone
17717         skips it). On Qt it required a bit of trickiness with
17718         the page's EventSender object to get it working fully though,
17719         so only unskipping Qt here.
17720
17721         https://bugs.webkit.org/show_bug.cgi?id=60722
17722
17723         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
17724         (WebCore::WebPage::WebPage):
17725         (WebCore::WebPage::eventSender):
17726         (WebCore::DumpRenderTree::DumpRenderTree):
17727         (WebCore::DumpRenderTree::initJSObjects):
17728         (WebCore::DumpRenderTree::createWindow):
17729         * DumpRenderTree/qt/DumpRenderTreeQt.h:
17730         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
17731         (webkit_test_plugin_new_instance):
17732         (webkit_test_plugin_set_window):
17733         (webkit_test_plugin_handle_event):
17734
17735 2011-05-14  Arno Renevier  <arno@renevier.net>
17736
17737         Reviewed by Martin Robinson.
17738
17739         make run-gtk-tests executable
17740         https://bugs.webkit.org/show_bug.cgi?id=59278
17741
17742         * Scripts/run-gtk-tests:
17743
17744 2011-05-14  Anders Carlsson  <andersca@apple.com>
17745
17746         Reviewed by Kevin Decker.
17747
17748         NPP_URLNotify is not called if a URL passed to NPN_GetURLNotify fails to load
17749         https://bugs.webkit.org/show_bug.cgi?id=60823
17750         <rdar://problem/9430386>
17751
17752         Add test that calls NPN_GetURLNotify on a URL that can't be loaded, and checks if NPP_URLNotify is called.
17753
17754         * DumpRenderTree/DumpRenderTree.gypi:
17755         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
17756         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLNotifyWithURLThatFailsToLoad.cpp: Added.
17757         (GetURLNotifyWithURLThatFailsToLoad::GetURLNotifyWithURLThatFailsToLoad):
17758         (GetURLNotifyWithURLThatFailsToLoad::NPP_New):
17759         (GetURLNotifyWithURLThatFailsToLoad::NPP_URLNotify):
17760         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
17761         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
17762         * GNUmakefile.am:
17763
17764 2011-05-14  Sheriff Bot  <webkit.review.bot@gmail.com>
17765
17766         Unreviewed, rolling out r86476.
17767         http://trac.webkit.org/changeset/86476
17768         https://bugs.webkit.org/show_bug.cgi?id=60832
17769
17770         It made plugins/get-url-notify-with-url-that-fails-to-
17771         load.html crash on GTK (Requested by Ossy on #webkit).
17772
17773         * DumpRenderTree/DumpRenderTree.gypi:
17774         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
17775         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLNotifyWithURLThatFailsToLoad.cpp: Removed.
17776         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
17777         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
17778         * GNUmakefile.am:
17779
17780 2011-05-13  Anders Carlsson  <andersca@apple.com>
17781
17782         Reviewed by Sam Weinig.
17783
17784         NPP_URLNotify is not called if a URL passed to NPN_GetURLNotify fails to load
17785         https://bugs.webkit.org/show_bug.cgi?id=60823
17786         <rdar://problem/9430386>
17787
17788         Add test that calls NPN_GetURLNotify on a URL that can't be loaded, and checks if NPP_URLNotify is called.
17789
17790         * DumpRenderTree/DumpRenderTree.gypi:
17791         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
17792         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLNotifyWithURLThatFailsToLoad.cpp: Added.
17793         (GetURLNotifyWithURLThatFailsToLoad::GetURLNotifyWithURLThatFailsToLoad):
17794         (GetURLNotifyWithURLThatFailsToLoad::NPP_New):
17795         (GetURLNotifyWithURLThatFailsToLoad::NPP_URLNotify):
17796         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
17797         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
17798         * GNUmakefile.am:
17799
17800 2011-05-13  Anders Carlsson  <andersca@apple.com>
17801
17802         Build fix.
17803
17804         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLWithJavaScriptURL.cpp:
17805
17806 2011-05-13  Anders Carlsson  <andersca@apple.com>
17807
17808         Reviewed by Sam Weinig.
17809
17810         NPN_GetURL with a javascript: URL doesn't send back the result in the stream
17811         https://bugs.webkit.org/show_bug.cgi?id=60810
17812
17813         Add a test that calls NPN_GetURL on a javascript: URL and checks that it gets back the expected result.
17814
17815         * DumpRenderTree/DumpRenderTree.gypi:
17816         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
17817         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
17818         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
17819         * GNUmakefile.am:
17820         Add files.
17821
17822         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
17823         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
17824         Add new NPN and NPP wrappers and sort them according to the order they appear in npfunctions.h
17825
17826         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLWithJavaScriptURL.cpp: Added.
17827         (GetURLWithJavaScriptURL::GetURLWithJavaScriptURL):
17828         (GetURLWithJavaScriptURL::NPP_New):
17829         (GetURLWithJavaScriptURL::NPP_NewStream):
17830         (GetURLWithJavaScriptURL::NPP_DestroyStream):
17831         (GetURLWithJavaScriptURL::NPP_WriteReady):
17832         (GetURLWithJavaScriptURL::NPP_Write):
17833         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
17834         (NPP_NewStream):
17835         (NPP_WriteReady):
17836         (NPP_Write):
17837         (NPP_URLNotify):
17838         call the PluginTest NPP wrappers.
17839
17840 2011-05-13  Sheriff Bot  <webkit.review.bot@gmail.com>
17841
17842         Unreviewed, rolling out r86447.
17843         http://trac.webkit.org/changeset/86447
17844         https://bugs.webkit.org/show_bug.cgi?id=60809
17845
17846         "Broke some uses of EventSender object on Qt" (Requested by
17847         mwenge on #webkit).
17848
17849         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
17850         (WebCore::WebPage::WebPage):
17851         (WebCore::DumpRenderTree::DumpRenderTree):
17852         (WebCore::DumpRenderTree::initJSObjects):
17853         (WebCore::DumpRenderTree::createWindow):
17854         * DumpRenderTree/qt/DumpRenderTreeQt.h:
17855         (WebCore::DumpRenderTree::eventSender):
17856         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
17857         (webkit_test_plugin_new_instance):
17858         (webkit_test_plugin_set_window):
17859         (webkit_test_plugin_handle_event):
17860
17861 2011-05-13  Robert Hogan  <robert@webkit.org>
17862
17863         Reviewed by Benjamin Poulain.
17864
17865         [Qt] fix http/tests/plugins/plugin-document-has-focus.html
17866
17867         The support for this test added to the unix test plugin here
17868         may allow other platforms to pass it (nearly everyone
17869         skips it). On Qt it required a bit of trickiness with
17870         the page's EventSender object to get it working fully though,
17871         so only unskipping Qt here.
17872
17873         https://bugs.webkit.org/show_bug.cgi?id=60722
17874
17875         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
17876         (WebCore::WebPage::WebPage):
17877         (WebCore::WebPage::eventSender):
17878         (WebCore::DumpRenderTree::DumpRenderTree):
17879         (WebCore::DumpRenderTree::initJSObjects):
17880         (WebCore::DumpRenderTree::createWindow):
17881         * DumpRenderTree/qt/DumpRenderTreeQt.h:
17882         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
17883         (webkit_test_plugin_new_instance):
17884         (webkit_test_plugin_set_window):
17885         (webkit_test_plugin_handle_event):
17886
17887 2011-05-13  Adam Roben  <aroben@apple.com>
17888
17889         Exclude the Leopard NRWT bot from the core builder set
17890
17891         Fixes <http://webkit.org/b/60392> Leopard Intel Release (NRWT) shouldn't be a core builder
17892
17893         Reviewed by Darin Adler.
17894
17895         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
17896         (BuildBot.__init__): Tighten up the Leopard regexp to only allow the non-NRWT Leopard bots
17897         in the core builder set.
17898
17899         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
17900         (BuildBotTest.test_builder_name_regexps): Updated to match the current set of builders and
17901         regexps.
17902
17903 2011-05-13  Adam Roben  <aroben@apple.com>
17904
17905         Group all the NRWT bots together
17906
17907         This also gets the new NRWT bots out of the middle of Apple's bots. Since these new NRWT
17908         bots are experimental, it seems best not to clutter up the non-experimental bots with
17909         experimental ones.
17910
17911         Fixes <http://webkit.org/b/60769> New "(NRWT)" bots are right in the middle of Apple's bots
17912         on build.webkit.org
17913
17914         Reviewed by Dan Bernstein.
17915
17916         * BuildSlaveSupport/build.webkit.org-config/config.json: Moved the "(NRWT)" bots next to the
17917         Chromium testers, which also use new-run-webkit-tests.
17918
17919 2011-05-13  Adam Roben  <aroben@apple.com>
17920
17921         Add a plugin test that dumps the plugin window's rect
17922
17923         Part of the test for <http://webkit.org/b/60194> <rdar://problem/9383760> REGRESSION
17924         (r79167): Windowed plugins in Google Reader don't move when the article list is scrolled
17925
17926         Reviewed by Dan Bernstein.
17927
17928         * DumpRenderTree/TestNetscapePlugIn/Tests/win/DumpWindowRect.cpp: Added.
17929         (DumpWindowRect::DumpWindowRect): Just call up to the base class.
17930         (DumpWindowRect::performWindowGeometryTest): Find our window rect relative to the test
17931         harness window and log it.
17932
17933         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Added DumpWindowRect.
17934
17935 2011-05-13  Adam Roben  <aroben@apple.com>
17936
17937         Move code required for most tests of plugin window geometry into a base class
17938
17939         WebKit2's asynchronous plugin window positioning makes testing window geometry tricky. By
17940         having all the trickiness in a shared base class, writing new window geometry tests will be
17941         much easier.
17942
17943         Fixes <http://webkit.org/b/60776> It's hard to write tests that check plugin window geometry
17944
17945         Reviewed by Anders Carlsson.
17946
17947         * DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowRegionIsSetToClipRect.cpp:
17948         (WindowRegionIsSetToClipRect::WindowRegionIsSetToClipRect): Changed to inherit from the new
17949         WindowGeometryTest base class. Most of the code from this class was moved to the new base
17950         class.
17951         (WindowRegionIsSetToClipRect::performWindowGeometryTest): Renamed from checkWindowRegion.
17952         This is called by the base class when it is safe to check the window's geometry. Note that
17953         the m_didCheckWindowRegion logic was removed completely; it has never been needed and was
17954         just vestigial code from an earlier version of this test.
17955
17956         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Added new
17957         WindowGeometryTest class.
17958
17959         * DumpRenderTree/TestNetscapePlugIn/win/WindowGeometryTest.cpp: Added. Code came from
17960         WindowRegionIsSetToClipRect.
17961         (WindowGeometryTest::finishTest): This is the only change from WindowRegionIsSetToClipRect.
17962         We now call the new virtual performWindowGeometryTest function instead of checkWindowRegion.
17963         Subclasses can implement that function to perform their own tests.
17964
17965         * DumpRenderTree/TestNetscapePlugIn/win/WindowGeometryTest.h: Added.
17966
17967 2011-05-13  Zoltan Herczeg  <zherczeg@webkit.org>
17968
17969         Reviewed by Csaba Osztrogonác.
17970
17971         Missing 2nd #EOF if WebProcess is killed.
17972         https://bugs.webkit.org/show_bug.cgi?id=60761
17973
17974         If the WebProcess is killed, only one #EOF is yielded
17975         by WebKitTestRunner. However, run-webkit-test wait for
17976         a second #EOF, which signals the finishing of pixel dump.
17977         Since this never happens, run-webkit-test waits forever.
17978
17979         * WebKitTestRunner/TestInvocation.cpp:
17980         (WTR::TestInvocation::dump):
17981         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
17982         * WebKitTestRunner/TestInvocation.h:
17983
17984 2011-05-13  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
17985
17986         Reviewed by Csaba Osztrogonác.
17987
17988         [NRWT] chromium-linux tests shouldn't run on other platforms
17989         https://bugs.webkit.org/show_bug.cgi?id=60162
17990
17991         Remove _tests_for_other_platforms defs because r82753
17992         implemented it properly in the baseclass in webkit.py.
17993
17994         * Scripts/webkitpy/layout_tests/port/gtk.py:
17995         * Scripts/webkitpy/layout_tests/port/qt.py:
17996         * Scripts/webkitpy/layout_tests/port/win.py:
17997
17998 2011-05-12  Ryuan Choi  <ryuan.choi@samsung.com>
17999
18000         Reviewed by Antonio Gomes.
18001
18002         [EFL] Remove unused option of EWeblauncher
18003         https://bugs.webkit.org/show_bug.cgi?id=60747
18004
18005         Remove sudo-workaround option.
18006         This option is not used and make other option wrong.
18007
18008         * EWebLauncher/main.c:
18009
18010 2011-05-12  Lucas Forschler  <lforschler@apple.com>
18011
18012         Reviewed by Stephanie Lewis.
18013
18014         Fix config.json to address missing WK2 builder/tester config.
18015         Fix naming of machines (xserve not macpro)
18016
18017         * BuildSlaveSupport/build.webkit.org-config/config.json:
18018
18019 2011-05-12  Lucas Forschler  <lforschler@apple.com>
18020
18021         Reviewed by Stephanie Lewis.
18022
18023         Add SnowLeopard Debug builder and Test bot.
18024         apple-xserve-8 = SL Debug Tester.
18025         apple-xserve-9 = SL Debug Builder.
18026
18027         * BuildSlaveSupport/build.webkit.org-config/config.json:
18028
18029 2011-05-12  Adam Barth  <abarth@webkit.org>
18030
18031         This patch disables sheriffbot's bug postings.  He's just been way too
18032         spammy recently due to test flakiness.
18033
18034         * Scripts/webkitpy/tool/commands/sheriffbot.py:
18035         * Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
18036
18037 2011-05-11  Brian Weinstein  <bweinstein@apple.com>
18038
18039         Reviewed by Anders Carlsson.
18040
18041         After relaunching an unresponsive web process, web page stops responding to mouse movement
18042         https://bugs.webkit.org/show_bug.cgi?id=60669
18043         <rdar://problem/9409934>
18044
18045         This test follows the steps needed to reproduce this bug:
18046         1) Hang the web process
18047         2) While the web process is hung, move the mouse (so WebPageProxy thinks it is processing a mouse move event)
18048         3) Kill and restart the WebProcess
18049         4) Move the mouse
18050
18051         We are testing that we receive mouse move events after the WebProcess has restarted.
18052
18053         * TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash.cpp: Added.
18054         (TestWebKitAPI::didFinishLoadForFrame): Set a flag that we finished loading.
18055         (TestWebKitAPI::processDidBecomeUnresponsive): Kill the web process once we become unresponsive.
18056         (TestWebKitAPI::processDidCrash): Reload the page when the process crashes, which restarts the web process.
18057         (TestWebKitAPI::setPageLoaderClient):
18058         (TestWebKitAPI::TEST):
18059         * TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash_Bundle.cpp: Added.
18060         (TestWebKitAPI::MouseMoveAfterCrashTest::MouseMoveAfterCrashTest):
18061         (TestWebKitAPI::MouseMoveAfterCrashTest::didReceiveMessage):
18062         * TestWebKitAPI/Tests/WebKit2/mouse-move-listener.html: Added.
18063
18064         * TestWebKitAPI/PlatformWebView.h:
18065         * TestWebKitAPI/mac/PlatformWebViewMac.mm:
18066         (TestWebKitAPI::PlatformWebView::simulateMouseMove): Sends the messages that simulate a mouse move.
18067         * TestWebKitAPI/win/PlatformWebViewWin.cpp:
18068         (TestWebKitAPI::PlatformWebView::simulateMouseMove): Ditto.
18069
18070         Add new files.
18071         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
18072         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
18073         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
18074         * TestWebKitAPI/win/copy-resources.cmd:
18075
18076 2011-05-12  Tomasz Morawski  <t.morawski@samsung.com>
18077
18078         Reviewed by Antonio Gomes.
18079
18080         [EFL] Added initialization/shutdown for ecore_file library
18081         https://bugs.webkit.org/show_bug.cgi?id=60457
18082
18083         Added initialization and shuthown for ecore_file library and replaced
18084         realpath function to ecore_file_realpath.
18085
18086         * EWebLauncher/main.c:
18087         (quit):
18088         (findThemePath):
18089         (main):
18090
18091 2011-05-12  Young Han Lee  <joybro@company100.net>
18092
18093         Reviewed by Kenneth Rohde Christiansen.
18094
18095         [Qt] Implement layoutTestController.layerTreeAsText()
18096         https://bugs.webkit.org/show_bug.cgi?id=60367
18097
18098         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
18099         (LayoutTestController::layerTreeAsText):
18100         * DumpRenderTree/qt/LayoutTestControllerQt.h:
18101
18102 2011-05-11  Eric Seidel  <eric@webkit.org>
18103
18104         Reviewed by David Levin.
18105
18106         Teach validate-committer-lists how to validate committers bugzilla_email()
18107         https://bugs.webkit.org/show_bug.cgi?id=60651
18108
18109         It turns out we currently have 9 committers with invalid values for their
18110         bugzilla_email field (meaning we don't know what email is their bugzilla account).
18111         "Alexander Kellett" <lypanov@mac.com>
18112         "Andre Boule" <aboule@apple.com>
18113         "Cameron McCormack" <cam@webkit.org>
18114         "Chris Petersen" <cpetersen@apple.com>
18115         "Eli Fidler" <eli@staikos.net>
18116         "Matt Lilek" <webkit@mattlilek.com>
18117         "Pierre d'Herbemont" <pdherbemont@free.fr>
18118         "Scott Violet" <sky@chromium.org>
18119         "Adam Treat" <treat@kde.org>
18120
18121         Most of those haven't committed in a very long time, so it's not suprising
18122         that we haven't noticed that their bugzilla emails are bogus.
18123
18124         * Scripts/validate-committer-lists:
18125         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
18126         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
18127
18128 2011-05-11  Dmitry Lomov  <dslomov@google.com>
18129
18130         Reviewed by David Levin.
18131
18132         https://bugs.webkit.org/show_bug.cgi?id=59561
18133         Switch TestWebKitAPI to GTest
18134
18135         Fixes Makefiles and WinCairo build as well
18136
18137         * Makefile:
18138         * Scripts/build-api-tests:
18139         * Scripts/run-api-tests:
18140         * TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
18141         * TestWebKitAPI/Test.h:
18142         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
18143         * TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
18144         (TestWebKitAPI::TEST):
18145         * TestWebKitAPI/TestsController.cpp:
18146         (TestWebKitAPI::TestsController::TestsController):
18147         (TestWebKitAPI::TestsController::dumpTestNames):
18148         (TestWebKitAPI::TestsController::runTestNamed):
18149         (TestWebKitAPI::TestsController::runAllTests):
18150         * TestWebKitAPI/TestsController.h:
18151         * TestWebKitAPI/mac/main.mm:
18152         (main):
18153
18154 2011-05-11  Eric Seidel  <eric@webkit.org>
18155
18156         Reviewed by Alexey Proskuryakov.
18157
18158         Add Luke Macpherson to committers.py as a Contributor
18159         https://bugs.webkit.org/show_bug.cgi?id=60639
18160
18161         * Scripts/webkitpy/common/config/committers.py:
18162
18163 2011-05-11  Noam Rosenthal  <noam.rosenthal@nokia.com>
18164
18165         Reviewed by Kenneth Rohde Christiansen.
18166
18167         [Texmap][Qt] Upstream texture-mapper changes from Qt's WebKit2 branch
18168         https://bugs.webkit.org/show_bug.cgi?id=60439
18169
18170         Patch 4/12: Update the build script to enable 3D rendering by default for Qt.
18171
18172         * Scripts/build-webkit:
18173
18174 2011-05-11  Brian Weinstein  <bweinstein@apple.com>
18175
18176         Reviewed by Adam Roben.
18177
18178         REGRESSION (r86161): WKPreferences API test failing on Windows
18179         https://bugs.webkit.org/show_bug.cgi?id=60658
18180
18181         * TestWebKitAPI/Tests/WebKit2/WKPreferences.cpp:
18182         (TestWebKitAPI::TEST):
18183
18184 2011-05-11  Alice Boxhall  <aboxhall@chromium.org>
18185
18186         Reviewed by Ojan Vafai.
18187
18188         Modify jsonresults_unittest.py to use a dict format for its test data, and modify jsonresults.py to flatten hierarchical directory structures in input JSON.
18189         https://bugs.webkit.org/show_bug.cgi?id=60521
18190
18191         First steps towards making the results JSON all hierarchical.
18192
18193         * TestResultServer/model/jsonresults.py:
18194         * TestResultServer/model/jsonresults_unittest.py:
18195
18196 2011-05-11  Kevin Ollivier  <kevino@theolliviers.com>
18197
18198         [wx] Build fix, adding new LayoutTestController method stubs.
18199
18200         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
18201         (LayoutTestController::localStorageDiskUsageForOrigin):
18202         (LayoutTestController::applicationCacheDiskUsageForOrigin):
18203
18204 2011-05-11  Jessie Berlin  <jberlin@apple.com>
18205
18206         Reviewed by Darin Adler.
18207
18208         [WebKit2] exceededDatabaseQuota function on WKPageUIClient needs to provide information
18209         about the usage for both the origin overall and the database in question.
18210         https://bugs.webkit.org/show_bug.cgi?id=60573
18211
18212         * WebKitTestRunner/TestController.cpp:
18213         (WTR::exceededDatabaseQuota):
18214
18215 2011-05-10  Stephen White  <senorblanco@chromium.org>
18216
18217         Reviewed by David Levin.
18218
18219         Adding myself as a reviewer.
18220         https://bugs.webkit.org/show_bug.cgi?id=60587
18221
18222         * Scripts/webkitpy/common/config/committers.py:
18223
18224 2011-05-11  Eric Seidel  <eric@webkit.org>
18225
18226         Reviewed by Csaba Osztrogonác.
18227
18228         queues.webkit.org should show 1000+ instead of 1000 when we hit the query limit
18229         https://bugs.webkit.org/show_bug.cgi?id=60614
18230
18231         AppEngine has a result limit of 1000.  We hit that when trying
18232         to find out how many patches the style-queue (and other queues)
18233         have passed in the last 30 days.  So lets display 1000+ instead
18234         of 1000 to avoid confusion. (Ossy was asking about this in #webkit tonight.)
18235
18236         I also made the status pages show "never" instead of just "ago" when we
18237         haven't seen a pass or boot message in the last 1000 messages.
18238
18239         * QueueStatusServer/handlers/queuestatus.py:
18240         * QueueStatusServer/templates/queuestatus.html:
18241
18242 2011-05-11  Adam Barth  <abarth@webkit.org>
18243
18244         Reviewed by Eric Seidel.
18245
18246         sheriffbot shouldn't spam bugs when it can't figure out who to blame
18247         https://bugs.webkit.org/show_bug.cgi?id=60613
18248
18249         This patch limits the number of bugs sheriffbot will spam to three.
18250         Ideally, we'd come up with some smarter huerstics, but hopefully this
18251         will cut down on the bugmail spam for slow bots.
18252
18253         * Scripts/webkitpy/tool/commands/sheriffbot.py:
18254
18255 2011-05-10  Eric Seidel  <eric@webkit.org>
18256
18257         Reviewed by Adam Barth.
18258
18259         EWS bots should be robust against test-webkitpy hangs
18260         https://bugs.webkit.org/show_bug.cgi?id=60579
18261
18262         Added a killAfterTimeout subroutine, which takes a child_pid
18263         and timeout, and will setup a watchdog process to kill that child
18264         after the timeout.  killAfterTimeout will waitpid on the child.
18265
18266         * EWSTools/start-queue.sh:
18267
18268 2011-05-10  MORITA Hajime  <morrita@google.com>
18269
18270         Unreviewed, moved myself to the reviewers list.
18271
18272         * Scripts/webkitpy/common/config/committers.py:
18273
18274 2011-05-10  Anton D'Auria  <adauria@apple.com>
18275
18276         Reviewed by David Levin.
18277
18278         StorageTracker should report actual local storage usage on disk
18279         https://bugs.webkit.org/show_bug.cgi?id=60558
18280
18281         * DumpRenderTree/LayoutTestController.cpp:
18282         (localStorageDiskUsageForOriginCallback):
18283         (LayoutTestController::staticFunctions):
18284         * DumpRenderTree/LayoutTestController.h:
18285         * DumpRenderTree/chromium/LayoutTestController.cpp:
18286         (LayoutTestController::LayoutTestController):
18287         (LayoutTestController::localStorageDiskUsageForOrigin):
18288         * DumpRenderTree/chromium/LayoutTestController.h:
18289         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
18290         (LayoutTestController::localStorageDiskUsageForOrigin):
18291         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
18292         (LayoutTestController::localStorageDiskUsageForOrigin):
18293         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
18294         (LayoutTestController::localStorageDiskUsageForOrigin):
18295         * DumpRenderTree/qt/LayoutTestControllerQt.h:
18296         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
18297         (LayoutTestController::localStorageDiskUsageForOrigin):
18298
18299 2011-05-10  Anton D'Auria  <adauria@apple.com>
18300
18301         Reviewed by Alexey Proskuryakov.
18302
18303         Add ApplicationCache API to get disk usage per origin
18304         https://bugs.webkit.org/show_bug.cgi?id=57127
18305
18306         Added test that loads two frames with ~15Kb image, checks usage,
18307         deletes app cache, verifies usage is 0.
18308
18309         * DumpRenderTree/LayoutTestController.cpp:
18310         (applicationCacheDiskUsageForOriginCallback):
18311         (LayoutTestController::staticFunctions):
18312         * DumpRenderTree/LayoutTestController.h:
18313         * DumpRenderTree/chromium/LayoutTestController.cpp: Added stub.
18314         (LayoutTestController::LayoutTestController):
18315         (LayoutTestController::applicationCacheDiskUsageForOrigin):
18316         * DumpRenderTree/chromium/LayoutTestController.h:
18317         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: Added stub.
18318         (LayoutTestController::applicationCacheDiskUsageForOrigin):
18319         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
18320         (LayoutTestController::applicationCacheDiskUsageForOrigin):
18321         * DumpRenderTree/qt/LayoutTestControllerQt.cpp: Added stub.
18322         (LayoutTestController::applicationCacheDiskUsageForOrigin):
18323         * DumpRenderTree/qt/LayoutTestControllerQt.h:
18324         * DumpRenderTree/win/LayoutTestControllerWin.cpp: Added stub.
18325         (LayoutTestController::applicationCacheDiskUsageForOrigin):
18326
18327 2011-05-10  Lucas De Marchi  <lucas.demarchi@profusion.mobi>
18328
18329         Unreviewed, adding Rafael Antognolli as a committer.
18330
18331         * Scripts/webkitpy/common/config/committers.py:
18332
18333 2011-05-10  Chris Evans  <cevans@chromium.org>
18334
18335         Reviewed by Adam Barth.
18336
18337         Add WebCore::Setting to block displaying and/or running insecure content on secure pages
18338         https://bugs.webkit.org/show_bug.cgi?id=58378
18339
18340         * DumpRenderTree/chromium/LayoutTestController.cpp:
18341         (LayoutTestController::LayoutTestController):
18342         (LayoutTestController::setAllowDisplayOfInsecureContent):
18343         (LayoutTestController::setAllowRunningOfInsecureContent):
18344         * DumpRenderTree/chromium/LayoutTestController.h:
18345         * DumpRenderTree/chromium/WebPreferences.cpp:
18346         (WebPreferences::reset):
18347         (WebPreferences::applyTo):
18348         * DumpRenderTree/chromium/WebPreferences.h:
18349         Add plumbing for new settings to Chromium DRT.
18350
18351 2011-05-10  Enrica Casucci  <enrica@apple.com>
18352
18353         Reviewed by Beth Dakin.
18354
18355         Added my name to the reviewers list and removed my name from
18356         to committer-only list.
18357
18358         * Scripts/webkitpy/common/config/committers.py:
18359
18360 2011-05-10  David Levin  <levin@chromium.org>
18361
18362         Reviewed by Shinichiro Hamaji.
18363
18364         check-webkit-style shouldn't filter the errors using the modified lines for the xml and test expectations checkers.
18365         https://bugs.webkit.org/show_bug.cgi?id=60466
18366
18367         * Scripts/webkitpy/style/checker.py: (unrelated) style fix.
18368         * Scripts/webkitpy/style/checker_unittest.py: Remove the file_path check since
18369           that property was removed.
18370         * Scripts/webkitpy/style/checkers/test_expectations.py: Turn off the line
18371           filtering for errors.
18372         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Add testing to
18373           verify that the line filtering is turned off.
18374         * Scripts/webkitpy/style/checkers/xml.py: Turn off the line filtering for errors
18375           and remove the unused file_path parameter.
18376         * Scripts/webkitpy/style/checkers/xml_unittest.py: Add testing to verify that the
18377           line filtering is turned off (and fix test_no_error).
18378         * Scripts/webkitpy/style/error_handlers.py: Added a way to turn off the filtering
18379           of errors based on the lines changed in the patch.
18380         * Scripts/webkitpy/style/error_handlers_unittest.py: Added a test for turning off
18381           the filtering.
18382
18383 2011-05-10  Adam Roben  <aroben@apple.com>
18384
18385         Update some plugin tests to work with WebKit2's asynchronous windowed plugin positioning on Windows
18386
18387         Fixes <http://webkit.org/b/60442> <rdar://problem/9403200> REGRESSION (r85795):
18388         platform/win/plugins/window-geometry-initialized-before-set-window.html and
18389         platform/win/plugins/window-region-is-set-to-clip-rect.html failing on Windows 7 Release
18390         (WebKit2 Tests)
18391
18392         Reviewed by Anders Carlsson.
18393
18394         * DumpRenderTree/TestNetscapePlugIn/Tests/win/NPNInvalidateRectInvalidatesWindow.cpp:
18395         (NPNInvalidateRectInvalidatesWindow::NPP_SetWindow): Moved some code from here to the new
18396         WindowedPluginTest::testHarnessWindow function.
18397
18398         * DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowGeometryInitializedBeforeSetWindow.cpp:
18399         (WindowGeometryInitializedBeforeSetWindow::NPP_SetWindow): Only check the window's size, not
18400         its position. In WebKit2, the position won't have been set yet, so we can't test it. But it
18401         turns out only the size mattered for <http://webkit.org/b/46716>, which this test was
18402         created for.
18403
18404         * DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowRegionIsSetToClipRect.cpp:
18405         (WindowRegionIsSetToClipRect::WindowRegionIsSetToClipRect): Converted to be a
18406         WindowedPluginTest, and initialized new members.
18407         (WindowRegionIsSetToClipRect::startTest): Make our plugin window visible, and set things up
18408         to ensure we get a WM_WINDOWPOSCHANGED message about it becoming visible.
18409         (WindowRegionIsSetToClipRect::finishTest): Check that the window region is what we expect,
18410         then clean up and tell the test harness we're done.
18411         (WindowRegionIsSetToClipRect::checkWindowRegion): Moved code to check the window region here
18412         from NPP_SetWindow. Changed to use the fancy new log() function instead of pluginLog().
18413         (WindowRegionIsSetToClipRect::showTestHarnessWindowIfNeeded): Show the test harness's window
18414         if it is hidden. This is needed to ensure we get a WM_WINDOWPOSCHANGED message when our own
18415         window becomes visible.
18416         (WindowRegionIsSetToClipRect::hideTestHarnessWindowIfNeeded): Hide the test harness's window
18417         if it was hidden at the start of the test.
18418         (WindowRegionIsSetToClipRect::wndProc): When our timer fires, tell the UI process to paint.
18419         This should cause our window to become visible and our window region to get set. When we get
18420         a WM_WINDOWPOSCHANGED message, we can finally finish the test.
18421         (WindowRegionIsSetToClipRect::NPP_GetValue): Return our script object.
18422         (WindowRegionIsSetToClipRect::ScriptObject::hasMethod): We only have one method:
18423         "startTest".
18424         (WindowRegionIsSetToClipRect::ScriptObject::invoke): Call through to the PluginTest object.
18425
18426         * DumpRenderTree/TestNetscapePlugIn/win/WindowedPluginTest.cpp:
18427         (WindowedPluginTest::testHarnessWindow):
18428         * DumpRenderTree/TestNetscapePlugIn/win/WindowedPluginTest.h:
18429         Added this helper function. Code came from NPNInvalidateRectInvalidatesWindow.
18430
18431 2011-05-10  Csaba Osztrogonác  <ossy@webkit.org>
18432
18433         Reviewed by Kenneth Rohde Christiansen.
18434
18435         [Qt] Add new qt-4.8 platform for layout testing
18436         https://bugs.webkit.org/show_bug.cgi?id=60544
18437
18438         * Scripts/old-run-webkit-tests:
18439         * Scripts/webkitdirs.pm:
18440
18441 2011-05-09  Sheriff Bot  <webkit.review.bot@gmail.com>
18442
18443         Unreviewed, rolling out r86108.
18444         http://trac.webkit.org/changeset/86108
18445         https://bugs.webkit.org/show_bug.cgi?id=60533
18446
18447         Broke makefile based builds and Cairo (Requested by weinig on
18448         #webkit).
18449
18450         * Scripts/build-api-tests:
18451         * Scripts/run-api-tests:
18452         * TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
18453         * TestWebKitAPI/Test.h:
18454         (TestWebKitAPI::Test::~Test):
18455         (TestWebKitAPI::Test::name):
18456         (TestWebKitAPI::Test::Register::Register):
18457         (TestWebKitAPI::Test::Register::create):
18458         (TestWebKitAPI::Test::Test):
18459         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
18460         * TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
18461         (TestWebKitAPI::TEST):
18462         * TestWebKitAPI/TestsController.cpp:
18463         (TestWebKitAPI::TestsController::TestsController):
18464         (TestWebKitAPI::TestsController::dumpTestNames):
18465         (TestWebKitAPI::TestsController::runTestNamed):
18466         (TestWebKitAPI::TestsController::testFailed):
18467         (TestWebKitAPI::TestsController::registerCreateTestFunction):
18468         * TestWebKitAPI/TestsController.h:
18469         * TestWebKitAPI/mac/main.mm:
18470         (main):
18471
18472 2011-05-09  Hayato Ito  <hayato@chromium.org>
18473
18474         Reviewed by Ojan Vafai.
18475
18476         Skip reftests if pixel tests are disabled.
18477
18478         https://bugs.webkit.org/show_bug.cgi?id=59078
18479
18480         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
18481         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
18482         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
18483         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
18484
18485 2011-05-09  Chang Shu  <cshu@webkit.org>
18486
18487         Reviewed by Kenneth Rohde Christiansen.
18488
18489         [Qt] DumpRenderTreeQt needs an implementation of unmarkText
18490         https://bugs.webkit.org/show_bug.cgi?id=60499
18491
18492         The implementation is similar to gtk.
18493
18494         * DumpRenderTree/qt/TextInputControllerQt.cpp:
18495         (TextInputController::unmarkText):
18496         * DumpRenderTree/qt/TextInputControllerQt.h:
18497
18498 2011-05-09  James Robinson  <jamesr@chromium.org>
18499
18500         Reviewed by Eric Seidel.
18501
18502         [chromium] Teach new-run-webkit-tests to follow symlinks when trying to determine the arch of DumpRenderTree on linux
18503         https://bugs.webkit.org/show_bug.cgi?id=60520
18504
18505         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
18506
18507 2011-05-09  Dmitry Lomov  <dslomov@google.com>
18508
18509         Reviewed by Sam Weinig.
18510
18511         Switch TestWebKitAPI to GTest
18512         https://bugs.webkit.org/show_bug.cgi?id=59561
18513
18514         * Scripts/build-api-tests:
18515         * Scripts/run-api-tests:
18516         * TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
18517         * TestWebKitAPI/Test.h:
18518         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
18519         * TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
18520         (TestWebKitAPI::TEST):
18521         * TestWebKitAPI/TestsController.cpp:
18522         (TestWebKitAPI::TestsController::TestsController):
18523         (TestWebKitAPI::TestsController::dumpTestNames):
18524         (TestWebKitAPI::TestsController::runTestNamed):
18525         (TestWebKitAPI::TestsController::runAllTests):
18526         * TestWebKitAPI/TestsController.h:
18527         * TestWebKitAPI/mac/main.mm:
18528         (main):
18529
18530 2011-05-09  Igor Oliveira  <igor.oliveira@openbossa.org>
18531
18532         Reviewed by Andreas Kling.
18533
18534         [Qt] DRT eventSender.addTouchPoint method is not setting the start position
18535         https://bugs.webkit.org/show_bug.cgi?id=60487
18536
18537         When adding a new touch point, the start position of this point needs to be set.
18538
18539         * DumpRenderTree/qt/EventSenderQt.cpp:
18540         (EventSender::addTouchPoint):
18541
18542 2011-05-09  Robert Hogan  <robert@webkit.org>
18543
18544         Reviewed by Adam Roben.
18545
18546         https://bugs.webkit.org/show_bug.cgi?id=57137
18547
18548         Unskip plugins/embed-prefers-plugins-for-images.html
18549
18550         Add image/png mime-type to test netscape plugin on Unix.
18551
18552         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
18553         (NP_GetMIMEDescription):
18554
18555 2011-05-09  Alice Boxhall  <aboxhall@chromium.org>
18556
18557         Reviewed by Ojan Vafai.
18558
18559         Pull jsonresults_unittest test cases out into separate test methods.
18560         https://bugs.webkit.org/show_bug.cgi?id=60418
18561
18562         * TestResultServer/model/jsonresults_unittest.py:
18563
18564 2011-05-09  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
18565
18566         Reviewed by Andreas Kling.
18567
18568         [Qt] fix fast/dom/Geolocation/window-close-crash.html
18569         https://bugs.webkit.org/show_bug.cgi?id=60458
18570
18571         Update all pages with Mock position/error/permission.
18572
18573         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
18574         (LayoutTestController::setGeolocationPermission):
18575         (LayoutTestController::setMockGeolocationError):
18576         (LayoutTestController::setMockGeolocationPosition):
18577
18578 2011-05-09  Shishir Agrawal  <shishir@chromium.org>
18579
18580         Reviewed by Tony Gentilcore.
18581
18582         Implement Page Visibility API.
18583         https://bugs.webkit.org/show_bug.cgi?id=54181
18584
18585         * DumpRenderTree/LayoutTestController.cpp:
18586         (setPageVisibilityCallback):
18587         (resetPageVisibilityCallback):
18588         (LayoutTestController::staticFunctions):
18589         * DumpRenderTree/LayoutTestController.h:
18590         (LayoutTestController::setPageVisibility):
18591         (LayoutTestController::resetPageVisibility):
18592         * DumpRenderTree/chromium/LayoutTestController.cpp:
18593         (LayoutTestController::LayoutTestController):
18594         (LayoutTestController::resetPageVisibility):
18595         (LayoutTestController::setPageVisibility):
18596         * DumpRenderTree/chromium/LayoutTestController.h:
18597
18598 2011-05-07  Adam Barth  <abarth@webkit.org>
18599
18600         Reviewed by Eric Seidel.
18601
18602         Testing EWS spins on patches with a large number of failures
18603         https://bugs.webkit.org/show_bug.cgi?id=60441
18604
18605         In cases where the tree has a small number of persistent failures but a
18606         patch has a large number of failures, we weren't believing any results.
18607         That lead to us retrying the patch forever while we waited for
18608         trustworthy results that would never come.
18609
18610         This patch loosens the semantics of unexpected_failures to return a
18611         subset (rather than the exact set) of unexpected failures.  In the case
18612         where the tree has a bounded number of failures and the patch has an
18613         unbounded number of failures, we can accurately compute such a subset
18614         and reject the patch.
18615
18616         * Scripts/webkitpy/tool/bot/expectedfailures.py:
18617         * Scripts/webkitpy/tool/bot/expectedfailures_unittest.py:
18618         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
18619         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
18620         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
18621
18622 2011-05-07  Joe Mason  <jmason@rim.com>
18623
18624         Reviewed by Daniel Bates.
18625
18626         FrameLoader::isProcessingUserGesture is wrong in dispatchWillPerformClientRedirect
18627         https://bugs.webkit.org/show_bug.cgi?id=52211
18628
18629         Dump isUserProcessingGesture in willPerformClientRedirect so that the layout tests can be extended.
18630         (For the Mac and Chromium ports - other ports don't support dumping user gestures in DRT.)
18631
18632         * DumpRenderTree/chromium/WebViewHost.cpp:
18633         (WebViewHost::willPerformClientRedirect):
18634         (WebViewHost::didStartProvisionalLoad):
18635         * DumpRenderTree/mac/FrameLoadDelegate.mm:
18636         (-[FrameLoadDelegate webView:willPerformClientRedirectToURL:delay:fireDate:forFrame:]):
18637
18638 2011-05-07  Daniel Bates  <dbates@webkit.org>
18639
18640         Attempt to fix the build; remove some more references to isTiger(), which was removed in changeset 86007
18641         <http://trac.webkit.org/changeset/86007> (bug #60438).
18642
18643         * Scripts/old-run-webkit-tests:
18644           - Removed reference to isTiger() from openDumpTool(), captureSavedCrashLog(), and
18645             file local variable.
18646
18647 2011-05-07  Pratik Solanki  <psolanki@apple.com>
18648
18649         Reviewed by Darin Adler.
18650
18651         Remove references to Tiger in scripts
18652         https://bugs.webkit.org/show_bug.cgi?id=60438
18653
18654         Remove isTiger() and all of the code that depends on it.
18655
18656         * Scripts/build-webkit:
18657         * Scripts/ensure-valid-python:
18658         * Scripts/gdb-safari:
18659         * Scripts/webkitdirs.pm:
18660
18661 2011-05-06  David Grogan  <dgrogan@chromium.org>
18662
18663         Reviewed by Eric Seidel.
18664
18665         update link to message about mac 64bit git binaries
18666         https://bugs.webkit.org/show_bug.cgi?id=60411
18667
18668         The original message implied (correctly, at-the-time) that
18669         http://code.google.com/p/git-osx-installer/ doesn't provide 64-bit
18670         binaries. The new message notes that they do. This might prevent someone
18671         else from spending a few minutes confused.
18672
18673         * Scripts/webkitpy/common/checkout/scm/git.py:
18674
18675 2011-05-06  Dirk Pranke  <dpranke@chromium.org>
18676
18677         Reviewed by Ojan Vafai.
18678
18679         new-run-webkit-tests: crashes in race with crashing DRT
18680         https://bugs.webkit.org/show_bug.cgi?id=60238
18681
18682         This fixes a race where if we wrote the URL to DRT and
18683         DRT crashed immediately, we could get a broken pipe
18684         error back during the write() and not handle it correctly,
18685         causing NRWT to also crash.
18686
18687         * Scripts/webkitpy/layout_tests/port/server_process.py:
18688         * Scripts/webkitpy/layout_tests/port/webkit.py:
18689
18690 2011-05-06  Adam Barth  <abarth@webkit.org>
18691
18692         Reviewed by Eric Seidel.
18693
18694         Add a --skip-integrationtests mode to test-webkitpy
18695         https://bugs.webkit.org/show_bug.cgi?id=60354
18696
18697         These two test suites account for 2/3 of test-webkitpy's running time
18698         because they're not actually unit tests.  They're integration tests.
18699         This patch adds the ability to skip integration tests by passing a
18700         command line argument to test-webkitpy.  The buildbots will continue to
18701         run these integration tests to make sure they don't regress.  This
18702         option is for use by developers when working on parts of the code that
18703         are well-covered by unit tests.
18704
18705         * Scripts/webkitpy/common/system/file_lock_integrationtest.py: Copied from Tools/Scripts/webkitpy/common/system/file_lock_unittest.py.
18706         * Scripts/webkitpy/common/system/file_lock_unittest.py: Removed.
18707         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py: Copied from Tools/Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py.
18708         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py: Removed.
18709         * Scripts/webkitpy/test/main.py:
18710
18711 2011-05-06  Adam Barth  <abarth@webkit.org>
18712
18713         Reviewed by Eric Seidel.
18714
18715         The results.zip generated by the Chromium EWS are too big
18716         https://bugs.webkit.org/show_bug.cgi?id=60352
18717
18718         Currently results.zip created by the Chromium EWS bot contain roughly
18719         1000 tests that we expect to fail.  These test results are not helpful
18720         to contributors.
18721
18722         There are several approaches to solving this problem:
18723
18724         1) We could be more selective about which files to include in the zip.
18725         2) We could ask new-run-webkit-tests not to generate results for
18726            expected failures.
18727         3) We could ask new-run-webkit-tests not to run tests that we expect to
18728            fail.
18729
18730         This patch takes approach (3).  Currently, we're ignoring the outcome
18731         of failing tests anyway, so there's not much point in actually running
18732         them.  As a side benefit, this will speed up the test run by about 6%.
18733         In the future, we might revisit this decision and choose another
18734         approach.
18735
18736         * Scripts/webkitpy/common/config/ports.py:
18737         * Scripts/webkitpy/common/config/ports_unittest.py:
18738
18739 2011-05-06  Gabor Rapcsanyi  <rgabor@webkit.org>
18740
18741         Reviewed by Csaba Osztrogonác.
18742
18743         [Qt] Use qt-arm platform instead of qt-linux on ARM platform.
18744
18745         * Scripts/old-run-webkit-tests:
18746
18747 2011-05-06  Adam Barth  <abarth@webkit.org>
18748
18749         Reviewed by Ojan Vafai.
18750
18751         Add an option to NRWT to skip failing tests
18752         https://bugs.webkit.org/show_bug.cgi?id=60295
18753
18754         This option is similar to the way old-run-webkit-tests works.  We're
18755         planning to use this option on the testing EWS bots because they don't
18756         use information from failing tests yet.
18757
18758         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
18759         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
18760         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
18761
18762 2011-05-04  Philippe Normand  <pnormand@igalia.com>
18763
18764         Reviewed by Dimitri Glazkov.
18765
18766         Implement LayoutTestController::pseudoShadowId()
18767         https://bugs.webkit.org/show_bug.cgi?id=60034
18768
18769         shadowPseudoId support in the various LayoutTestController
18770         implementations.
18771
18772         * DumpRenderTree/LayoutTestController.cpp:
18773         (shadowPseudoIdCallback):
18774         (LayoutTestController::staticFunctions):
18775         * DumpRenderTree/LayoutTestController.h:
18776         * DumpRenderTree/chromium/LayoutTestController.cpp:
18777         (LayoutTestController::LayoutTestController):
18778         (LayoutTestController::shadowPseudoId):
18779         * DumpRenderTree/chromium/LayoutTestController.h:
18780         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
18781         (LayoutTestController::shadowPseudoId):
18782         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
18783         (LayoutTestController::shadowPseudoId):
18784         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
18785         (LayoutTestController::shadowPseudoId):
18786         * DumpRenderTree/qt/LayoutTestControllerQt.h:
18787         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
18788         (LayoutTestController::shadowPseudoId):
18789         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
18790         (LayoutTestController::shadowPseudoId):
18791         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
18792         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
18793         (WTR::LayoutTestController::shadowPseudoId):
18794         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
18795
18796 2011-05-06  Adam Barth  <abarth@webkit.org>
18797
18798         EWS can't reject patches
18799         https://bugs.webkit.org/show_bug.cgi?id=60345
18800
18801         We're supposed to use patch.id() for the attachment ID, silly.
18802
18803         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
18804         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
18805
18806 2011-05-05  Steve Lacey  <sjl@chromium.org>
18807
18808         Reviewed by Ojan Vafai.
18809
18810         Fix generation of results html file by passing relative path to scm
18811         methods.
18812         Generate diff images for results html file.
18813
18814         https://bugs.webkit.org/show_bug.cgi?id=60086
18815
18816         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
18817         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
18818
18819 2011-05-05  Andy Estes  <aestes@apple.com>
18820
18821         Reviewed by Daniel Bates.
18822
18823         update-webkit should handle Internal using a different VCS than OpenSource.
18824         https://bugs.webkit.org/show_bug.cgi?id=60310
18825
18826         isGit() and isSVN() cache their results, so Internal will always be
18827         updated using the VCS commands used for OpenSource. If Internal uses a
18828         different VCS, this will fail. Fix this by calling isGitDirectory() and
18829         isSVNDirectory() instead, which do not cache their results.
18830
18831         * Scripts/update-webkit:
18832
18833 2011-05-05  Dirk Pranke  <dpranke@chromium.org>
18834
18835         Reviewed by Ojan Vafai.
18836
18837         nrwt: revamp integration tests incl. tests for servers
18838         https://bugs.webkit.org/show_bug.cgi?id=59985
18839
18840         This change adds the ability to run "integration tests" that
18841         depend on (and affect) the local environment on your machine.
18842         These tests can be run manually by invoking python on the
18843         test file in question and passing the '-i' flag, but are
18844         not run automatically as part of test-webkitpy.
18845
18846         Integration tests are identified by being methods named as
18847         "integration_test_*" instead of "test_*", and we use a
18848         custom TestLoader instead of the default one provided by
18849         unittest.
18850
18851         This change also adds a protected "_default_to_apache" method
18852         to the port object so we can do the right thing when starting
18853         and stopping web servers with less mechanism.
18854
18855         * Scripts/webkitpy/layout_tests/port/base.py:
18856         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
18857         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
18858         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
18859         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
18860         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
18861
18862 2011-05-05  Adam Roben  <aroben@apple.com>
18863
18864         Robustify DRT's pixel-dumping code against semi-random alpha channel values on Windows
18865
18866         DRT generates a pixel dump on Windows using the WM_PRINT message. This message (just like
18867         most of GDI) doesn't write anything meaningful into the alpha channel. In some cases, the
18868         alpha channel will even get filled with zeros (like when ::GradientFill is used). To
18869         safeguard against this, we fill the alpha channel with constant full opacity after the
18870         WM_PRINT message is handled to ensure we end up with an opaque image. (Note that we can't
18871         just ignore the alpha channel when creating the PNG; Mac's pixel results have an alpha
18872         channel, and we want to be able to compare Mac and Windows pixel results, so Windows pixel
18873         results need an alpha channel, too.)
18874
18875         Fixes <http://webkit.org/b/60300> draws-gradient.html's pixel results can't be correctly
18876         generated due to alpha channel weirdness
18877
18878         Reviewed by Sam Weinig.
18879
18880         * DumpRenderTree/win/PixelDumpSupportWin.cpp:
18881         (makeAlphaChannelOpaque): Added this helper function to fill a bitmap's alpha channel with
18882         constant full opacity.
18883         (createBitmapContextFromWebView): Make the bitmap's alpha channel opaque before wrapping it
18884         in a CGContext or cairo_surface.
18885
18886 2011-05-05  Adam Roben  <aroben@apple.com>
18887
18888         Add support for layoutTestController.dumpAsText(true) to WebKitTestRunner
18889
18890         Passing true tells the test harness to dump pixels even though we're not dumping a render
18891         tree.
18892
18893         Fixes <http://webkit.org/b/60298> Calling layoutTestController.dumpAsText(true) in
18894         WebKitTestRunner doesn't result in pixel results being generated
18895
18896         Reviewed by Sam Weinig.
18897
18898         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl: Added dumpPixels
18899         argument to dumpAsText.
18900
18901         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
18902         (WTR::InjectedBundlePage::dump): Updated for dumpAsText change.
18903
18904         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
18905         (WTR::LayoutTestController::dumpAsText): Store the dumpPixels argument.
18906
18907         * WebKitTestRunner/InjectedBundle/LayoutTestController.h: Updated for IDL change.
18908
18909         * WebKitTestRunner/win/InjectedBundle.vcproj: Added .idl files as a convenience for Windows
18910         developers.
18911
18912 2011-05-05  Dirk Pranke  <dpranke@chromium.org>
18913
18914         Reviewed by Ojan Vafai.
18915
18916         new-run-webkit-tests: merge os, version fields in test config
18917         https://bugs.webkit.org/show_bug.cgi?id=60004
18918
18919         This change removes the 'o/s' field from the TestConfiguration
18920         object, and changes the 'MAC', 'WIN', and 'LINUX' modifiers in
18921         the expectations file to macros that expand to the version names
18922         for each platform.
18923
18924         This means that you can once again put any subset of the
18925         operating system versions on a single line, e.g.:
18926           BUGX LEOPARD XP LINUX : fast/html/foo.html = FAIL
18927
18928         This change will break any existing test_expectations.txt
18929         file that uses the "OS-VERSION" macros, which are no longer
18930         needed. It turns out they were never needed, as just VERSION
18931         by itself was sufficient. The test_expectations.txt file should
18932         be updated accordingly prior to landing this change.
18933
18934         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
18935         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
18936         * Scripts/webkitpy/layout_tests/port/base.py:
18937
18938 2011-05-05  Tony Chang  <tony@chromium.org>
18939
18940         Reviewed by Adam Barth.
18941
18942         [chromium] switch from Lucida to free fonts for Hindi and Thai
18943         https://bugs.webkit.org/show_bug.cgi?id=60287
18944
18945         The fonts are provided by the ttf-thai-tlwg and ttf-indic-fonts
18946         packages on Lucid. There are actually multiple Thai fonts in the
18947         package, so I just picked one.
18948
18949         * DumpRenderTree/chromium/TestShellGtk.cpp:
18950         (setupFontconfig):
18951
18952 2011-05-05  Tony Chang  <tony@chromium.org>
18953
18954         Reviewed by Ojan Vafai.
18955
18956         [chromium] google-chrome-64 builds should now fallback to chromium-linux-x86_64 results
18957         https://bugs.webkit.org/show_bug.cgi?id=60290
18958
18959         It used to be that google-chrome-32 was hardy32, google-chrome-64 was
18960         hardy64, chromium-linux was hardy32 and chromium-linux-x86_64 was
18961         lucid64.  Because of that, google-chrome-64 used to skip the
18962         chromium-linux-x86_64 results since they were more incorrect.
18963
18964         Now that we're switching to lucid for google-chrome, we want
18965         google-chrome-64 to fall back to chromium-linux-x86_64.
18966
18967         * Scripts/webkitpy/layout_tests/port/google_chrome.py:
18968
18969 2011-05-05  Dirk Pranke  <dpranke@chromium.org>
18970
18971         Reviewed by Ojan Vafai.
18972
18973         new-run-webkit-tests: merge TestExpectations, TestExpectationsFile classes
18974         https://bugs.webkit.org/show_bug.cgi?id=60002
18975
18976         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
18977         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
18978         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
18979         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
18980         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
18981         * Scripts/webkitpy/style/checkers/test_expectations.py:
18982
18983 2011-05-05  Eric Seidel  <eric@webkit.org>
18984
18985         Reviewed by Adam Barth.
18986
18987         commit-queue should reject patches which fail ewses
18988         https://bugs.webkit.org/show_bug.cgi?id=47534
18989
18990         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
18991         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
18992         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
18993
18994 2011-05-05  Chang Shu  <cshu@webkit.org>
18995
18996         Reviewed by Andreas Kling.
18997
18998         [Qt][WK2] Support PageGroup in QGraphicsWKView/QWKPage constructors
18999         https://bugs.webkit.org/show_bug.cgi?id=59638
19000
19001         1. Call QGraphicsWKView constructor with pageGroupRef.
19002         2. Disable setFontFamily calls as it breaks Qt test fonts.
19003
19004         * WebKitTestRunner/TestController.cpp:
19005         (WTR::TestController::resetStateToConsistentValues):
19006         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
19007         (WTR::WebView::WebView):
19008         (WTR::PlatformWebView::PlatformWebView):
19009
19010 2011-05-05  Joseph Pecoraro  <joepeck@webkit.org>
19011
19012         Reviewed by David Kilzer.
19013
19014         Add --only-webkit option to build-webkit
19015         https://bugs.webkit.org/show_bug.cgi?id=60245
19016
19017         * Scripts/build-webkit: If --only-webkit was specified
19018         just build the WebKit project, and not its dependencies.
19019
19020 2011-05-05  Jeff Miller  <jeffm@apple.com>
19021
19022         Reviewed by Darin Adler.
19023
19024         Add didNotHandleWheelEvent callback to WKPageUIClient
19025         https://bugs.webkit.org/show_bug.cgi?id=60203
19026
19027         * MiniBrowser/gtk/BrowserWindow.c:
19028         (browserWindowUIClientInit): Stub out didNotHandleWheelEvent callback.
19029         * MiniBrowser/mac/BrowserWindowController.m:
19030         (-[BrowserWindowController awakeFromNib]): Stub out didNotHandleWheelEvent callback.
19031         * MiniBrowser/win/BrowserView.cpp:
19032         (BrowserView::create): Stub out didNotHandleWheelEvent callback.
19033         * WebKitTestRunner/TestController.cpp:
19034         (WTR::TestController::createOtherPage): Stub out didNotHandleWheelEvent callback.
19035         (WTR::TestController::initialize): Stub out didNotHandleWheelEvent callback.
19036
19037 2011-05-04  Eric Seidel  <eric@webkit.org>
19038
19039         Reviewed by Adam Barth.
19040
19041         Add more missing imports from scm refactor
19042         https://bugs.webkit.org/show_bug.cgi?id=60239
19043
19044         * Scripts/webkitpy/common/checkout/scm/git.py:
19045         * Scripts/webkitpy/common/checkout/scm/svn.py:
19046
19047 2011-05-04  Eric Seidel  <eric@webkit.org>
19048
19049         Reviewed by Adam Barth.
19050
19051         sheriff-bot is having more exception trouble
19052         https://bugs.webkit.org/show_bug.cgi?id=60233
19053
19054         A webpage-forced build was causing SB to throw an exception.
19055         We've long had this bug, but we just didn't notice it until now.
19056
19057         In debugging this I found we fetch an absurd amount of data now
19058         that we use buildbot's json (instead of xmlrpc), so I added
19059         the filter=1 parameter to reduce it a little.  That required
19060         me to handle the case where filter=1 would strip
19061         result=0 in some cases (it's unclear why it does that).
19062
19063         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
19064         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
19065
19066 2011-05-04  Cris Neckar  <cdn@chromium.org>
19067
19068         Unreviewed, Adding Cris Neckar as a committer
19069
19070         * Scripts/webkitpy/common/config/committers.py:
19071
19072 2011-05-04  Siddharth Mathur  <siddharth.mathur@nokia.com>
19073
19074         Reviewed by Laszlo Gombos.
19075
19076         [Qt][Wk2][Symbian] Capabilities for Qt MiniBrowser and QtWebProcess on Symbian
19077         https://bugs.webkit.org/show_bug.cgi?id=59462
19078
19079         * MiniBrowser/qt/MiniBrowser.pro: Add Symbian OS capabilities to executable. Also enable data paging.
19080
19081 2011-05-04  Eric Seidel  <eric@webkit.org>
19082
19083         Reviewed by Adam Barth.
19084
19085         The testing EWS bots should upload layout-test-results.zip to bugs after failures
19086         https://bugs.webkit.org/show_bug.cgi?id=60223
19087
19088         This required sharing a bit of code between the commit-queue
19089         (which already knew how to do this) and the new EWS testing bots.
19090
19091         In the process I also cleaned up EWS testing a little and
19092         removed some dead code from the commit-queue.
19093
19094         * Scripts/webkitpy/common/config/committervalidator.py:
19095         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
19096         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
19097         * Scripts/webkitpy/tool/commands/queues.py:
19098         * Scripts/webkitpy/tool/commands/queues_unittest.py:
19099
19100 2011-05-04  James Kozianski  <koz@chromium.org>
19101
19102         Reviewed by Eric Seidel.
19103
19104         Add method to make a Port able to retrieve TestOutputSets from its builders.
19105         https://bugs.webkit.org/show_bug.cgi?id=58211
19106
19107         This will be used in the new rebaseline script - TestOutputs are
19108         potential new baselines.
19109
19110         * Scripts/webkitpy/common/config/urls.py:
19111         * Scripts/webkitpy/layout_tests/port/base.py:
19112         * Scripts/webkitpy/layout_tests/port/chromium.py:
19113         * Scripts/webkitpy/layout_tests/port/factory.py:
19114         * Scripts/webkitpy/layout_tests/port/webkit.py:
19115
19116 2011-05-04  Eric Seidel  <eric@webkit.org>
19117
19118         Reviewed by Ojan Vafai.
19119
19120         sheriff-bot can't parse results.html
19121         https://bugs.webkit.org/show_bug.cgi?id=60196
19122
19123         The problem was that the NRWT bots are currently core
19124         and so sheriff-bot was trying to diagnose failures on
19125         them.  It was trying to parse their results.html file
19126         and BeautifulSoup was throwing an exception.
19127
19128         It seems our version of BeautifulSoup was quite old
19129         and unable to handle <script> tags correctly.
19130         This moves us to 3.2 which doesn't seem to have this bug.
19131
19132         * Scripts/webkitpy/thirdparty/BeautifulSoup.py:
19133
19134 2011-05-04  Adam Barth  <abarth@webkit.org>
19135
19136         Reviewed by Eric Seidel.
19137
19138         Make the EWS test logging less verbose
19139         https://bugs.webkit.org/show_bug.cgi?id=60201
19140
19141         This should help us generate less than a hundred MB of logs per hour.
19142
19143         * Scripts/webkitpy/common/config/ports.py:
19144         * Scripts/webkitpy/common/config/ports_unittest.py:
19145
19146 2011-05-04  Tao Bai  <michaelbai@chromium.org>
19147
19148         Reviewed by David Kilzer.
19149
19150         Populate touch-icon url to FrameLoaderClient
19151         https://bugs.webkit.org/show_bug.cgi?id=59143
19152
19153         Disable touch icon by default.
19154
19155         * Scripts/build-webkit:
19156
19157 2011-05-04  Justin Novosad  <junov@chromium.org>
19158
19159         Reviewed by Kenneth Russell.
19160
19161         [Chromium] Add gpu acceleration flags to DumpRenderTree
19162         https://bugs.webkit.org/show_bug.cgi?id=60168
19163         Adding support for command line flags
19164         enable-accelerated-2d-canvas an
19165         enable-accelerated-drawing in DumpRenderTree.
19166
19167         * DumpRenderTree/chromium/DumpRenderTree.cpp:
19168         (main):
19169         * DumpRenderTree/chromium/TestShell.cpp:
19170         (TestShell::TestShell):
19171         (TestShell::resetWebSettings):
19172         * DumpRenderTree/chromium/TestShell.h:
19173         (TestShell::setLegacyAccelerated2dCanvasEnabled):
19174         (TestShell::setAcceleratedDrawingEnabled):
19175         * DumpRenderTree/chromium/WebPreferences.cpp:
19176         (WebPreferences::reset):
19177         (WebPreferences::applyTo):
19178         * DumpRenderTree/chromium/WebPreferences.h:
19179
19180 2011-05-04  Ademar de Souza Reis Jr.  <ademar.reis@openbossa.org>
19181
19182         Reviewed by Csaba Osztrogonác.
19183
19184         [Qt] run-qtwebkit-tests misses exec bit
19185         https://bugs.webkit.org/show_bug.cgi?id=60188
19186
19187         Also removing an unecessary import (mostly to let this bugfix goes
19188         through the system, as this is just a chmod).
19189
19190         * Scripts/run-qtwebkit-tests:
19191
19192 2011-05-04  Adam Barth  <abarth@webkit.org>
19193
19194         Fix import of ircproxy module to actually import the module.  I don't
19195         really understand how this code could have worked before.
19196
19197         * Scripts/webkitpy/common/host.py:
19198
19199 2011-05-04  Siddharth Mathur  <siddharth.mathur@nokia.com>
19200
19201         Reviewed by Andreas Kling.
19202
19203         [Qt][Wk2][Symbian] Minor UI facelift for MiniBrowser
19204         https://bugs.webkit.org/show_bug.cgi?id=60172
19205
19206         * MiniBrowser/qt/BrowserWindow.cpp: Bring in UI layout changes from QtTestBrowser
19207         (BrowserWindow::BrowserWindow):
19208
19209 2011-05-04  Mihai Parparita  <mihaip@chromium.org>
19210
19211         Reviewed by Ojan Vafai.
19212
19213         [NRWT] --new-baseline should put baselines in the directory specified by --additional-platform-directory
19214         https://bugs.webkit.org/show_bug.cgi?id=60088
19215
19216         Makes --new-baseline put new baselines into the first entry in the
19217         baseline search path, including any additional directories passed in via
19218         --additional-platform-directory.
19219
19220         Simplifies baseline_path() by making it return the first entry in the
19221         baseline search path instead of requiring each port to have an
19222         implementation (should not result in any behavior change).
19223
19224         * Scripts/webkitpy/layout_tests/port/base.py:
19225         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
19226         * Scripts/webkitpy/layout_tests/port/chromium.py:
19227         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
19228         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
19229         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
19230         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
19231         * Scripts/webkitpy/layout_tests/port/mac.py:
19232         * Scripts/webkitpy/layout_tests/port/test.py:
19233         * Scripts/webkitpy/layout_tests/port/webkit.py:
19234
19235 2011-05-04  Adam Barth  <abarth@webkit.org>
19236
19237         Reviewed by Eric Seidel.
19238
19239         Fix circular dependency in webkitpy
19240         https://bugs.webkit.org/show_bug.cgi?id=60075
19241
19242         These functions don't belong in bugzilla.py.  They only exist there
19243         because they are old.  Really, these functions shouldn't be free
19244         functions at all, but that's a patch for another day.
19245
19246         * Scripts/webkitpy/common/checkout/changelog.py:
19247         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
19248         * Scripts/webkitpy/common/checkout/checkout.py:
19249         * Scripts/webkitpy/common/config/urls.py:
19250         * Scripts/webkitpy/common/net/bugzilla/__init__.py:
19251         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
19252         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
19253         * Scripts/webkitpy/style/checkers/changelog.py:
19254         * Scripts/webkitpy/tool/bot/irc_command.py:
19255         * Scripts/webkitpy/tool/bot/sheriff.py:
19256         * Scripts/webkitpy/tool/commands/upload.py:
19257
19258 2011-05-03  Pratik Solanki  <psolanki@apple.com>
19259
19260         Reviewed by Dan Bernstein.
19261
19262         https://bugs.webkit.org/show_bug.cgi?id=60141
19263         Remove reference to libWebKitSystemInterfaceTiger.a from build-webkit
19264
19265         * Scripts/build-webkit:
19266
19267 2011-05-03  Mihai Parparita  <mihaip@chromium.org>
19268
19269         Reviewed by Ojan Vafai.
19270
19271         Update rebaseline queue server to handle hierarchical test results
19272         https://bugs.webkit.org/show_bug.cgi?id=60063
19273
19274         Update test result parsing done by the rebaseline queue server to handle
19275         the hierarchical test result output that is generated as of r85254.
19276
19277         * RebaselineQueueServer/templates/builder-queue-edit.html:
19278
19279 2011-05-03  Eric Seidel  <eric@webkit.org>
19280
19281         Reviewed by Adam Roben.
19282
19283         webkit-patch post-commits HEAD^.. throws an exception
19284         https://bugs.webkit.org/show_bug.cgi?id=60027
19285
19286         * Scripts/webkitpy/common/checkout/scm/git.py:
19287
19288 2011-05-03  Luiz Agostini  <luiz.agostini@openbossa.org>
19289
19290         Reviewed by Kenneth Rohde Christiansen.
19291
19292         [Qt] Layout tests may timeout if loadFinished receives false
19293         https://bugs.webkit.org/show_bug.cgi?id=60064
19294
19295         When using layoutTestController.queueLoad(), if QWebPage::loadFinished(bool) receives false
19296         then the layout test may timout. It happens that QWebPage::loadFinished() receives false when
19297         a response is ignored, like it should be for 204 responses. It is a problem for creating
19298         tests for 204 responses for example.
19299
19300         This change has been tested and it causes no regressions in layout tests.
19301
19302         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
19303         (LayoutTestController::maybeDump):
19304
19305 2011-05-03  Dirk Pranke  <dpranke@chromium.org>
19306
19307         Unreviewed, rolling out r85659.
19308         http://trac.webkit.org/changeset/85659
19309         https://bugs.webkit.org/show_bug.cgi?id=59977
19310
19311         restarting isn't working correctly
19312
19313         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
19314         * Scripts/webkitpy/layout_tests/port/http_server.py:
19315         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
19316         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
19317
19318 2011-05-03  Dirk Pranke  <dpranke@chromium.org>
19319
19320         Reviewed by Ojan Vafai.
19321
19322         new-run-webkit-tests: fix http server startup/shutdown
19323
19324         Previous versions of the code had three problems that made startup
19325         and shutdown flaky. The first is that it would throw exceptions
19326         if it couldn't delete stale log files, which was overly
19327         paranoid. The second is that some of the exceptions weren't
19328         defined properly. The third, and most important, is that it was
19329         using urllib to check if ports were available, which was leaving
19330         sockets in a half-closed state, and keeping ports from being
19331         reused. By switching to raw sockets, we are able to now reliably
19332         restart.
19333
19334         This change also switches the code to using Executives to stop
19335         processes, which will let us delete a bunch of code in the
19336         port/* implementations and fix a weird layering problem in a
19337         subsequent patch.
19338
19339         https://bugs.webkit.org/show_bug.cgi?id=59977
19340
19341         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
19342         * Scripts/webkitpy/layout_tests/port/http_server.py:
19343         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
19344         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
19345
19346 2011-05-03  Dirk Pranke  <dpranke@chromium.org>
19347
19348         Reviewed by Adam Barth.
19349
19350         fix failing printing_unittest on win32
19351         https://bugs.webkit.org/show_bug.cgi?id=59963
19352
19353         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
19354
19355 2011-05-03  Adam Roben  <aroben@apple.com>
19356
19357         Fix most strict PassOwnPtr violations on Windows
19358
19359         Fixes <http://webkit.org/b/60025> Windows should (almost) build with strict PassOwnPtr
19360         enabled
19361
19362         Reviewed by Anders Carlsson.
19363
19364         * WebKitAPITest/Test.h: Use adoptPtr.
19365
19366 2011-05-03  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
19367
19368         Reviewed by Andreas Kling.
19369
19370         Add myself to the committers list.
19371
19372         * Scripts/webkitpy/common/config/committers.py:
19373
19374 2011-05-03  James Robinson  <jamesr@chromium.org>
19375
19376         Reviewed by Kenneth Russell.
19377
19378         [chromium] Layout test animations/3d/replace-filling-transform.html failing
19379         https://bugs.webkit.org/show_bug.cgi?id=59839
19380
19381         Add animations/3d to the list of tests run in the chromium-gpu configuration since these tests exercise the compositor.
19382
19383         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
19384
19385 2011-05-03  Lucas De Marchi  <lucas.demarchi@profusion.mobi>
19386
19387         Unreviewed, rolling out r85471.
19388         http://trac.webkit.org/changeset/85471
19389         https://bugs.webkit.org/show_bug.cgi?id=59922
19390
19391         The rule of thumb in EFL is that you init in all the places
19392         you use it. So, if in your application you call ecore_evas_*
19393         functions, you have to call ecore_evas_init() first.
19394
19395         * EWebLauncher/main.c: always call *_init() functions when you use
19396         that namespace.
19397         (quit):
19398         (main):
19399
19400 2011-05-02  Ojan Vafai  <ojan@chromium.org>
19401
19402         Reviewed by Tony Chang.
19403
19404         remove no longer used method
19405         https://bugs.webkit.org/show_bug.cgi?id=59981
19406
19407         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
19408         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
19409
19410 2011-05-02  Ojan Vafai  <ojan@chromium.org>
19411
19412         Reviewed by Tony Chang.
19413
19414         add image toggling behavior from old-run-webkit-tests
19415         https://bugs.webkit.org/show_bug.cgi?id=59978
19416
19417         Output -diffs.html files. This matches old-run-webkit-tests.
19418         In theory, we could use data urls in the results.html file, but
19419         data urls don't currently have access to local files.
19420
19421         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
19422
19423 2011-05-02  Tony Chang  <tony@chromium.org>
19424
19425         Reviewed by Ojan Vafai.
19426
19427         [nrwt] remove some obsolete references to .checksum files
19428         https://bugs.webkit.org/show_bug.cgi?id=59971
19429
19430         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
19431         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
19432         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
19433         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
19434         * Scripts/webkitpy/layout_tests/port/test.py:
19435         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
19436
19437 2011-05-02  Adam Roben  <aroben@apple.com>
19438
19439         Take advantage of implicit nullptr_t -> PassOwnPtr conversion
19440
19441         Fixes <http://webkit.org/b/59964> Implicit conversion from std::nullptr_t to PassOwnPtr
19442         doesn't work, but should
19443
19444         Reviewed by Adam Barth.
19445
19446         * DumpRenderTree/chromium/CppBoundClass.h:
19447         * DumpRenderTree/chromium/WebViewHost.cpp:
19448
19449 2011-05-02  Martin Robinson  <mrobinson@igalia.com>
19450
19451         Reviewed by Xan Lopez.
19452
19453         [GTK] Compiler warning in Tools/MiniBrowser/main.c
19454         https://bugs.webkit.org/show_bug.cgi?id=59986
19455
19456         Remove a compiler warning.
19457
19458         * MiniBrowser/gtk/main.c: Use full array initializers to prevent some compilers
19459         from complaining.
19460
19461 2011-05-02  Adam Barth  <abarth@webkit.org>
19462
19463         Reviewed by Eric Seidel.
19464
19465         Add runtests_unittest.py
19466         https://bugs.webkit.org/show_bug.cgi?id=59955
19467
19468         I forgot to add this file in a previous patch.
19469
19470         * Scripts/webkitpy/tool/steps/runtests_unittest.py: Added.
19471
19472 2011-05-02  Adam Barth  <abarth@webkit.org>
19473
19474         Change the testing EWS to log failures explicitly to avoid an infinite loop.
19475
19476         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
19477
19478 2011-05-02  Adrienne Walker  <enne@google.com>
19479
19480         Unreviewed, fix misspelled contributor name.
19481
19482         * Scripts/webkitpy/common/config/committers.py:
19483
19484 2011-05-02  Adam Barth  <abarth@webkit.org>
19485
19486         Reviewed by Eric Seidel.
19487
19488         cr-linux-ews should run tests
19489         https://bugs.webkit.org/show_bug.cgi?id=59272
19490
19491         This patch turns on testing for the cr-linux-ews.  Yay!
19492
19493         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
19494         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
19495         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
19496         * Scripts/webkitpy/tool/steps/runtests.py:
19497         * Scripts/webkitpy/tool/steps/steps_unittest.py:
19498
19499 2011-05-02  Martin Robinson  <mrobinson@igalia.com>
19500
19501         Incorporate a review comment that I mistakenly omitted from my last commit.
19502
19503         * DumpRenderTree/gtk/EventSender.cpp: Use unsigned here instead of unsigned int.
19504
19505 2011-05-02  Adam Barth  <abarth@webkit.org>
19506
19507         Reviewed by Eric Seidel.
19508
19509         EWS should report the list of unexpected failures
19510         https://bugs.webkit.org/show_bug.cgi?id=59923
19511
19512         We cache the results of the first layout test results run on the task
19513         so the EWS can grab them later.  Another option is to include them in
19514         the exception that's thrown, but this approach seemed slightly more
19515         aesthetic.
19516
19517         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
19518         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
19519
19520 2011-05-02  Adam Barth  <abarth@webkit.org>
19521
19522         Reviewed by Eric Seidel.
19523
19524         Add base case for a test-running EWS
19525         https://bugs.webkit.org/show_bug.cgi?id=59920
19526
19527         Later I will subclass this to run the Chromium Linux EWS.
19528
19529         * Scripts/webkitpy/tool/bot/earlywarningsystemtask.py: Added.
19530         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
19531         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
19532
19533 2011-05-02  Anders Carlsson  <andersca@apple.com>
19534
19535         Build fix.
19536
19537         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
19538         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
19539
19540 2011-05-02  Nico Weber  <thakis@chromium.org>
19541
19542         Reviewed by Mihai Parparita.
19543
19544         Add --no-webkit2 to the output of "build-webkit -h"
19545         https://bugs.webkit.org/show_bug.cgi?id=59942
19546
19547         * Scripts/build-webkit:
19548
19549 2011-05-02  Martin Robinson  <mrobinson@igalia.com>
19550
19551         Reviewed by Xan Lopez.
19552
19553         [GTK] old-run-webkit-tests should disable the Ubuntu overlay scrollbar
19554         https://bugs.webkit.org/show_bug.cgi?id=59943
19555
19556         Disable the Ubuntu overlay scrollbar while testing, by injecting the
19557         LIBOVERLAY_SCROLLBAR=0 environment variable.
19558
19559         * Scripts/old-run-webkit-tests:
19560
19561 2011-05-02  Martin Robinson  <mrobinson@igalia.com>
19562
19563         Reviewed by Eric Seidel.
19564
19565         [GTK] editing/selection/select-out-of-editable.html failing after r83967
19566         https://bugs.webkit.org/show_bug.cgi?id=58667
19567
19568         Do not trigger mouse down events when the EventSender knows that a mouse button is
19569         already down. These type of situations can confuse Xvfb and cause dragging failures.
19570
19571         * DumpRenderTree/gtk/EventSender.cpp:
19572         (mouseDownCallback): Do not trigger another mouse button down event if the mouse button
19573         is already down. This can interfere with drags in Xvfb.
19574
19575 2011-05-02  Zsolt Fehér  <h490267@stud.u-szeged.hu>
19576
19577         Reviewed by Csaba Osztrogonác.
19578
19579         DRT sideeffect revealed by r71268.
19580         https://bugs.webkit.org/show_bug.cgi?id=48958
19581
19582         Reset QWebSettings::ZoomTextOnly attribute before all tests.
19583
19584         Call setTextSizeMultiplier(1.0) to reset TextZoomFactor and
19585         PageZoomFactor too. It should be done before resetsettings()
19586         to guarantee resetting QWebSettings::ZoomTextOnly correctly.
19587
19588         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
19589         (WebCore::WebPage::resetSettings):
19590         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
19591
19592 2011-05-02  Sheriff Bot  <webkit.review.bot@gmail.com>
19593
19594         Unreviewed, rolling out r85469.
19595         http://trac.webkit.org/changeset/85469
19596         https://bugs.webkit.org/show_bug.cgi?id=59928
19597
19598         Broke  webkitpy-tests on Windows 7, GTK and Qt. (Requested by
19599         bbandix on #webkit).
19600
19601         * Scripts/webkitpy/tool/bot/earlywarningsystemtask.py: Removed.
19602         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
19603         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
19604
19605 2011-05-02  Tomasz Morawski  <t.morawski@samsung.com>
19606
19607         Reviewed by Kenneth Rohde Christiansen.
19608
19609         Cleanup EFL initialization/shutdown in the EWebLauncher
19610         https://bugs.webkit.org/show_bug.cgi?id=59922
19611
19612         Removed ecore_evas_init and edje_init form main function due to they
19613         are also called in the ewk_init. Moved ewk_shutdown to quit function.
19614
19615         * EWebLauncher/main.c:
19616         (quit):
19617         (main):
19618
19619 2011-05-02  Adam Barth  <abarth@webkit.org>
19620
19621         Reviewed by Eric Seidel.
19622
19623         Add base case for a test-running EWS
19624         https://bugs.webkit.org/show_bug.cgi?id=59920
19625
19626         Later I will subclass this to run the Chromium Linux EWS.
19627
19628         * Scripts/webkitpy/tool/bot/earlywarningsystemtask.py: Added.
19629         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
19630         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
19631
19632 2011-05-02  Adam Barth  <abarth@webkit.org>
19633
19634         Reviewed by Eric Seidel.
19635
19636         Add the ability to skip python and perl unit tests on EC2
19637         https://bugs.webkit.org/show_bug.cgi?id=59918
19638
19639         * Scripts/webkitpy/common/config/ports.py:
19640             - This is slightly ugly, but it's useful...
19641         * Scripts/webkitpy/common/config/ports_unittest.py:
19642         * Scripts/webkitpy/tool/mocktool.py:
19643         * Scripts/webkitpy/tool/steps/runtests.py:
19644
19645 2011-05-02  Adam Barth  <abarth@webkit.org>
19646
19647         Reviewed by Eric Seidel.
19648
19649         webkit-patch --non-interactive is incompatible with new-run-webkit-tests
19650         https://bugs.webkit.org/show_bug.cgi?id=59919
19651
19652         new-run-webkit-tests doesn't support --wait-for-httpd (because it
19653         always waits for HTTPD, as I understand things).  It turns out we're
19654         not really using this flag for anything, so we should just remove it.
19655
19656         * Scripts/webkitpy/tool/steps/runtests.py:
19657         * Scripts/webkitpy/tool/steps/steps_unittest.py:
19658
19659 2011-05-01  Eric Seidel  <eric@webkit.org>
19660
19661         Reviewed by Adam Barth.
19662
19663         The bots should learn from expected failures without having to retry
19664         https://bugs.webkit.org/show_bug.cgi?id=59279
19665
19666         Rather than re-trying the whole patch, we can just check whether we now
19667         expect the failures we see with the patch applied, saving a lot of time.
19668
19669         This patch also teaches the CommitQueue how to reject patches even
19670         when the tree is red using the new unexpected_failures() method on
19671         ExpectedFailures.
19672
19673         Finally in testing we found a bug in ExpectedFailures where it would
19674         "trust" results which had many many failures, due to using an
19675         equal sign instead of a less-than sign.
19676
19677         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py
19678         * Scripts/webkitpy/tool/bot/expectedfailures.py
19679         * Scripts/webkitpy/tool/bot/expectedfailures_unittest.py
19680         * Scripts/webkitpy/tool/bot/patchanalysistask.py
19681
19682
19683 2011-05-02  Adam Barth  <abarth@webkit.org>
19684
19685         Reviewed by Eric Seidel.
19686
19687         workspace.py fails to import ScriptError
19688         https://bugs.webkit.org/show_bug.cgi?id=59915
19689
19690         If the workspace fails to create a zip, webkitpy crashes instead of
19691         handling the error as expected.
19692
19693         * Scripts/webkitpy/common/system/workspace.py:
19694         * Scripts/webkitpy/common/system/workspace_unittest.py:
19695         * Scripts/webkitpy/tool/mocktool.py:
19696
19697 2011-05-01  Adam Barth  <abarth@webkit.org>
19698
19699         Remove unadvertantly added file.
19700
19701         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
19702
19703 2011-05-01  Adam Barth  <abarth@webkit.org>
19704
19705         Reviewed by Eric Seidel.
19706
19707         expectedfailures.py should report the unexpected failures when asked
19708         https://bugs.webkit.org/show_bug.cgi?id=59916
19709
19710         This information is useful when the EWS runs tests.
19711
19712         * Scripts/webkitpy/tool/bot/expectedfailures.py:
19713         * Scripts/webkitpy/tool/bot/expectedfailures_unittest.py:
19714
19715 2011-05-01  Adam Barth  <abarth@webkit.org>
19716
19717         Reviewed by Eric Seidel.
19718
19719         JSONTestResult needs to handle multiple results
19720         https://bugs.webkit.org/show_bug.cgi?id=59269
19721
19722         This case occurs when a test is flaky because NRWT will run the test
19723         twice and report both results as "actual."
19724
19725         * Scripts/webkitpy/common/net/resultsjsonparser.py:
19726         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
19727
19728 2011-05-01  Adam Barth  <abarth@webkit.org>
19729
19730         Reviewed by Eric Seidel.
19731
19732         Move checkout/api.py to checkout/checkout.py
19733         https://bugs.webkit.org/show_bug.cgi?id=59906
19734
19735         This is more consistent with Python style.
19736
19737         * Scripts/webkitpy/common/checkout/__init__.py:
19738         * Scripts/webkitpy/common/checkout/scm_unittest.py:
19739         * Scripts/webkitpy/common/host.py:
19740         * Scripts/webkitpy/tool/main.py:
19741
19742 2011-05-01  Eric Seidel  <eric@webkit.org>
19743
19744         Reviewed by Adam Barth.
19745
19746         scm.py should be split into many pieces
19747         https://bugs.webkit.org/show_bug.cgi?id=59908
19748
19749         Forgot an AuthenticationError import.
19750
19751         * Scripts/webkitpy/common/checkout/scm/git.py:
19752         * Scripts/webkitpy/common/checkout/scm/svn.py:
19753
19754 2011-05-01  Arno Renevier  <arno@renevier.net>
19755
19756         Reviewed by Eric Seidel.
19757
19758         wrong exception raised in Lighttpd start method
19759         https://bugs.webkit.org/show_bug.cgi?id=59283
19760
19761         raise standard Exception instead of non existent
19762         google.httpd_utils.HttpdNotStarted
19763
19764         * Scripts/webkitpy/layout_tests/port/http_server.py:
19765
19766 2011-05-01  Eric Seidel  <eric@webkit.org>
19767
19768         Reviewed by Adam Barth.
19769
19770         scm.py should be split into many pieces
19771         https://bugs.webkit.org/show_bug.cgi?id=59908
19772
19773         * Scripts/webkitpy/common/checkout/scm/__init__.py:
19774         * Scripts/webkitpy/common/checkout/scm/commitmessage.py: Added.
19775         * Scripts/webkitpy/common/checkout/scm/git.py: Added.
19776         * Scripts/webkitpy/common/checkout/scm/scm.py:
19777         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
19778         * Scripts/webkitpy/common/checkout/scm/svn.py: Added.
19779
19780 2011-05-01  Eric Seidel  <eric@webkit.org>
19781
19782         Reviewed by Adam Barth.
19783
19784         Make most scm.py tests pass in preparation for splitting them up
19785         https://bugs.webkit.org/show_bug.cgi?id=59904
19786
19787         There is one remaining failure:
19788         webkitpy.common.checkout.scm.scm_unittest.GitTestWithMock.test_create_patch
19789         which identifies a difference between the SVN and Git versions of the
19790         SCM api.  I'm not going to bother fixing that until we split out all
19791         the tests.
19792
19793         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
19794
19795 2011-05-01  Patrick Gansterer  <paroga@webkit.org>
19796
19797         [WIN] Unreviewed buildfix after r85440.
19798
19799         * DumpRenderTree/win/UIDelegate.cpp:
19800         (UIDelegate::UIDelegate): Removed accidental added adoptPtr.
19801
19802 2011-05-01  Patrick Gansterer  <paroga@webkit.org>
19803
19804         [WIN] Unreviewed buildfix after r85434.
19805
19806         * DumpRenderTree/win/FrameLoadDelegate.cpp: Added missing header.
19807         * DumpRenderTree/win/UIDelegate.cpp: Ditto.
19808
19809 2011-05-01  Patrick Gansterer  <paroga@webkit.org>
19810
19811         [WIN] Unreviewed buildfix after r85434.
19812
19813         * DumpRenderTree/win/FrameLoadDelegate.cpp:
19814         (FrameLoadDelegate::FrameLoadDelegate):
19815         * DumpRenderTree/win/UIDelegate.cpp:
19816         (DRTUndoManager::DRTUndoManager):
19817         (UIDelegate::UIDelegate):
19818         (UIDelegate::resetUndoManager):
19819
19820 2011-05-01  Eric Seidel  <eric@webkit.org>
19821
19822         Reviewed by Adam Barth.
19823
19824         Move scm.py into its own module in preparation for splitting it up
19825         https://bugs.webkit.org/show_bug.cgi?id=59884
19826
19827         * Scripts/webkitpy/common/checkout/scm/__init__.py: Added.
19828         * Scripts/webkitpy/common/checkout/scm/scm.py: Moved.
19829         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py: Moved.
19830         * Scripts/webkitpy/test/main.py:
19831
19832 2011-05-01  Adam Barth  <abarth@webkit.org>
19833
19834         Reviewed by Eric Seidel.
19835
19836         webkit-patch fails when not run from root
19837         https://bugs.webkit.org/show_bug.cgi?id=59882
19838
19839         We need to set the CWD when running svn so that svn interprets file
19840         paths relative to the root of the working copy.
19841
19842         * Scripts/webkitpy/common/checkout/scm.py:
19843
19844 2011-04-29  Eric Seidel  <eric@webkit.org>
19845
19846         Reviewed by Adam Barth.
19847
19848         land-safely patches should show show their place in the commit-queue
19849         https://bugs.webkit.org/show_bug.cgi?id=59853
19850
19851         This complicated the "should we show the bubbles or the submit-to-ews button"
19852         logic substantially. :(  But now we'll show both the commit-queue position
19853         as well as the submit-to-ews button for land-safely (cq+ only) patches.
19854         Previously we would only show the submit-to-ews button.
19855
19856         * QueueStatusServer/handlers/statusbubble.py:
19857         * QueueStatusServer/templates/statusbubble.html:
19858
19859 2011-04-29  Naoki Takano  <takano.naoki@gmail.com>
19860
19861         Reviewed by Adam Barth.
19862
19863         Add my name to contributers.
19864         https://bugs.webkit.org/show_bug.cgi?id=59854
19865
19866         * Scripts/webkitpy/common/config/committers.py:
19867
19868 2011-04-29  Keith Kyzivat  <keith.kyzivat@nokia.com>
19869
19870         Reviewed by Csaba Osztrogonác.
19871
19872         [Qt] Don't build TestNetscapePlugin if ENABLE_NETSCAPE_PLUGIN_API=0
19873         https://bugs.webkit.org/show_bug.cgi?id=59772
19874
19875         * Tools.pro:
19876
19877 2011-04-29  Stephanie Lewis  <slewis@apple.com>
19878
19879         Rubber-stamped by Adam Roben.
19880
19881         Remove the production configuration guard since installs only happen in production anyway.
19882         Works around a bug in Xcode
19883
19884         * DumpRenderTree/mac/PerlSupport/Makefile:
19885
19886 2011-04-29  Adam Roben  <aroben@apple.com>
19887
19888         Remove a PHP setting that is now provided by httpd.conf
19889
19890         r85155 added this setting, but putting it in httpd.conf makes more sense (since the setting
19891         isn't dependent on paths that could be different on each user's machine).
19892
19893         Reviewed by Sam Weinig.
19894
19895         * Scripts/webkitperl/httpd.pm:
19896         (getDefaultConfigForTestDirectory): Removed log_errors setting.
19897
19898 2011-04-29  Jon Lee  <jonlee@apple.com>
19899
19900         Unreviewed, Adding Jon Lee as a committer
19901
19902         * Scripts/webkitpy/common/config/committers.py:
19903
19904 2011-04-28  Ojan Vafai  <ojan@chromium.org>
19905
19906         Reviewed by Tony Chang.
19907
19908         fix pretty_patch_available and wdiff_available
19909         https://bugs.webkit.org/show_bug.cgi?id=59761
19910
19911         These methods rely on the check_* method being called first.
19912         This would only happen when we actually tried to compute a diff,
19913         but computing the diff happens on a different process.
19914
19915         Also fix setting _wdiff_avialable to True by default.
19916
19917         * Scripts/webkitpy/layout_tests/port/base.py:
19918         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
19919         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
19920
19921 2011-04-29  Ojan Vafai  <ojan@chromium.org>
19922
19923         Reviewed by Eric Seidel.
19924
19925         exclude skipped tests from the summarized results
19926         https://bugs.webkit.org/show_bug.cgi?id=59813
19927
19928         If we find a need fo skipped tests later, we should either create
19929         a skipped.json file or expose this some other way. As it is, the data
19930         is unused and is hugely bloating the size of full_results.json.
19931
19932         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
19933
19934 2011-04-29  Adam Roben  <aroben@apple.com>
19935
19936         Make DumpRenderTreeSupport get installed when the default configuration is built
19937
19938         The default configuration is Production, but in that case the CONFIGURATION environment
19939         variable is set to the empty string.
19940
19941         Reviewed by Sam Weinig.
19942
19943         * DumpRenderTree/mac/PerlSupport/Makefile: Set CONFIGURATION to Production if it isn't
19944         already set so that later logic that depends on CONFIGURATION having that value will work.
19945
19946 2011-04-29  Dean Jackson  <dino@apple.com>
19947
19948         Reviewed by Simon Fraser.
19949
19950         Add ENABLE macro for WebKitAnimation
19951         https://bugs.webkit.org/show_bug.cgi?id=59729
19952
19953         Add new feature to toggle WebKit Animation API.
19954
19955         * Scripts/build-webkit:
19956
19957 2011-04-29  Sam Weinig  <sam@webkit.org>
19958
19959         Reviewed by Adam Roben.
19960
19961         Make production the default build type for WebKitTestRunner, matching
19962         all other projects.
19963
19964         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
19965
19966 2011-04-29  Kevin Ollivier  <kevino@theolliviers.com>
19967
19968         [wx] Unreviewed build fix, add new directory to the build.
19969
19970         * waf/build/settings.py:
19971
19972 2011-04-29  Carlos Garcia Campos  <cgarcia@igalia.com>
19973
19974         Reviewed by Martin Robinson.
19975
19976         [GTK] Implement UI client in minibrowser
19977         https://bugs.webkit.org/show_bug.cgi?id=58416
19978
19979         It also allows to open more than one minibrowser window from the
19980         command line.
19981
19982         * MiniBrowser/gtk/BrowserWindow.c:
19983         (browserWindowFinalize):
19984         (browser_window_init):
19985         (browserWindowConstructed):
19986         (createNewPage):
19987         (showPage):
19988         (closePage):
19989         (runJavaScriptAlert):
19990         (browserWindowUIClientInit):
19991         * MiniBrowser/gtk/main.c:
19992         (loadURI):
19993         (main):
19994
19995 2011-04-29  Carlos Garcia Campos  <cgarcia@igalia.com>
19996
19997         Reviewed by Martin Robinson.
19998
19999         [GTK] Implement page loader client in MiniBrowser
20000         https://bugs.webkit.org/show_bug.cgi?id=58024
20001
20002         A new class BrowserWindow has been added to implement the page
20003         loader client. It makes supporting multiple windows easier. Window
20004         title and url entry are now updated during the loading process.
20005
20006         * MiniBrowser/gtk/BrowserWindow.c: Added.
20007         (activateUriEntryCallback):
20008         (goBackCallback):
20009         (goForwardCallback):
20010         (browserWindowFinalize):
20011         (browserWindowGetProperty):
20012         (browserWindowSetProperty):
20013         (browser_window_init):
20014         (browserWindowConstructed):
20015         (browser_window_class_init):
20016         (browserWindowUpdateTitle):
20017         (browserWindowSetTitle):
20018         (browserWindowSetLoadProgress):
20019         (browserWindowUpdateURL):
20020         (didStartProvisionalLoadForFrame):
20021         (didReceiveServerRedirectForProvisionalLoadForFrame):
20022         (didFailProvisionalLoadWithErrorForFrame):
20023         (didCommitLoadForFrame):
20024         (didFinishDocumentLoadForFrame):
20025         (didFinishLoadForFrame):
20026         (didFailLoadWithErrorForFrame):
20027         (didReceiveTitleForFrame):
20028         (didFirstLayoutForFrame):
20029         (didFirstVisuallyNonEmptyLayoutForFrame):
20030         (didRemoveFrameFromHierarchy):
20031         (didStartProgress):
20032         (didChangeProgress):
20033         (didFinishProgress):
20034         (didBecomeUnresponsive):
20035         (didBecomeResponsive):
20036         (browserWindowLoaderClientInit):
20037         (browser_window_new):
20038         (browser_window_get_view):
20039         * MiniBrowser/gtk/BrowserWindow.h: Added.
20040         * MiniBrowser/gtk/GNUmakefile.am:
20041         * MiniBrowser/gtk/main.c:
20042         (createWindow):
20043
20044 2011-04-28  Stephanie Lewis  <slewis@apple.com>
20045
20046         Reviewed by Mark Rowe.
20047
20048         http://bugs.webkit.org/show_bug.cgi?id=59743
20049         <rdar://problem/9040413> Add an install target for production builds.
20050
20051         * DumpRenderTree/mac/PerlSupport/Makefile:
20052
20053 2011-04-28  Adam Roben  <aroben@apple.com>
20054
20055         Stop and restart ATSServer before each test run on Leopard
20056
20057         This is to work around an ATSServer leak that eventually causes tests
20058         to fail.
20059
20060         Fixes <http://webkit.org/b/51807> <rdar://problem/9273894>
20061         fast/blockflow/broken-ideograph-[small-caps|font].html sometimes fails on Leopard buildbots
20062         (due to ATSServer leak)
20063
20064         Reviewed by Dan Bernstein.
20065
20066         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
20067         (StopATSServer):
20068         (StartATSServer):
20069         Added these new commands to stop and start ATSServer.
20070
20071         (TestFactory.__init__): Stop and start ATSServer on Leopard before running tests.
20072
20073 2011-04-28  Dirk Pranke  <dpranke@chromium.org>
20074
20075         Reviewed by Adam Roben.
20076
20077         add Leopard NRWT bot to the scheduler
20078         https://bugs.webkit.org/show_bug.cgi?id=59766
20079
20080         * BuildSlaveSupport/build.webkit.org-config/config.json:
20081
20082 2011-04-28  Sam Weinig  <sam@webkit.org>
20083
20084         Fix WebKitTestRunner build.
20085
20086         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
20087         The Derived Sources target needs a product name.
20088
20089 2011-04-28  Sam Weinig  <sam@webkit.org>
20090
20091         Reviewed by Mark Rowe.
20092
20093         Add install support for WebKitTestRunner.
20094         <rdar://problem/9349055>
20095
20096         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
20097         Add missing CompilerVersion.xcconfig.
20098
20099         * WebKitTestRunner/Configurations/InjectedBundle.xcconfig:
20100         * WebKitTestRunner/Configurations/WebKitTestRunner.xcconfig:
20101         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
20102         Add install path and tell it not to skip installation.
20103
20104         Add missing CompilerVersion.xcconfig.
20105
20106 2011-04-28  David Levin  <levin@chromium.org>
20107
20108         Reviewed by Adam Barth.
20109
20110         Remove IMAGE_RESIZER related code.
20111         https://bugs.webkit.org/show_bug.cgi?id=59735
20112
20113         * Scripts/build-webkit:
20114
20115 2011-04-28  Sam Weinig  <sam@webkit.org>
20116
20117         Reviewed by Adam Roben.
20118
20119         Make reading skipped files from additional platform
20120         directories work in WebKit2 mode.
20121
20122         * Scripts/old-run-webkit-tests:
20123         (buildPlatformTestHierarchy): Simplify logic to always use the first
20124         and last platforms in the build hierarchy as the test hierarchy, and to
20125         insert the WebKit2 platform in the middle if it exists. Yum.
20126
20127 2011-04-28  Adam Barth  <abarth@webkit.org>
20128
20129         Reviewed by Eric Seidel.
20130
20131         Remove WML
20132         https://bugs.webkit.org/show_bug.cgi?id=59678
20133
20134         Remove build system integration and test harness integration for WML.
20135
20136         * Scripts/build-webkit:
20137         * Scripts/old-run-webkit-tests:
20138         * Scripts/webkitperl/features.pm:
20139         * Scripts/webkitpy/layout_tests/port/webkit.py:
20140         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
20141
20142 2011-04-28  Ojan Vafai  <ojan@chromium.org>
20143
20144         Reviewed by Mihai Parparita.
20145
20146         use a hierarchical data structure to store results json
20147         https://bugs.webkit.org/show_bug.cgi?id=59736
20148
20149         Instead of a map from test name to results dict, we now store a hierarchical
20150         data structure.
20151
20152         Old: { foo/bar/baz.html: results_dict, foo/bar/baz2.html: results_dict }
20153         New:
20154         { foo: {
20155             bar: {
20156               baz.html: results_dict,
20157               baz2.html: results_dict } } }
20158
20159         This cut ~52% of the size of times_ms.json. I expect it will cut
20160         30-40% of full_results.json and unexpected_results.json.
20161
20162         * Scripts/webkitpy/common/net/resultsjsonparser.py:
20163         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
20164         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
20165         Using os.sep was just wrong since relative_test_filename canonicalizes separators.
20166         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
20167         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
20168         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
20169         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
20170         Instead of plumbing through the change to all of the rebaselineserver
20171         python and JS, just convert to the old format after reading in the json.
20172
20173 2011-04-28  Eric Seidel  <eric@webkit.org>
20174
20175         Reviewed by Adam Barth.
20176
20177         EWS Bubbles should show commit-queue position when patch is in the commit-queue
20178         https://bugs.webkit.org/show_bug.cgi?id=59744
20179
20180         Add target="_top" to the bubble links.  This was
20181         forgotten when I moved from using an onclick to a normal link.
20182
20183         * QueueStatusServer/templates/statusbubble.html:
20184
20185 2011-04-28  Eric Seidel  <eric@webkit.org>
20186
20187         Reviewed by Adam Barth.
20188
20189         EWS bubbles should show "submit to ews" button if they've not been processed
20190         https://bugs.webkit.org/show_bug.cgi?id=59751
20191
20192         * QueueStatusServer/handlers/statusbubble.py:
20193         * QueueStatusServer/handlers/submittoews.py:
20194         * QueueStatusServer/templates/statusbubble.html:
20195         * QueueStatusServer/templates/submittoews.html:
20196
20197 2011-04-28  Eric Seidel  <eric@webkit.org>
20198
20199         Reviewed by Adam Barth.
20200
20201         EWS Bubbles should show commit-queue position when patch is in the commit-queue
20202         https://bugs.webkit.org/show_bug.cgi?id=59744
20203
20204         This was actually just a 2 line change to statusbubble.py.
20205         I also cleaned up statusbubble.html a little while I was there
20206         (the links are now actual <a> tags and have hover).
20207
20208         * QueueStatusServer/handlers/statusbubble.py:
20209         * QueueStatusServer/templates/statusbubble.html:
20210
20211 2011-04-28  Dirk Pranke  <dpranke@chromium.org>
20212
20213         Reviewed by Eric Seidel.
20214
20215         add a mac leopard release NRWT bot
20216         https://bugs.webkit.org/show_bug.cgi?id=59733
20217
20218         * BuildSlaveSupport/build.webkit.org-config/config.json:
20219
20220 2011-04-28  Ojan Vafai  <ojan@chromium.org>
20221
20222         Avoid crashing when running new-run-webkit-tests outside of
20223         version control.
20224
20225         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
20226
20227 2011-04-28  Xan Lopez  <xlopez@igalia.com>
20228
20229         Unreviewed build fix.
20230
20231         * GNUmakefile.am:
20232
20233 2011-04-28  Ojan Vafai  <ojan@chromium.org>
20234
20235         Reviewed by Eric Seidel.
20236
20237         fix python printing test
20238         https://bugs.webkit.org/show_bug.cgi?id=59719
20239
20240         Now that we run scm commands in summarize_results, --verbose prints
20241         debug logging to stderr.
20242
20243         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
20244
20245 2011-04-28  Ojan Vafai  <ojan@chromium.org>
20246
20247         Reviewed by Tony Chang.
20248
20249         link to tests at a specific revision in the results.html file
20250         https://bugs.webkit.org/show_bug.cgi?id=59711
20251
20252         * Scripts/webkitpy/common/checkout/scm.py:
20253         Add a method to get the head svn revision.
20254         * Scripts/webkitpy/common/checkout/scm_unittest.py:
20255         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
20256         Include the svn revision in the results json.
20257
20258 2011-04-28  Carlos Garcia Campos  <cgarcia@igalia.com>
20259
20260         Reviewed by Sam Weinig.
20261
20262         [GTK] Remove WKURLCreateWithURL()
20263         https://bugs.webkit.org/show_bug.cgi?id=59691
20264
20265         Use WKURLCreateWithUTF8CString() instead of WKURLCreateWithURL().
20266
20267         * MiniBrowser/gtk/main.c:
20268         (activateUriEntryCallback):
20269         (main):
20270
20271 2011-04-28  Carlos Garcia Campos  <cgarcia@igalia.com>
20272
20273         Reviewed by Martin Robinson.
20274
20275         [GTK] WKView should be a GtkWidget
20276         https://bugs.webkit.org/show_bug.cgi?id=59603
20277
20278         Don't use WKViewGetWindow(), use WKView as a GtkWidget instead.
20279
20280         * MiniBrowser/gtk/main.c:
20281         (createWindow):
20282         (main):
20283
20284 2011-04-28  Gabor Rapcsanyi  <rgabor@webkit.org>
20285
20286         Reviewed by Csaba Osztrogonác.
20287
20288         [Qt] Add ARM specific Skipped list
20289         https://bugs.webkit.org/show_bug.cgi?id=59680
20290
20291         * Scripts/old-run-webkit-tests:
20292         * Scripts/webkitdirs.pm:
20293
20294 2011-04-27  Eric Seidel  <eric@webkit.org>
20295
20296         Reviewed by Adam Barth.
20297
20298         error_log.txt should include PHP errors
20299         https://bugs.webkit.org/show_bug.cgi?id=59643
20300
20301         Turns out this was really easy after I finally learned
20302         enough about PHP to do it...
20303
20304         I'm adding this in hopes of better tracking down
20305         our random http failures.
20306
20307         * Scripts/webkitperl/httpd.pm:
20308
20309 2011-04-27  David Levin  <levin@chromium.org>
20310
20311         Reviewed by Adam Barth.
20312
20313         check-webkit-style should flags spacing errors for braces on the same line.
20314         https://bugs.webkit.org/show_bug.cgi?id=59665
20315
20316         * Scripts/webkitpy/style/checkers/cpp.py: Added the check.
20317         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added some tests for the check
20318           and fixed some test cases to not trigger this error since it isn't what they are verifying.
20319
20320 2011-04-27  David Levin  <levin@chromium.org>
20321
20322         Reviewed by Adam Barth.
20323
20324         check-webkit-style error for spaces after periods in a comment has a bad error
20325         message and warns about end of line spaces.
20326         https://bugs.webkit.org/show_bug.cgi?id=59663
20327
20328         * Scripts/webkitpy/style/checkers/cpp.py: Fix both issues.
20329         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added tests for spaces at the end of line
20330         and changed the error messages.
20331
20332 2011-04-27  Ojan Vafai  <ojan@chromium.org>
20333
20334         remove assertion now that platform/mac-tiger no longer exists
20335         https://bugs.webkit.org/show_bug.cgi?id=59660
20336
20337         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
20338
20339 2011-04-27  Ojan Vafai  <ojan@chromium.org>
20340
20341         Reviewed by Eric Seidel.
20342
20343         remove test that didn't make sense
20344         https://bugs.webkit.org/show_bug.cgi?id=59658
20345
20346         It was artifically saying the test wasn't expected to pass,
20347         but the expectations string was empty. The case of unexpected
20348         passes with a real expectations string is tested later in the
20349         same test.
20350
20351         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
20352
20353 2011-04-27  Sam Weinig  <sam@webkit.org>
20354
20355         Reviewed by Adam Roben.
20356
20357         WebKit2 needs to notify the client when a plugin fails to initialize
20358         https://bugs.webkit.org/show_bug.cgi?id=59657
20359
20360         * MiniBrowser/mac/BrowserWindowController.m:
20361         (-[BrowserWindowController awakeFromNib]):
20362         * WebKitTestRunner/TestController.cpp:
20363         (WTR::TestController::initialize):
20364         Add stubs.
20365
20366 2011-04-26  Sam Weinig  <sam@webkit.org>
20367
20368         Reviewed by David Levin.
20369
20370         Add a few more missing options to new-run-webkit-tests
20371         https://bugs.webkit.org/show_bug.cgi?id=59567
20372
20373         Add support for:
20374             --threaded
20375             --gc-between-tests
20376             --complex-text
20377
20378         * Scripts/webkitpy/layout_tests/port/webkit.py:
20379         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
20380         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
20381
20382 2011-04-27  Mark Rowe  <mrowe@apple.com>
20383
20384         Build fix.
20385
20386         Fix an OwnPtr-related issue that was missed before the strictness was turned on.
20387
20388         * WebKitTestRunner/TestController.cpp:
20389         (WTR::TestController::runTest):
20390
20391 2011-04-27  Ojan Vafai  <ojan@chromium.org>
20392
20393         Forgot to upload the new times_ms.json file after splitting
20394         it out from full_results.json.
20395
20396         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
20397
20398 2011-04-27  Ojan Vafai  <ojan@chromium.org>
20399
20400         Fix snafu in generating the JSON for the chromium tests.
20401
20402         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
20403
20404 2011-04-27  Ojan Vafai  <ojan@chromium.org>
20405
20406         Fix bad merge. This broke the chromium tests generating JSON files.
20407         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
20408
20409 2011-04-27  Mark Rowe  <mrowe@apple.com>
20410
20411         Rubber-stamped by Geoff Garen.
20412
20413         Don't look for results in a location that no longer exists.
20414
20415         * Scripts/old-run-webkit-tests:
20416
20417 2011-04-27  Adrienne Walker  <enne@google.com>
20418
20419         Unreviewed, add Chrome GPU folks to contributors that I'm tired of manually CCing
20420
20421         * Scripts/webkitpy/common/config/committers.py:
20422
20423 2011-04-22  Ojan Vafai  <ojan@chromium.org>
20424
20425         Reviewed by Tony Chang.
20426
20427         only include failures in full_results.json
20428         https://bugs.webkit.org/show_bug.cgi?id=59256
20429
20430         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
20431
20432 2011-04-22  Ojan Vafai  <ojan@chromium.org>
20433
20434         Reviewed by Adam Roben.
20435
20436         move times into their own json file
20437         https://bugs.webkit.org/show_bug.cgi?id=59239
20438
20439         The test runtimes are the only case where we need an entry for
20440         every test that was run, including passing tests. Move them into
20441         their own json file so we can prune the other json files to only
20442         include failing tests.
20443
20444         Use a trie-like data structure to hold the tests to avoid including
20445         full test paths. This cuts 52% of the filesize. The end result is ~700k
20446         pre-gzip.
20447
20448         * Scripts/webkitpy/common/net/resultsjsonparser.py:
20449         Exclude times from the parsed TestResults.
20450         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
20451         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
20452         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
20453         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
20454
20455 2011-04-27  Anders Carlsson  <andersca@apple.com>
20456
20457         More casts and more clang++ build fixes.
20458
20459         * DumpRenderTree/mac/DumpRenderTree.mm:
20460         (createWebViewAndOffscreenWindow):
20461         * TestWebKitAPI/mac/PlatformWebViewMac.mm:
20462         (TestWebKitAPI::PlatformWebView::PlatformWebView):
20463         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
20464         (WTR::PlatformWebView::PlatformWebView):
20465
20466 2011-04-27  Tony Chang  <tony@chromium.org>
20467
20468         Reviewed by Ojan Vafai.
20469
20470         lazily autoinstall thirdparty python libraries
20471         https://bugs.webkit.org/show_bug.cgi?id=55723
20472
20473         This helps by only having pywebsocket installed when the bots run.  It
20474         doesn't help much for devs since webkit-patch pulls in almost everything.
20475
20476         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
20477         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
20478         * Scripts/webkitpy/common/net/networktransaction.py:
20479         * Scripts/webkitpy/common/net/networktransaction_unittest.py:
20480         * Scripts/webkitpy/common/net/statusserver.py:
20481         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
20482         * Scripts/webkitpy/thirdparty/__init__.py: Add an import hook so we can install based on
20483             the import command rather than just installing everything.
20484         * Scripts/webkitpy/thirdparty/__init___unittest.py: Added.
20485
20486 2011-04-27  Girish Ramakrishnan  <girish@forwardbias.in>
20487
20488         Add my IRC nick.
20489
20490         * Scripts/webkitpy/common/config/committers.py:
20491
20492 2011-04-27  Eric Seidel  <eric@webkit.org>
20493
20494         Reviewed by Mihai Parparita.
20495
20496         sherrifbot create-bug shouldn't assign bugs to webkit.review.bot
20497         https://bugs.webkit.org/show_bug.cgi?id=59545
20498
20499         To do this, I needed a way to look up contributors by irc-name
20500         (since anyone in #webkit who might use this command may not be a committer).
20501         To lookup contributors, I had to make Contributor a real object.
20502         Which led me to redesign parts of committers.py...
20503         and finally fix one spot in changelog.py where we wanted to be
20504         looking up contributors and not committers.
20505
20506         Overall a pretty simple fix, once you wade through the yak-hair.
20507
20508         This may not prevent *all* possible ways that bugs would get assigned
20509         to webkit.review.bot.  If we don't recognize the requester we will
20510         go through the previous code path (which shouldn't change the assignee
20511         on the bug from the default as far as I can tell).
20512
20513         * Scripts/webkitpy/common/checkout/changelog.py:
20514         * Scripts/webkitpy/common/config/committers.py:
20515         * Scripts/webkitpy/common/config/committers_unittest.py:
20516         * Scripts/webkitpy/tool/bot/irc_command.py:
20517         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
20518
20519 2011-04-27  Yi Shen  <yi.4.shen@nokia.com>
20520
20521         Reviewed by Kenneth Rohde Christiansen.
20522
20523         [Qt] Upstream Symbian platform plugin
20524         https://bugs.webkit.org/show_bug.cgi?id=58435
20525
20526         Exempting directory WebKit/qt/symbian/platformplugin from style guide.
20527
20528         * Scripts/webkitpy/style/checker.py:
20529         * Scripts/webkitpy/style/checker_unittest.py:
20530
20531 2011-04-27  Jeff Miller  <jeffm@apple.com>
20532
20533         Correct my IRC nickname.
20534
20535         * Scripts/webkitpy/common/config/committers.py:
20536
20537 2011-04-27  Lucas De Marchi  <lucas.demarchi@profusion.mobi>
20538
20539         Reviewed by Darin Adler.
20540
20541         Change the email used on bugzilla
20542         https://bugs.webkit.org/show_bug.cgi?id=59582
20543
20544         * Scripts/webkitpy/common/config/committers.py:
20545
20546 2011-04-27  James Kozianski  <koz@chromium.org>
20547
20548         Reviewed by Ojan Vafai.
20549
20550         Extract map from port to builder name.
20551         https://bugs.webkit.org/show_bug.cgi?id=58301
20552
20553         * Scripts/webkitpy/layout_tests/port/__init__.py:
20554         * Scripts/webkitpy/layout_tests/port/builders.py: Added.
20555         * Scripts/webkitpy/layout_tests/port/builders_unittest.py: Added.
20556         * Scripts/webkitpy/layout_tests/port/factory.py:
20557
20558 2011-04-27  Ryuan Choi  <ryuan.choi@samsung.com>
20559
20560         Reviewed by Martin Robinson.
20561
20562         [GTK] Add proxy support to GtkLauncher
20563         https://bugs.webkit.org/show_bug.cgi?id=58852
20564
20565         Add SOUP_TYPE_PROXY_RESOLVER_DEFAULT feature or set SOUP_SESSION_PROXY_URI for developers behind proxy.
20566
20567         * GNUmakefile.am:
20568         * GtkLauncher/main.c:
20569         (main):
20570
20571 2011-04-27  Annie Sullivan  <sullivan@chromium.org>
20572
20573         Reviewed by David Levin.
20574
20575         Adding myself as a Chromium contributor to committers.py.
20576
20577         https://bugs.webkit.org/show_bug.cgi?id=59576
20578
20579         * Scripts/webkitpy/common/config/committers.py:
20580
20581 2011-04-26  David Grogan  <dgrogan@chromium.org>
20582
20583         Reviewed by David Levin.
20584
20585         Adding myself to Chromium contributors list in committers.py
20586         https://bugs.webkit.org/show_bug.cgi?id=59565
20587
20588         * Scripts/webkitpy/common/config/committers.py:
20589
20590 2011-04-26  Chang Shu  <cshu@webkit.org>
20591
20592         Reviewed by Laszlo Gombos.
20593
20594         WebKitTestRunner needs layoutTestController.addOriginAccessWhitelistEntry
20595         https://bugs.webkit.org/show_bug.cgi?id=42544
20596
20597         Also implemented removeOriginAccessWhitelistEntry and resetOriginAccessWhitelists.
20598
20599         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
20600         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
20601         (WTR::InjectedBundle::beginTesting):
20602         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
20603         (WTR::LayoutTestController::addOriginAccessWhitelistEntry):
20604         (WTR::LayoutTestController::removeOriginAccessWhitelistEntry):
20605         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
20606
20607 2011-04-26  Chris Fleizach  <cfleizach@apple.com>
20608
20609         Reviewed by Adele Peterson.
20610
20611         AX: Add more text marker routines to DRT
20612         https://bugs.webkit.org/show_bug.cgi?id=58949
20613
20614         Add a few more methods to help test the text marker system, including nextMarker, previousMarker and
20615         stringForMarkerRange.
20616
20617         * DumpRenderTree/AccessibilityUIElement.cpp:
20618         (nextTextMarkerCallback):
20619         (previousTextMarkerCallback):
20620         (stringForTextMarkerRangeCallback):
20621         (AccessibilityUIElement::previousTextMarker):
20622         (AccessibilityUIElement::nextTextMarker):
20623         (AccessibilityUIElement::stringForTextMarkerRange):
20624         (AccessibilityUIElement::getJSClass):
20625         * DumpRenderTree/AccessibilityUIElement.h:
20626         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
20627         (AccessibilityUIElement::previousTextMarker):
20628         (AccessibilityUIElement::nextTextMarker):
20629         (AccessibilityUIElement::stringForTextMarkerRange):
20630
20631 2011-04-26  Dan Bernstein  <mitz@apple.com>
20632
20633         Build fix.
20634
20635         * DumpRenderTree/mac/Configurations/CompilerVersion.xcconfig:
20636         * MiniBrowser/Configurations/CompilerVersion.xcconfig:
20637         * WebKitTestRunner/Configurations/CompilerVersion.xcconfig:
20638
20639 2011-04-26  Kevin Ollivier  <kevino@theolliviers.com>
20640
20641         Rubberstamped by Eric Seidel.
20642
20643         Enable waf to be used to build other ports
20644         https://bugs.webkit.org/show_bug.cgi?id=58213
20645
20646         * Scripts/build-webkit:
20647         * Scripts/webkitdirs.pm:
20648         * waf: Added.
20649         * waf/build: Added.
20650         * waf/build/build_utils.py: Added.
20651         * waf/build/settings.py: Added.
20652         * waf/build/waf_extensions.py: Added.
20653         * waf/build/wxpresets.py: Added.
20654
20655 2011-04-26  Sam Weinig  <sam@webkit.org>
20656
20657         Reviewed by David Hyatt.
20658
20659         Remove Datagrid from the tree
20660         https://bugs.webkit.org/show_bug.cgi?id=59543
20661
20662         * Scripts/build-webkit:
20663         Remove feature.
20664
20665         * iExploder/iexploder-1.3.2/htdocs/htmltags.in:
20666         * iExploder/iexploder-1.7.2/src/html-tags/webkit:
20667         Remove from fuzzer.
20668
20669 2011-04-26  Brent Fulgham  <bfulgham@webkit.org>
20670
20671         Reviewed by Adam Roben.
20672
20673         Update WinLauncher with command-line arguments to allow the
20674         transparency to be tested.  Use --transparent to create a
20675         transparent view.  Use --desktop to have the view fill the
20676         desktop area of the screen.
20677         https://bugs.webkit.org/show_bug.cgi?id=58300
20678
20679         * WinLauncher/WinLauncher.cpp:
20680         (usesLayeredWebView): New method to indicate that the
20681         program is running with layered windows (alpha-blended
20682         transparent windows).
20683         (shouldUseFullDesktop): New method to indicate that the
20684         program is running across the full desktop (less any
20685         task bars.)
20686         (resizeSubViews): Layered window mode doesn't have any
20687         winapi child windows, so this routine is exited early
20688         for that case.
20689         (subclassForLayeredWindow): Subclass the window so we can
20690         override its defaut message loop.
20691         (computeFullDesktopFrame): Convenience function to get the
20692         desktop region and update the frame dimensions appropriately.
20693         (_tWinMain): Updated to accept the new '--transparent' and
20694         '--desktop' arguments.
20695         (InitInstance): Modified to exit early for the transparent
20696         background case.  WebKit will construct our primary window (with
20697         appropriate settings) so we don't need to create one ourselves.
20698         (WndProc):
20699         (TransparentWndProc): New message loop for transparent windows,
20700         since they need special handling to support dragging.
20701
20702 2011-04-26  Alexandre MAzari  <amazari@igalia.com>
20703
20704         Reviewed by Xan Lopez.
20705
20706         [Gtk] Implement LayoutTestController::setPluginsEnabled
20707         https://bugs.webkit.org/show_bug.cgi?id=59184
20708
20709         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
20710         (LayoutTestController::setPluginsEnabled):
20711
20712 2011-04-26  Chang Shu  <cshu@webkit.org>
20713
20714         Reviewed by Laszlo Gombos.
20715
20716         WebKitTestRunner needs layoutTestController.setAllowFileAccessFromFileURLs
20717         https://bugs.webkit.org/show_bug.cgi?id=57572
20718
20719         Remove the unnecessary code that updates preference.
20720
20721         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
20722         (WTR::InjectedBundle::beginTesting):
20723         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
20724         (WTR::LayoutTestController::setAllowUniversalAccessFromFileURLs):
20725
20726 2011-04-26  Mario Sanchez Prada  <msanchez@igalia.com>
20727
20728         Reviewed by Eric Seidel.
20729
20730         [GTK] accessibility/aria-combobox-expected.txt
20731         https://bugs.webkit.org/show_bug.cgi?id=59399
20732
20733         Implement isExpanded(), as it's used in the layout test.
20734
20735         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
20736         (AccessibilityUIElement::isExpanded): Implemented.
20737
20738 2011-04-26  Adrienne Walker  <enne@google.com>
20739
20740         Reviewed by Kent Tamura.
20741
20742         [chromium] Fix memory leak in DRT due to OwnPtr<T*>
20743         https://bugs.webkit.org/show_bug.cgi?id=59489
20744
20745         * DumpRenderTree/chromium/WebViewHost.h:
20746
20747 2011-04-26  Mario Sanchez Prada  <msanchez@igalia.com>
20748
20749         Reviewed by Eric Seidel.
20750
20751         [GTK] Implement isChecked in DRT
20752         https://bugs.webkit.org/show_bug.cgi?id=59385
20753
20754         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
20755         (AccessibilityUIElement::isChecked): Implemented.
20756
20757 2011-04-26  Alejandro G. Castro  <alex@igalia.com>
20758
20759         Unreviewed, rolling out r84949.
20760         http://trac.webkit.org/changeset/84949
20761         https://bugs.webkit.org/show_bug.cgi?id=58852
20762
20763         Broke GTK+ compilation
20764
20765         * GtkLauncher/main.c:
20766         (main):
20767
20768 2011-04-26  Eric Seidel  <eric@webkit.org>
20769
20770         Reviewed by Mihai Parparita.
20771
20772         sheriff-bot should have a create-bug command
20773         https://bugs.webkit.org/show_bug.cgi?id=59491
20774
20775         This patch has some stderr unittest spew, but I will
20776         fix it after my live demo.
20777
20778         * Scripts/webkitpy/tool/bot/irc_command.py:
20779         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
20780         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
20781
20782 2011-04-26  Ryuan Choi  <ryuan.choi@samsung.com>
20783
20784         Reviewed by Martin Robinson.
20785
20786         [GTK] Add proxy support to GtkLauncher
20787         https://bugs.webkit.org/show_bug.cgi?id=58852
20788
20789         Add SOUP_TYPE_PROXY_RESOLVER_DEFAULT feature or set SOUP_SESSION_PROXY_URI for developers behind proxy.
20790
20791         * GtkLauncher/main.c:
20792         (main):
20793
20794 2011-04-26  Dan Bernstein  <mitz@apple.com>
20795
20796         Reviewed by Mark Rowe.
20797
20798         Choose the compiler based on the Xcode version for Snow Leopard debug builds.
20799
20800         * DumpRenderTree/mac/Configurations/Base.xcconfig:
20801         * DumpRenderTree/mac/Configurations/CompilerVersion.xcconfig: Added.
20802         * MiniBrowser/Configurations/Base.xcconfig:
20803         * MiniBrowser/Configurations/CompilerVersion.xcconfig: Added.
20804         * TestWebKitAPI/Configurations/Base.xcconfig:
20805         * TestWebKitAPI/Configurations/CompilerVersion.xcconfig: Added.
20806         * WebKitTestRunner/Configurations/Base.xcconfig:
20807         * WebKitTestRunner/Configurations/CompilerVersion.xcconfig: Added.
20808
20809 2011-04-26  Finnur Thorarinsson  <finnur.webkit@gmail.com>
20810
20811         Reviewed by David Levin.
20812
20813         Adding myself as a Chromium contributors to committers.py.
20814
20815         https://bugs.webkit.org/show_bug.cgi?id=59444
20816
20817         * Scripts/webkitpy/common/config/committers.py:
20818
20819 2011-04-26  Adam Barth  <abarth@webkit.org>
20820
20821         Reviewed by Eric Seidel.
20822
20823         DumpRenderTree/chromium should play nice with strict OwnPtrs
20824         https://bugs.webkit.org/show_bug.cgi?id=59458
20825
20826         All these changes are pretty straight forward.
20827
20828         * DumpRenderTree/chromium/CppBoundClass.cpp:
20829         (GetterPropertyCallback::GetterPropertyCallback):
20830         (CppBoundClass::bindGetterCallback):
20831         * DumpRenderTree/chromium/CppBoundClass.h:
20832         (CppBoundClass::bindProperty):
20833         (CppBoundClass::bindFallbackCallback):
20834         (CppBoundClass::bindFallbackMethod):
20835         * DumpRenderTree/chromium/DRTDevToolsClient.cpp:
20836         (DRTDevToolsClient::DRTDevToolsClient):
20837         * DumpRenderTree/chromium/TestEventPrinter.cpp:
20838         (TestEventPrinter::createDRTPrinter):
20839         (TestEventPrinter::createTestShellPrinter):
20840         * DumpRenderTree/chromium/TestEventPrinter.h:
20841         * DumpRenderTree/chromium/TestShell.cpp:
20842         (TestShell::TestShell):
20843         (TestShell::createMainWindow):
20844         (TestShell::createDRTDevToolsClient):
20845         * DumpRenderTree/chromium/WebViewHost.cpp:
20846         (WebViewHost::geolocationClientMock):
20847         (WebViewHost::speechInputController):
20848         (WebViewHost::deviceOrientationClientMock):
20849         (WebViewHost::reset):
20850         (WebViewHost::navigate):
20851         (WebViewHost::setPendingExtraData):
20852         (WebViewHost::canvas):
20853         * DumpRenderTree/chromium/WebViewHost.h:
20854
20855 2011-04-26  Adam Roben  <aroben@apple.com>
20856
20857         Make DRT/win log loading errors the same way DRT/mac does
20858
20859         Fixes <http://webkit.org/b/59303> [Windows WebKit1 Tests]
20860         http/tests/misc/will-send-request-returns-null-on-redirect.html failing since r84742
20861
20862         Reviewed by Alexey Proskuryakov.
20863
20864         * DumpRenderTree/win/ResourceLoadDelegate.cpp:
20865         (ResourceLoadDelegate::descriptionSuitableForTestResult): Removed the workaround for
20866         <rdar://problem/5064234>, which has long since been fixed, and then made this function match
20867         -[NSError(DRTExtras) _drt_descriptionSuitableForTestResult] more closely.
20868
20869 2011-04-26  Dan Bernstein  <mitz@apple.com>
20870
20871         REGRESSION (84876): build-webkit produces an error message
20872         https://bugs.webkit.org/show_bug.cgi?id=59460
20873
20874         * Scripts/webkitdirs.pm:
20875
20876 2011-04-26  Mihai Parparita  <mihaip@chromium.org>
20877
20878         Reviewed by Eric Seidel.
20879
20880         Fix timestamp comparison in check-for-weak-vtables-and-externals
20881         https://bugs.webkit.org/show_bug.cgi?id=59416
20882
20883         We should only re-run the check-for-weak-vtables-and-externals script
20884         if the executable is more recent than the last run.
20885
20886         * Scripts/check-for-weak-vtables-and-externals:
20887
20888 2011-04-26  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
20889
20890         Reviewed by Csaba Osztrogonác.
20891
20892         [Qt] Implement LayoutTestController::setAutofilled
20893         https://bugs.webkit.org/show_bug.cgi?id=59439
20894
20895         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
20896         (LayoutTestController::setAutofilled):
20897         * DumpRenderTree/qt/LayoutTestControllerQt.h:
20898
20899 2011-04-25  Dan Bernstein  <mitz@apple.com>
20900
20901         Reviewed by Mark Rowe.
20902
20903         Teach build-webkit where newer Xcode versions store the application-wide build settings user default.
20904
20905         * Scripts/webkitdirs.pm:
20906
20907 2011-04-25  Sam Weinig  <sam@webkit.org>
20908
20909         Reviewed by Adam Roben.
20910
20911         Leaks Viewer throws "Cannot post cyclic structures" in Safari 5
20912         https://bugs.webkit.org/show_bug.cgi?id=56090
20913
20914         Make LeaksViewer work with versions of browsers that don't support message
20915         passing of cyclic structures.
20916
20917         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParser.js:
20918         (LeaksParser.this._worker.onmessage):
20919         (LeaksParser):
20920         (LeaksParser.prototype.addLeaksFile):
20921         Use LeaksParserImpl directly if we don't support passing cyclic structures.
20922
20923         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserImpl.js: Copied from BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js.
20924         (LeaksParserImpl):
20925         (LeaksParserImpl.prototype.addLeaksFile):
20926         (LeaksParserImpl.prototype._incorporateLeaks):
20927         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js:
20928         (onmessage):
20929         Factor out LeaksParser implementation into LeaksParserImpl.js.
20930
20931         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Utilities.js:
20932         (workersSupportCyclicStructures):
20933         Add function to detect if workers support passing cyclic structures.
20934
20935         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html:
20936         Add LeaksParserImpl.js include.
20937
20938 2011-04-25  Jeff Miller  <jeffm@apple.com>
20939
20940         Reviewed by Timothy Hatcher.
20941
20942         Treat strings defined with UI_STRING_LOCALIZE_LATER_KEY() as debugging strings in extract-localizable-strings
20943         https://bugs.webkit.org/show_bug.cgi?id=59373
20944
20945         * Scripts/extract-localizable-strings: Add UI_STRING_LOCALIZE_LATER_KEY() to isDebugMacro var.
20946
20947 2011-04-25  Andrew Scherkus  <scherkus@chromium.org>
20948
20949         Reviewed by Dimitri Glazkov.
20950
20951         Adding a few Chromium contributors to committers.py.
20952
20953         https://bugs.webkit.org/show_bug.cgi?id=59327
20954
20955         * Scripts/webkitpy/common/config/committers.py:
20956
20957 2011-04-23  Dominic Cooney  <dominicc@chromium.org>
20958
20959         Reviewed by Dimitri Glazkov.
20960
20961         [V8] Nodes in shadow DOM should not be GCed while their hosts are alive
20962         https://bugs.webkit.org/show_bug.cgi?id=59284
20963
20964         Chromium DRT over-refcounted the NPObject wrappers it used to
20965         marshal shadowRoot nodes back to tests (WebBindings::makeNode
20966         births NPObjects with referenceCount == 1). This masked the GC bug
20967         59284 in tests by making shadowRoot nodes live forever.
20968
20969         * DumpRenderTree/chromium/LayoutTestController.cpp:
20970         (LayoutTestController::shadowRoot):
20971         (LayoutTestController::ensureShadowRoot):
20972
20973 2011-04-23  Kevin Ollivier  <kevino@theolliviers.com>
20974
20975         [wx] Unreviewed build fixes. Add new methods to wx LayoutTestController,
20976         and update the URL for downloading waf.
20977
20978         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
20979         (LayoutTestController::ensureShadowRoot):
20980         (LayoutTestController::removeShadowRoot):
20981         * Scripts/webkitdirs.pm:
20982
20983 2011-04-22  Sheriff Bot  <webkit.review.bot@gmail.com>
20984
20985         Unreviewed, rolling out r84627.
20986         http://trac.webkit.org/changeset/84627
20987         https://bugs.webkit.org/show_bug.cgi?id=59271
20988
20989         It broke Symbian build (Requested by Ossy on #webkit).
20990
20991         * Scripts/webkitpy/style/checker.py:
20992         * Scripts/webkitpy/style/checker_unittest.py:
20993
20994 2011-04-22  Dirk Pranke  <dpranke@chromium.org>
20995
20996         Reviewed by Adam Barth.
20997
20998         rebaseline-chromium-webkit-tests: clean up output
20999         https://bugs.webkit.org/show_bug.cgi?id=59240
21000
21001         This change dramatically changes the output the tool provides
21002         by default. The previous output (more or less) is still there
21003         in -v / verbose / debug mode, but the default output is
21004         way terser and more readable. It mostly says which files are
21005         updated, which are duplicates, and which are not found in an
21006         archive. Feedback desired :).
21007
21008         This change also adds a scm.exists() method to the SCM object
21009         so that we can cleanly figure out what we're doing while
21010         rebaselining.
21011
21012         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
21013         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
21014         * Scripts/webkitpy/common/system/filesystem_mock.py:
21015         * Scripts/webkitpy/common/checkout/scm.py:
21016         * Scripts/webkitpy/common/checkout/scm_unittest.py:
21017
21018 2011-04-22  Ilya Sherman  <isherman@chromium.org>
21019
21020         Reviewed by Alexey Proskuryakov.
21021
21022         Add Ilya Sherman (isherman@chromium.org) to the list of contributors
21023         https://bugs.webkit.org/show_bug.cgi?id=59231
21024
21025         * Scripts/webkitpy/common/config/committers.py:
21026
21027 2011-04-22  Yuta Kitamura  <yutak@chromium.org>
21028
21029         Reviewed by Ojan Vafai.
21030
21031         new-run-webkit-tests: Exception raised when stderr contains binary data
21032         https://bugs.webkit.org/show_bug.cgi?id=59083
21033
21034         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
21035         Use write_binary_file() instead of write_text_file().
21036
21037 2011-04-22  Chang Shu  <cshu@webkit.org>
21038
21039         Reviewed by Kenneth Rohde Christiansen.
21040
21041         WebKitTestRunner needs layoutTestController.setFrameFlatteningEnabled
21042         https://bugs.webkit.org/show_bug.cgi?id=42536
21043
21044         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
21045         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
21046         (WTR::LayoutTestController::setFrameFlatteningEnabled):
21047         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
21048
21049 2011-04-22  Adam Barth  <abarth@webkit.org>
21050
21051         Reviewed by Eric Seidel.
21052
21053         CommitQueue should remember the expected failures from patch to patch
21054         https://bugs.webkit.org/show_bug.cgi?id=59253
21055
21056         Previously, we stored the set of expected failures on the task object,
21057         which meant the commit-queue forgot about them after each patch.  This
21058         patch moves them to the CommitQueue object so they will have a longer
21059         lifetime.
21060
21061         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
21062         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
21063         * Scripts/webkitpy/tool/commands/queues.py:
21064
21065 2011-04-22  Sam Weinig  <sam@webkit.org>
21066
21067         Reviewed by Brian Weinstein.
21068
21069         WebKit2: Implement JavaScript unresponsiveness timer/callbacks
21070         https://bugs.webkit.org/show_bug.cgi?id=59248
21071
21072         Stub out shouldInterruptJavaScript callbacks.
21073
21074         * MiniBrowser/mac/BrowserWindowController.m:
21075         (-[BrowserWindowController awakeFromNib]):
21076         * MiniBrowser/win/BrowserView.cpp:
21077         (BrowserView::create):
21078         * WebKitTestRunner/TestController.cpp:
21079         (WTR::TestController::createOtherPage):
21080         (WTR::TestController::initialize):
21081
21082 2011-04-22  Daniel Bates  <dbates@webkit.org>
21083
21084         Reviewed by Csaba Osztrogonác.
21085
21086         webkit-patch land hangs if svn prompts for credentials
21087         https://bugs.webkit.org/show_bug.cgi?id=31500
21088
21089         Make has_authorization_for_realm() return True only if there exists a credentials file
21090         that contains either the word "password" or "passtype". We assume that these words don't
21091         coincide with the actual credential data (e.g. a person's username is "password").
21092
21093         * Scripts/webkitpy/common/checkout/scm.py:
21094         * Scripts/webkitpy/common/checkout/scm_unittest.py:
21095           - Added test cases:
21096             test_has_authorization_for_realm_using_credentials_with_passtype(),
21097             test_has_authorization_for_realm_using_credentials_with_password(),
21098             test_not_have_authorization_for_realm_with_credentials_missing_password_and_passtype()
21099           - Renamed test_not_have_authorization_for_realm() to test_not_have_authorization_for_realm_when_missing_credentials_file()
21100             to better describe what it's testing.
21101           - Repurposed test_has_authorization_for_realm() to take realm and credential data to use
21102             and return the result of calling has_authorization_for_realm() so that the caller can
21103             assert the result; Renamed to _test_has_authorization_for_realm_using_credentials() to
21104             better reflect its new purpose.
21105
21106 2011-04-22  Adam Barth  <abarth@webkit.org>
21107
21108         Remove double-negative.  (Sorry this was bugging me.)
21109
21110         * Scripts/webkitpy/tool/bot/commitqueuetask.py:
21111
21112 2011-04-22  Adam Barth  <abarth@webkit.org>
21113
21114         Reviewed by Eric Seidel.
21115
21116         Factor LayoutTestResultsReader out of CommitQueue
21117         https://bugs.webkit.org/show_bug.cgi?id=59244
21118
21119         This code will be shared with the EWS when they start running tests.
21120
21121         * Scripts/webkitpy/tool/bot/layouttestresultsreader.py: Added.
21122         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py: Added.
21123         * Scripts/webkitpy/tool/commands/queues.py:
21124         * Scripts/webkitpy/tool/commands/queues_unittest.py:
21125
21126 2011-04-22  Eric Seidel  <eric@webkit.org>
21127
21128         Reviewed by Dimitri Glazkov.
21129
21130         webkit-patch rebaseline is broken
21131         https://bugs.webkit.org/show_bug.cgi?id=59207
21132
21133         My recent changes to how LayoutTestResults.results_from_string
21134         uncovered a bug in our use of urllib2.
21135         We now are treating the return value from _fetch_results_html()
21136         (which returns urllib2.openurl()) as a string.  That's incorrect,
21137         the return value is a file-like object (which auto-converts to a
21138         string in most cases).
21139
21140         I've updated our urllib2-using code to correctly call .read() on the
21141         resulting object before treating it like a string.
21142         Unfortunately there is no easy way to test this mis-understanding
21143         of the openurl() semantics.
21144
21145         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
21146
21147 2011-04-22  Adam Barth  <abarth@webkit.org>
21148
21149         Reviewed by Eric Seidel.
21150
21151         Factor PatchAnalysis task out of CommitQueueTask
21152         https://bugs.webkit.org/show_bug.cgi?id=59220
21153
21154         This is a first step towards teaching the EWS to run tests.
21155
21156         * Scripts/webkitpy/tool/bot/commitqueuetask.py:
21157         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
21158         * Scripts/webkitpy/tool/bot/patchanalysistask.py: Added.
21159
21160 2011-04-22  Philippe Normand  <pnormand@igalia.com>
21161
21162         Reviewed by Martin Robinson.
21163
21164         [GTK] DRT: implement eventSender.scalePageBy
21165         https://bugs.webkit.org/show_bug.cgi?id=59082
21166
21167         * DumpRenderTree/gtk/EventSender.cpp:
21168         (scalePageByCallback): scalePageBy() support.
21169
21170 2011-04-21  Eric Seidel  <eric@webkit.org>
21171
21172         Reviewed by Adam Barth.
21173
21174         LayoutTestResults should know how to handle NRWT json files
21175         https://bugs.webkit.org/show_bug.cgi?id=59168
21176
21177         LayoutTestResults was originally created to be a parallel
21178         to existing NRWT architecture for dealing with results files.
21179         But then it turns out that NRWT has no such architecture. :(
21180         So this patch writes the necessary code to handle reading
21181         full_results.json and unexpected_results.json files from
21182         NRWT layout-test-results directories.
21183
21184         LayoutTestResults has thus morphed from being ORWT-only to
21185         supporting both ORWT and NRWT.  It's possible at some future
21186         point that other pieces of the NRWT architecture will learn
21187         how to read JSON files in which case this can go away.
21188
21189         This is all done for making it possible for the commit-queue
21190         to run the tests using NWRT and be able to understand the results
21191         (for flaky test reporting, etc.)
21192
21193         * Scripts/webkitpy/common/net/layouttestresults.py:
21194         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
21195         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
21196         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
21197         * Scripts/webkitpy/common/net/resultsjsonparser.py: Added.
21198         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py: Added.
21199
21200 2011-04-22  Yi Shen  <yi.4.shen@nokia.com>
21201
21202         Reviewed by Kenneth Rohde Christiansen.
21203
21204         [Qt] Upstream Symbian platform plugin
21205         https://bugs.webkit.org/show_bug.cgi?id=58435
21206
21207         Exempting directory WebKit/qt/symbian/platformplugin from style guide.
21208
21209         * Scripts/webkitpy/style/checker.py:
21210         * Scripts/webkitpy/style/checker_unittest.py:
21211
21212 2011-04-22  Balazs Kelemen  <kbalazs@webkit.org>
21213
21214         Reviewed by Csaba Osztrogonác.
21215
21216         [Qt][WK2] Qt should use the mac-wk2 skipped list
21217         https://bugs.webkit.org/show_bug.cgi?id=58536
21218
21219         * Scripts/old-run-webkit-tests:
21220
21221 2011-04-21  Tony Chang  <tony@chromium.org>
21222
21223         Reviewed by Mihai Parparita.
21224
21225         Remove find-mismatched-layout-test-results since it's not needed in a
21226         checksum free world.
21227         https://bugs.webkit.org/show_bug.cgi?id=59156
21228
21229         * Scripts/find-mismatched-layout-test-results: Removed.
21230
21231 2011-04-21  Brent Fulgham  <bfulgham@webkit.org>
21232
21233         Reviewed by Martin Robinson.
21234
21235         Teach build-webkit and update-webkit that the WinCairo port does
21236         not need the non-redistributable WebKitSupportLibraries.zip file.
21237         Also append the _Cairo_CFLite label to the build target for the
21238         default (no argument) case when building with '--wincairo'
21239
21240         https://bugs.webkit.org/show_bug.cgi?id=58801
21241
21242         * Scripts/build-webkit: Don't die if WebKitSupportLibraries.zip
21243           doesn't exist.
21244         * Scripts/webkitdirs.pm: Append '_Cairo_CFLite' to the default
21245           'Release' build target when passed the '--wincairo' flag.
21246         * Scripts/update-webkit: Don't die if WebKitSupportLibraries.zip
21247           doesn't exist.
21248
21249 2011-04-21  Dirk Pranke  <dpranke@chromium.org>
21250
21251         Reviewed by Eric Seidel.
21252
21253         new-run-webkit-tests: obsolete old code, part 5: rename message_broker2 to message_broker
21254         https://bugs.webkit.org/show_bug.cgi?id=58758
21255
21256         Since worker_mixin is now only used by worker, there's no point
21257         in having it split across two files.
21258
21259         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
21260         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
21261         * Scripts/webkitpy/layout_tests/layout_package/message_broker.py: Added
21262         * Scripts/webkitpy/layout_tests/layout_package/message_broker2.py: Removed.
21263         * Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py: * Added
21264         * Scripts/webkitpy/layout_tests/layout_package/message_broker2_unittest.py: Removed.
21265
21266 2011-04-21  Dirk Pranke  <dpranke@chromium.org>
21267
21268         Reviewed by Tony Chang.
21269
21270         new-run-webkit-tests: obsolete old code, part 4: merge worker_mixin into worker
21271         https://bugs.webkit.org/show_bug.cgi?id=58756
21272
21273         Since worker_mixin is now only used by worker, there's no point
21274         in having it split across two files.
21275
21276         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
21277         * Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py: Removed.
21278
21279 2011-04-21  Dirk Pranke  <dpranke@chromium.org>
21280
21281         Reviewed by Tony Chang.
21282
21283         new-run-webkit-tests: obsolete old code, part 3 - rename test_runner to manager
21284         https://bugs.webkit.org/show_bug.cgi?id=58754
21285
21286         test_runner hasn't been a great name for some time, since all of
21287         the test running is actually done in worker and
21288         single_test_runner. Rename to manager to match the terminology
21289         elsewhere.
21290
21291         * Scripts/webkitpy/layout_tests/layout_package/manager.py: Added.
21292         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py: Added.
21293         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
21294         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py: Removed.
21295         * Scripts/webkitpy/layout_tests/layout_package/test_runner_unittest.py: Removed.
21296         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
21297
21298 2011-04-21  Dirk Pranke  <dpranke@chromium.org>
21299
21300         Reviewed by Tony Chang.
21301
21302         new-run-webkit-tests: obsolete old threading code, part 2
21303         https://bugs.webkit.org/show_bug.cgi?id=58753
21304
21305         This merges the test_runner2 code back into test_runner since
21306         there's only a single type of test runner now.
21307
21308         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
21309         * Scripts/webkitpy/layout_tests/layout_package/test_runner2.py: Removed.
21310         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
21311
21312 2011-04-21  Maciej Stachowiak  <mjs@apple.com>
21313
21314         Reviewed by Adam Roben.
21315
21316         Add a feature define to allow <details> and <summary> to be disabled
21317         https://bugs.webkit.org/show_bug.cgi?id=59118
21318         <rdar://problem/9257045>
21319
21320         * Scripts/build-webkit:
21321         * Tools.pro:
21322
21323 2011-04-21  Adam Barth  <abarth@webkit.org>
21324
21325         Reviewed by Eric Seidel.
21326
21327         webkit-patch upload should assign the bug to you
21328         https://bugs.webkit.org/show_bug.cgi?id=33699
21329
21330         This patch only re-assigns the bug if it is currently unassigned.  We
21331         can be more aggressive in the future if people want us to be more
21332         agressive.
21333
21334         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
21335         * Scripts/webkitpy/tool/commands/upload_unittest.py:
21336         * Scripts/webkitpy/tool/mocktool.py:
21337         * Scripts/webkitpy/tool/steps/postdiff.py:
21338
21339 2011-04-21  Ojan Vafai  <ojan@chromium.org>
21340
21341         Reviewed by Tony Chang.
21342
21343         move json_results.html over to LayoutTests/fast/harness
21344         https://bugs.webkit.org/show_bug.cgi?id=59154
21345
21346         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
21347
21348 2011-04-21  Adam Barth  <abarth@webkit.org>
21349
21350         Reviewed by Eric Seidel.
21351
21352         ValidateChangeLogs doesn't work on SVN
21353         https://bugs.webkit.org/show_bug.cgi?id=59115
21354
21355         svn-create-patch lies about the diff in the working copy by moving
21356         ChangeLog entries to the top of the diff.  That's fine on most cases,
21357         but causes problems for ValidateChangeLogs, which is trying validate
21358         the where the ChangeLog entry appears.
21359
21360         I haven't added a test for this change because I couldn't figure out
21361         how to write one.  The issue is more of an integration issue, which
21362         we're not really set up to test in our unit testing framework.  If this
21363         patch had worked around the output from svn-create-patch, then I could
21364         have tested that we behave correctly on sample svn-create-patch output,
21365         but, in this case, I've removed the dependency on svn-create-patch.  I
21366         could test that we behave correctly on "svn diff" output, but we
21367         already have those tests.
21368
21369         * Scripts/webkitpy/tool/commands/download_unittest.py:
21370         * Scripts/webkitpy/tool/mocktool.py:
21371         * Scripts/webkitpy/tool/steps/validatechangelogs.py:
21372
21373 2011-04-21  Tony Chang  <tony@chromium.org>
21374
21375         Reviewed by Ojan Vafai.
21376
21377         old-run-webkit-tests/new-run-webkit-tests shouldn't read or write .checksum files
21378         https://bugs.webkit.org/show_bug.cgi?id=58402
21379
21380         * Scripts/old-run-webkit-tests: Don't read or write .checksum files.
21381         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py: Don't write .checksum files.
21382         * Scripts/webkitpy/layout_tests/port/base.py: Don't read .checksum files.
21383         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
21384         * Scripts/webkitpy/layout_tests/port/test.py: Update passing pngs to include a checksum.
21385         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
21386
21387 2011-04-21  Ojan Vafai  <ojan@chromium.org>
21388
21389         Reviewed by Tony Chang.
21390
21391         final tests for results.html file
21392         https://bugs.webkit.org/show_bug.cgi?id=59150
21393
21394         This gets us close to 100% code coverage. Once this is
21395         checked in, I'll move this file over to LayoutTests/fast/harness
21396         and make it a proper layout test.
21397
21398         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
21399         * Scripts/webkitpy/layout_tests/layout_package/json_results_test.js:
21400
21401 2011-04-21  Ojan Vafai  <ojan@chromium.org>
21402
21403         Reviewed by Tony Chang.
21404
21405         only expand visible rows
21406         https://bugs.webkit.org/show_bug.cgi?id=59130
21407
21408         For the chromium port, where there are many expected failures,
21409         expanding all of them is quite slow. Only expand the visible ones.
21410         This makes the default case. Also, make the expanding async to
21411         avoid totally killing the process.
21412
21413         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
21414         * Scripts/webkitpy/layout_tests/layout_package/json_results_test.js:
21415
21416 2011-04-21  Martin Robinson  <mrobinson@igalia.com>
21417
21418         [GTK] DRT: implement eventSender.scalePageBy
21419         https://bugs.webkit.org/show_bug.cgi?id=59082
21420
21421         Rolling out r84507 since it broke many GTK+ tests.
21422
21423         * DumpRenderTree/gtk/EventSender.cpp:
21424
21425 2011-04-21  Sam Weinig  <sam@webkit.org>
21426
21427         Reviewed by Simon Fraser.
21428
21429         Add way to garbage collect between each test when running layout tests
21430         https://bugs.webkit.org/show_bug.cgi?id=59126
21431
21432         * DumpRenderTree/mac/DumpRenderTree.mm:
21433         (initializeGlobalsFromCommandLineOptions):
21434         (runTest):
21435         * Scripts/old-run-webkit-tests:
21436         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
21437         (WTR::InjectedBundle::didReceiveMessage):
21438         * WebKitTestRunner/TestController.cpp:
21439         (WTR::TestController::TestController):
21440         (WTR::TestController::initialize):
21441         (WTR::TestController::resetStateToConsistentValues):
21442         * WebKitTestRunner/TestController.h:
21443
21444 2011-04-21  Ojan Vafai  <ojan@chromium.org>
21445
21446         Reviewed by Mihai Parparita.
21447
21448         fix another innocuous JS error in expand/collapseAllExpectations
21449         https://bugs.webkit.org/show_bug.cgi?id=59122
21450
21451         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
21452         * Scripts/webkitpy/layout_tests/layout_package/json_results_test.js:
21453
21454 2011-04-20  Ojan Vafai  <ojan@chromium.org>
21455
21456         Reviewed by Tony Chang.
21457
21458         put unexpected passes into their own table
21459         https://bugs.webkit.org/show_bug.cgi?id=59016
21460
21461         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
21462
21463 2011-04-20  Ojan Vafai  <ojan@chromium.org>
21464
21465         Reviewed by Tony Chang.
21466
21467         initial set of tests for new results.html file
21468         https://bugs.webkit.org/show_bug.cgi?id=59031
21469
21470         Also, fix the bugs (mostly innocuous JS errors) the tests found.
21471
21472         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
21473         * Scripts/webkitpy/layout_tests/layout_package/json_results_test.js: Added.
21474
21475 2011-04-21  Philippe Normand  <pnormand@igalia.com>
21476
21477         Reviewed by Martin Robinson.
21478
21479         [GTK] missing LayoutTestController::callShouldCloseOnWebView()
21480         https://bugs.webkit.org/show_bug.cgi?id=59086
21481
21482         callShouldCloseOnWebView() implementation.
21483
21484         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
21485         (LayoutTestController::callShouldCloseOnWebView):
21486
21487 2011-04-21  Philippe Normand  <pnormand@igalia.com>
21488
21489         Reviewed by Martin Robinson.
21490
21491         [GTK] DRT: implement eventSender.scalePageBy
21492         https://bugs.webkit.org/show_bug.cgi?id=59082
21493
21494         * DumpRenderTree/gtk/EventSender.cpp:
21495         (scalePageByCallback): scalePageBy() support.
21496
21497 2011-04-20  Dominic Cooney  <dominicc@chromium.org>
21498
21499         Reviewed by Dimitri Glazkov.
21500
21501         layoutTestController can create and destroy shadow DOM
21502         https://bugs.webkit.org/show_bug.cgi?id=59058
21503
21504         Chromium, Mac, GTK and Qt DRT; and WK2 test runner.
21505
21506         * DumpRenderTree/LayoutTestController.cpp:
21507         (ensureShadowRootCallback):
21508         (removeShadowRootCallback):
21509         (LayoutTestController::staticFunctions):
21510         * DumpRenderTree/LayoutTestController.h:
21511         * DumpRenderTree/chromium/LayoutTestController.cpp:
21512         (LayoutTestController::LayoutTestController):
21513         (LayoutTestController::ensureShadowRoot):
21514         (LayoutTestController::removeShadowRoot):
21515         * DumpRenderTree/chromium/LayoutTestController.h:
21516         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
21517         (LayoutTestController::ensureShadowRoot):
21518         (LayoutTestController::removeShadowRoot):
21519         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
21520         (LayoutTestController::shadowRoot):
21521         (LayoutTestController::ensureShadowRoot):
21522         (LayoutTestController::removeShadowRoot):
21523         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
21524         (LayoutTestController::ensureShadowRoot):
21525         (LayoutTestController::removeShadowRoot):
21526         * DumpRenderTree/qt/LayoutTestControllerQt.h:
21527         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
21528         (LayoutTestController::ensureShadowRoot):
21529         (LayoutTestController::removeShadowRoot):
21530         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
21531         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
21532         (WTR::LayoutTestController::shadowRoot):
21533         (WTR::LayoutTestController::ensureShadowRoot):
21534         (WTR::LayoutTestController::removeShadowRoot):
21535         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
21536
21537 2011-04-20  Dirk Pranke  <dpranke@chromium.org>
21538
21539         Reviewed by Eric Seidel.
21540
21541         new-run-webkit-tests: handle ctrl-c more cleanly
21542         https://bugs.webkit.org/show_bug.cgi?id=58849
21543
21544         * Scripts/new-run-webkit-tests:
21545         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
21546
21547 2011-04-20  Sheriff Bot  <webkit.review.bot@gmail.com>
21548
21549         Unreviewed, rolling out r84431.
21550         http://trac.webkit.org/changeset/84431
21551         https://bugs.webkit.org/show_bug.cgi?id=59061
21552
21553         Breaks for folks who don't have editbugs (Requested by abarth
21554         on #webkit).
21555
21556         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
21557         * Scripts/webkitpy/tool/commands/upload_unittest.py:
21558         * Scripts/webkitpy/tool/mocktool.py:
21559         * Scripts/webkitpy/tool/steps/postdiff.py:
21560
21561 2011-04-20  Leo Yang  <leo.yang@torchmobile.com.cn>
21562
21563         Unreviewed.
21564
21565         Adding myself to committers.py.
21566
21567         * Scripts/webkitpy/common/config/committers.py:
21568
21569 2011-04-20  Adam Barth  <abarth@webkit.org>
21570
21571         Reviewed by Eric Seidel.
21572
21573         webkit-patch upload should assign the bug to you
21574         https://bugs.webkit.org/show_bug.cgi?id=33699
21575
21576         This patch only re-assigns the bug if it is currently unassigned.  We
21577         can be more aggressive in the future if people want us to be more
21578         agressive.
21579
21580         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
21581         * Scripts/webkitpy/tool/commands/upload_unittest.py:
21582         * Scripts/webkitpy/tool/mocktool.py:
21583         * Scripts/webkitpy/tool/steps/postdiff.py:
21584
21585 2011-04-20 Peter Gal <galpeter@inf.u-szeged.hu>
21586
21587         Reviewed by Csaba Osztrogonác.
21588
21589         Modify builder.html template to allow easy access
21590         to list more builds on the builder's page.
21591
21592         https://bugs.webkit.org/show_bug.cgi?id=59008
21593
21594         * BuildSlaveSupport/build.webkit.org-config/templates/builder.html:
21595
21596 2011-04-20  Peter Gal  <galpeter@inf.u-szeged.hu>
21597
21598         Reviewed by Csaba Osztrogonác.
21599
21600         Add the default template for builder page.
21601
21602         https://bugs.webkit.org/show_bug.cgi?id=59008
21603
21604         * BuildSlaveSupport/build.webkit.org-config/templates/builder.html: Added.
21605
21606 2011-04-20  Ojan Vafai  <ojan@chromium.org>
21607
21608         Reviewed by Eric Seidel.
21609
21610         move page generation into a single function
21611         https://bugs.webkit.org/show_bug.cgi?id=59027
21612
21613         Now executing the script doesn't create any state. We should then
21614         be able to write tests for this by setting custom results data.
21615
21616         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
21617
21618 2011-04-20  Ojan Vafai  <ojan@chromium.org>
21619
21620         Reviewed by Eric Seidel.
21621
21622         move global state into a shared object
21623         https://bugs.webkit.org/show_bug.cgi?id=59023
21624
21625         This makes the code more testable.
21626
21627         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
21628
21629 2011-04-20  Ojan Vafai  <ojan@chromium.org>
21630
21631         Reviewed by Tony Chang.
21632
21633         move global state into a shared object
21634         https://bugs.webkit.org/show_bug.cgi?id=59023
21635
21636         This makes the code more testable.
21637
21638         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
21639
21640 2011-04-20  Ojan Vafai  <ojan@chromium.org>
21641
21642         Reviewed by Eric Seidel.
21643
21644         add expandAll/collapseAll links
21645         https://bugs.webkit.org/show_bug.cgi?id=59011
21646
21647         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
21648
21649 2011-04-20  Kevin Ollivier  <kevino@theolliviers.com>
21650
21651         [wx] Unreviewed build fixes for wx 2.9.1, Cocoa and 64-bit builds.
21652
21653         * Scripts/webkitdirs.pm:
21654         * wx/build/build_utils.py:
21655         * wx/build/settings.py:
21656         * wx/install-unix-extras:
21657         * wx/packaging/build-mac-installer.py:
21658         * wx/packaging/build-win-installer.py:
21659
21660 2011-04-20  Adam Roben  <aroben@apple.com>
21661
21662         Replace old-run-webkit-tests's win-7 platform with win-7sp0
21663
21664         It was an error that we had a win-7 platform to begin with. Windows 7 was the most recent
21665         version of Windows we cared about, so its results should just go in win. But now that
21666         Windows 7 SP1 exists (and has different results for some tests), we need a place to store
21667         results for Windows 7 without SP1 (and older versions of Windows).
21668
21669         Part of <http://webkit.org/b/56617> Several tests fail on Windows 7 SP1 due to font
21670         differences from Windows 7 "SP0"
21671
21672         Reviewed by David Kilzer.
21673
21674         * Scripts/old-run-webkit-tests: Replaced win-7 with win-7sp0. Newer versions of Windows 7
21675         will just use "win".
21676
21677         * Scripts/webkitdirs.pm:
21678         (isWindows7SP0): Added. Replaces isWindows7, because that function had no remaining callers.
21679
21680 2011-04-20  Adam Roben  <aroben@apple.com>
21681
21682         Teach webkitdirs::winVersion about build numbers
21683
21684         The build number isn't currently used, but soon will be.
21685
21686         As a bonus, winVersion now works on non-Cygwin variants of Perl!
21687
21688         Fixes <http://webkit.org/b/58991> Need a way to distinguish between Windows 7 and Windows 7
21689         SP1
21690
21691         Reviewed by Anders Carlsson.
21692
21693         * Scripts/webkitdirs.pm:
21694         (isAnyWindows): Added. Returns true if we're running on any Windows-ish Perl.
21695         (determineWinVersion): Use `cmd /c ver` to get the version number, including the build
21696         number. Parse the version number into a hash, similar to how osXVersion() works.
21697
21698         (isWindows7):
21699         (isWindowsVista):
21700         (isWindowsXP):
21701         Updated for changes to winVersion().
21702
21703 2011-04-20  W. James MacLean  <wjmaclean@chromium.org>
21704
21705         Reviewed by Kenneth Russell.
21706
21707         Add support for composite-to-texture to DumpRenderTree.
21708         https://bugs.webkit.org/show_bug.cgi?id=58675
21709
21710         * DumpRenderTree/chromium/DumpRenderTree.cpp:
21711         (main):
21712         * DumpRenderTree/chromium/TestShell.cpp:
21713         (TestShell::TestShell):
21714         (TestShell::resetWebSettings):
21715         * DumpRenderTree/chromium/TestShell.h:
21716         (TestShell::setCompositeToTexture):
21717         * DumpRenderTree/chromium/WebPreferences.cpp:
21718         (WebPreferences::reset):
21719         (WebPreferences::applyTo):
21720         * DumpRenderTree/chromium/WebPreferences.h:
21721
21722 2011-04-19  Eric Seidel  <eric@webkit.org>
21723
21724         Reviewed by Adam Barth.
21725
21726         The commit-queue is confused when non-layout tests fail
21727         https://bugs.webkit.org/show_bug.cgi?id=58955
21728
21729         As seen in http://queues.webkit.org/results/8474435
21730
21731         It fails to create an archive of the layout test results
21732         and then throws an exception.
21733
21734         I've fixed this by making it catch the exception as well
21735         as made it so it never calls that path in the common case.
21736
21737         I've updated MockFileSystem to understand rmtree's affect on
21738         directories as well as files.
21739
21740         * Scripts/webkitpy/common/system/filesystem_mock.py:
21741         * Scripts/webkitpy/common/system/workspace.py:
21742         * Scripts/webkitpy/tool/commands/queues.py:
21743         * Scripts/webkitpy/tool/commands/queues_unittest.py:
21744         * Scripts/webkitpy/tool/mocktool.py:
21745
21746 2011-04-19  Ojan Vafai  <ojan@chromium.org>
21747
21748         Fix check for whether a failure is expected to deal with
21749         flaky tests and FAIL expectations.
21750         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
21751
21752 2011-04-19  Ojan Vafai  <ojan@chromium.org>
21753
21754         Reviewed by Eric Seidel.
21755
21756         don't show missing text results for tests that only dump image results
21757         https://bugs.webkit.org/show_bug.cgi?id=58931
21758
21759         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
21760         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
21761
21762 2011-04-19  Ojan Vafai  <ojan@chromium.org>
21763
21764         Reviewed by Eric Seidel.
21765
21766         don't show failures table if there are only new tests
21767         https://bugs.webkit.org/show_bug.cgi?id=58929
21768
21769         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
21770
21771 2011-04-19  Ojan Vafai  <ojan@chromium.org>
21772
21773         Reviewed by Eric Seidel.
21774
21775         don't show failures table if there are only new tests
21776         https://bugs.webkit.org/show_bug.cgi?id=58929
21777
21778         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
21779
21780 2011-04-19  Ojan Vafai  <ojan@chromium.org>
21781
21782         Fix unittest failures from http://trac.webkit.org/changeset/84294.
21783
21784         * Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py:
21785         * Scripts/webkitpy/layout_tests/layout_package/test_runner_unittest.py:
21786
21787 2011-04-18  Ojan Vafai  <ojan@chromium.org>
21788
21789         Reviewed by Eric Seidel.
21790
21791         switch new-run-webkit-tests to using the new results file
21792         https://bugs.webkit.org/show_bug.cgi?id=58861
21793
21794         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
21795         * Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py:
21796         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
21797
21798 2011-04-19  Renata Hodovan  <reni@webkit.org>
21799
21800         Reviewed by Eric Seidel.
21801
21802         Move the alignment related macros from Vector.h to new Alignment.h.
21803         https://bugs.webkit.org/show_bug.cgi?id=56000
21804
21805         Adding the new file to the forwarding headers of Tools.
21806
21807         * DumpRenderTree/ForwardingHeaders/wtf/Alignment.h: Added.
21808
21809 2011-04-19  Kevin Ollivier  <kevino@theolliviers.com>
21810
21811         [wx] Unreviewed build fix for Win / wx 2.9 wxPython installation.
21812
21813         * wx/packaging/build-win-installer.py:
21814         * wx/packaging/wxWebKitInstaller.iss.in:
21815
21816 2011-04-19  Adam Roben  <aroben@apple.com>
21817
21818         Skip another accelerated-compositing-sensitive test when accelerated compositing is disabled
21819
21820         * Scripts/old-run-webkit-tests: Skip media/controls-without-preload.html on Windows when
21821         accelerated compositing is disabled.
21822
21823 2011-04-19  Ojan Vafai  <ojan@chromium.org>
21824
21825         Remove accidentally committed debug code.
21826         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
21827
21828 2011-04-19  Brent Fulgham  <bfulgham@webkit.org>
21829
21830         Unreviewed build correction.
21831
21832         Avoid duplicate stub implementation of test routine
21833         for WinCairo build.
21834
21835         * WebKitTestRunner/win/TestInvocationWin.cpp: Exclude
21836         WinCairo from using this stub.
21837
21838 2011-04-18  Ojan Vafai  <ojan@chromium.org>
21839
21840         Reviewed by Eric Seidel.
21841
21842         make results file work with audio and reftests
21843         https://bugs.webkit.org/show_bug.cgi?id=58860
21844
21845         Also fix bug with timeout tests and store a bit in the JSON
21846         for new image tests instead of loading the image result to check if it's there.
21847
21848         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
21849         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
21850
21851 2011-04-19  Jer Noble  <jer.noble@apple.com>
21852
21853         Reviewed by Adam Roben.
21854
21855         REGRESSION (r84206): 50 tests failing on SnowLeopard Intel Release (WebKit2 Tests) due to extra "supportsFullScreen() == true" output
21856         https://bugs.webkit.org/show_bug.cgi?id=58890
21857
21858         Add a layoutTestController flag which controls when full screen callbacks are written to stdout.
21859         Defaults to false.
21860
21861         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl: Added dumpFullScreenCallbacks().
21862         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
21863         (WTR::InjectedBundlePage::supportsFullScreen): Only write to stdout if shouldDumpFullScreenCallbacks is set.
21864         (WTR::InjectedBundlePage::enterFullScreenForElement): Ditto.
21865         (WTR::InjectedBundlePage::exitFullScreenForElement): Ditto.
21866         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
21867         (WTR::LayoutTestController::LayoutTestController): Initialize new ivar.
21868         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
21869         (WTR::LayoutTestController::dumpFullScreenCallbacks): Added.
21870         (WTR::LayoutTestController::shouldDumpFullScreenCallbacks): Added.
21871
21872 2011-04-19  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
21873
21874         Reviewed by Andreas Kling.
21875
21876         [Qt] fast/xmlhttprequest/xmlhttprequest-nonexistent-file.html fails (DRT sideeffect)
21877         https://bugs.webkit.org/show_bug.cgi?id=35086
21878
21879         fast/xmlhttprequest/xmlhttprequest-no-file-access.html caused
21880         a DRT sideeffect problem, because the resetSettings didn't set
21881         the default value of LocalContentCanAccessFileUrls.
21882
21883         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
21884         (WebCore::WebPage::resetSettings):
21885
21886 2011-04-18  Zan Dobersek  <zandobersek@gmail.com>
21887
21888         Reviewed by Martin Robinson.
21889
21890         [GTK] Need support for dumping focus rectangles in pixel results
21891         https://bugs.webkit.org/show_bug.cgi?id=53647
21892
21893         Dump the selection rect if it has been requested.
21894
21895         * DumpRenderTree/gtk/PixelDumpSupportGtk.cpp:
21896         (createBitmapContextFromWebView):
21897
21898 2011-04-18  Eric Seidel  <eric@webkit.org>
21899
21900         Reviewed by Adam Barth.
21901
21902         test-webkitpy should run the ActiveWorkItems tests
21903         https://bugs.webkit.org/show_bug.cgi?id=58859
21904
21905         Also make the unit tests pass. :)
21906
21907         * QueueStatusServer/model/activeworkitems.py:
21908         * QueueStatusServer/model/queue.py:
21909         * QueueStatusServer/model/workitems.py:
21910
21911 2011-04-18  Eric Seidel  <eric@webkit.org>
21912
21913         Reviewed by Adam Barth.
21914
21915         queues.webkit.org should show trailing-days pass counts for queues
21916         https://bugs.webkit.org/show_bug.cgi?id=58812
21917
21918         This lets us see if each bot is pulling its own weight or not.
21919         Before I wrote this patch our suspicious was that one of the
21920         cq bots was doing most of the work.  Turns out they're actually
21921         pretty even in their recent pass counts.
21922
21923         * QueueStatusServer/handlers/queuestatus.py:
21924         * QueueStatusServer/index.yaml:
21925         * QueueStatusServer/templates/queuestatus.html:
21926
21927 2011-04-18  Dirk Pranke  <dpranke@chromium.org>
21928
21929         Reviewed by Eric Seidel.
21930
21931         new-run-webkit-tests: obsolete old threading code, part 1
21932         https://bugs.webkit.org/show_bug.cgi?id=58752
21933
21934         This patch removes support for the 'old-inline' and
21935         'old-threads' worker models, and deletes the corresponding code.
21936         Now that the new code is stable, this stuff is no longer
21937         necessary.
21938
21939         * Scripts/webkitpy/layout_tests/layout_package/dump_render_tree_thread.py: Removed.
21940         * Scripts/webkitpy/layout_tests/layout_package/message_broker.py: Removed.
21941         * Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py: Removed.
21942         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
21943         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
21944         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
21945
21946 2011-04-18  Dirk Pranke  <dpranke@chromium.org>
21947
21948         Reviewed by Eric Seidel.
21949
21950         new-run-webkit-tests: add an --additional-drt-flag option
21951         https://bugs.webkit.org/show_bug.cgi?id=58680
21952
21953         NRWT has a bunch of command line flags that exist to pass
21954         custom flags to DRT, especially on chromium. It would be nice
21955         if there was a generic mechanism to pass through flags so we
21956         didn't have all the custom ones.
21957
21958         * Scripts/webkitpy/layout_tests/port/chromium.py:
21959         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
21960         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
21961         * Scripts/webkitpy/layout_tests/port/test.py:
21962         * Scripts/webkitpy/layout_tests/port/webkit.py:
21963         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
21964
21965 2011-04-05  Jer Noble  <jer.noble@apple.com>
21966
21967         Reviewed by Sam Weinig.
21968
21969         WebKit2: WKTR should support WebKit2 full screen APIs
21970         https://bugs.webkit.org/show_bug.cgi?id=56318
21971
21972         Respond to enterFullScreenForElement() and exitFullScreenForElement(), allowing
21973         WKTR to test the LayoutTest/fullscreen/ tests.
21974
21975         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
21976         (WTR::InjectedBundlePage::InjectedBundlePage): Add support for WKBundlePageFullScreenClient.
21977         (WTR::InjectedBundlePage::supportsFullScreen): Added.
21978         (WTR::InjectedBundlePage::enterFullScreenForElement): Added.
21979         (WTR::InjectedBundlePage::exitFullScreenForElement): Added.
21980         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
21981         * WebKitTestRunner/TestController.cpp:
21982         (WTR::TestController::resetStateToConsistentValues): Enable the full screen preference.
21983
21984 2011-04-18  Dirk Pranke  <dpranke@chromium.org>
21985
21986         Reviewed by Tony Chang.
21987
21988         new-run-webkit-tests: remove chromium-specific relpath hack for --results-directory
21989         https://bugs.webkit.org/show_bug.cgi?id=58388
21990
21991         * Scripts/webkitpy/layout_tests/port/chromium.py:
21992
21993 2011-04-18  Evan Martin  <evan@chromium.org>
21994
21995         Reviewed by Eric Seidel.
21996
21997         [chromium] expose title direction to webkit client
21998         https://bugs.webkit.org/show_bug.cgi?id=58823
21999
22000         * DumpRenderTree/chromium/LayoutTestController.cpp:
22001         (LayoutTestController::LayoutTestController):
22002         (LayoutTestController::reset):
22003         * DumpRenderTree/chromium/LayoutTestController.h:
22004         (LayoutTestController::setTitleTextDirection):
22005         Add member and setter to let us stash the title direction at load time.
22006
22007         * DumpRenderTree/chromium/WebViewHost.cpp:
22008         (WebViewHost::didReceiveTitle):
22009         * DumpRenderTree/chromium/WebViewHost.h:
22010         Adapt to new API; push the title into layoutTestController when we
22011         receive it.
22012
22013 2011-04-18  Dirk Pranke  <dpranke@chromium.org>
22014
22015         Unreviewed.
22016
22017         Re-land r84112. It was rolled out in r84127 but it didn't need
22018         to be. Only r84111 actually broke things.
22019
22020         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
22021         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
22022
22023 2011-04-18  Dirk Pranke  <dpranke@chromium.org>
22024
22025         Reviewed by Tony Chang.
22026
22027         new-run-webkit-tests: don't run more processes than necessary
22028         https://bugs.webkit.org/show_bug.cgi?id=58751
22029
22030         Previously NRWT would start up 1 process for each processor on
22031         the machine by default, which was a waste if we were only
22032         running a few shards of tests.
22033
22034         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
22035         * Scripts/webkitpy/layout_tests/layout_package/test_runner2.py:
22036         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
22037
22038 2011-04-18  Ojan Vafai  <ojan@chromium.org>
22039
22040         Reviewed by Eric Seidel.
22041
22042         include links to actual.png files for new pixel tests
22043         https://bugs.webkit.org/show_bug.cgi?id=58831
22044
22045         Load image elements to confirm whether the files exist.
22046         Image elements will load without being appended to the DOM.
22047
22048         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
22049
22050 2011-04-18  Ojan Vafai  <ojan@chromium.org>
22051
22052         Reviewed by Eric Seidel.
22053
22054         add a zoomed view for pixel results to the new results html file
22055         https://bugs.webkit.org/show_bug.cgi?id=58827
22056
22057         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
22058
22059 2011-04-18  Ojan Vafai  <ojan@chromium.org>
22060
22061         Reviewed by Eric Seidel.
22062
22063         add labels to results html page
22064         https://bugs.webkit.org/show_bug.cgi?id=58803
22065
22066         -fix iframe size to be the size of the png results
22067         -fix a bunch of aesthetic nits
22068         -fix some incorrectly styled code
22069
22070         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
22071
22072 2011-04-18  Chang Shu  <cshu@webkit.org>
22073
22074         Reviewed by Eric Seidel.
22075
22076         WebKitTestRunner needs layoutTestController.pathToLocalResource
22077         https://bugs.webkit.org/show_bug.cgi?id=42541
22078
22079         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
22080         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
22081         * WebKitTestRunner/InjectedBundle/mac/LayoutTestControllerMac.mm:
22082         (WTR::LayoutTestController::pathToLocalResource):
22083         * WebKitTestRunner/InjectedBundle/qt/LayoutTestControllerQt.cpp:
22084         (WTR::LayoutTestController::pathToLocalResource):
22085         * WebKitTestRunner/InjectedBundle/win/LayoutTestControllerWin.cpp:
22086         (WTR::LayoutTestController::pathToLocalResource):
22087
22088 2011-04-18  Timothy Hatcher  <timothy@apple.com>
22089
22090         Make update-webkit-localizable-strings put WebKit/win strings in WebCore
22091         now that all localized strings in WebKit/win use WEB_UI_STRING.
22092
22093         https://webkit.org/b/58747
22094
22095         Reviewed by Dan Bernstein.
22096
22097         * Scripts/update-webkit-localizable-strings: Remove the code that updated WebKit/win differently.
22098
22099 2011-04-18  Patrick Gansterer  <paroga@webkit.org>
22100
22101         Reviewed by Adam Roben.
22102
22103         [WIN] Use WCHAR instead of TCHAR
22104         https://bugs.webkit.org/show_bug.cgi?id=58755
22105
22106         We always use the UNICODE versions of windows functions, so
22107         the usage of TCHAR makes no sense and mixing them is bad style.
22108
22109         * WinLauncher/PrintWebUIDelegate.cpp:
22110           Also fix various style issues (including indentation and removal of ugly C-casts).
22111         * WinLauncher/WinLauncher.cpp:
22112
22113 2011-04-18  Dominic Cooney  <dominicc@chromium.org>
22114
22115         Reviewed by Andreas Kling.
22116
22117         Add layoutTestController.shadowRoot to Qt DRT.
22118         https://bugs.webkit.org/show_bug.cgi?id=58759
22119
22120         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
22121         (LayoutTestController::shadowRoot): Added.
22122         * DumpRenderTree/qt/LayoutTestControllerQt.h:
22123
22124 2011-04-18  Sheriff Bot  <webkit.review.bot@gmail.com>
22125
22126         Unreviewed, rolling out r84111.
22127         http://trac.webkit.org/changeset/84111
22128         https://bugs.webkit.org/show_bug.cgi?id=58771
22129
22130         Chromium linux layout tests are broken. (Requested by loislo2
22131         on #webkit).
22132
22133         * Scripts/webkitpy/layout_tests/port/chromium.py:
22134         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
22135         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
22136
22137 2011-04-18  Sheriff Bot  <webkit.review.bot@gmail.com>
22138
22139         Unreviewed, rolling out r84112.
22140         http://trac.webkit.org/changeset/84112
22141         https://bugs.webkit.org/show_bug.cgi?id=58770
22142
22143         Chromium linux layout tests are broken. (Requested by loislo2
22144         on #webkit).
22145
22146         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
22147         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
22148
22149 2011-04-17  Daniel Bates  <dbates@webkit.org>
22150
22151         Reviewed by Eric Seidel and Adam Barth.
22152
22153         webkit-patch land hangs if svn prompts for credentials
22154         https://bugs.webkit.org/show_bug.cgi?id=31500
22155
22156         Prompt for username and password when using git-svn and there aren't cached SVN credentials.
22157
22158         * Scripts/webkitpy/common/checkout/scm.py:
22159           - Added mixin class SVNRepository and made both class SVN and Git inherit from it.
22160           - Moved SVN.has_authorization_for_realm() to class SVNRepository and removed default value
22161             for argument realm; modified call sites as needed.
22162           - Modified AuthenticationError constructor to take optional prompt_for_password argument.
22163           - Modified {SCM, SVN, Git}.commit_with_message() to take optional password argument.
22164           - Modified Git._commit_on_branch() to take optional username and password argument.
22165           - Modified Git.push_local_commits_to_server() to take optional username and password
22166             argument and to call has_authorization_for_realm().
22167         * Scripts/webkitpy/common/checkout/scm_unittest.py:
22168           - Modified SVNTest.test_commit_without_authorization() to take dummy realm argument.
22169           - Modified SVNTest.test_not_have_authorization_for_realm() to pass realm argument to
22170             SVN.has_authorization_for_realm().
22171         * Scripts/webkitpy/common/net/credentials.py:
22172           - Modified Credentials.read_credentials() to call User.prompt_password() instead
22173             of using getpass.getpass() directly.
22174         * Scripts/webkitpy/common/system/user.py:
22175           - Added User.prompt_password().
22176         * Scripts/webkitpy/tool/steps/commit.py:
22177           - Modified Commit.run() to prompt for a password if needed.
22178
22179 2011-04-17  Dirk Pranke  <dpranke@chromium.org>
22180
22181         Reviewed by Ojan Vafai.
22182
22183         new-run-webkit-tests: save stderr into the layout-test-results dir
22184         https://bugs.webkit.org/show_bug.cgi?id=58690
22185
22186         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
22187         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
22188
22189 2011-04-17  Dirk Pranke  <dpranke@chromium.org>
22190
22191         Reviewed by Ojan Vafai.
22192
22193         new-run-webkit-tests: read stderr from chromium DRT separately
22194         https://bugs.webkit.org/show_bug.cgi?id=58708
22195
22196         * Scripts/webkitpy/layout_tests/port/chromium.py:
22197         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
22198         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
22199
22200 2011-04-17  Patrick Gansterer  <paroga@webkit.org>
22201
22202         Reviewed by Adam Barth.
22203
22204         Rename PLATFORM(CA) to USE(CA)
22205         https://bugs.webkit.org/show_bug.cgi?id=58742
22206
22207         * DumpRenderTree/mac/DumpRenderTree.mm:
22208         (resetDefaultsToConsistentValues):
22209
22210 2011-04-17  Patrick Gansterer  <paroga@webkit.org>
22211
22212         Reviewed by Adam Barth.
22213
22214         Rename PLATFORM(CG) to USE(CG)
22215         https://bugs.webkit.org/show_bug.cgi?id=58729
22216
22217         * DumpRenderTree/PixelDumpSupport.cpp:
22218         * DumpRenderTree/chromium/WebViewHost.cpp:
22219         (WebViewHost::paintRect):
22220         * DumpRenderTree/config.h:
22221         * DumpRenderTree/win/PixelDumpSupportWin.cpp:
22222         (createBitmapContextFromWebView):
22223         * TestWebKitAPI/PlatformWebView.h:
22224         * WebKitTestRunner/config.h:
22225         * WebKitTestRunner/win/TestInvocationWin.cpp:
22226
22227 2011-04-16  Patrick Gansterer  <paroga@webkit.org>
22228
22229         Reviewed by Eric Seidel.
22230
22231         Rename PLATFORM(CAIRO) to USE(CAIRO)
22232         https://bugs.webkit.org/show_bug.cgi?id=55192
22233
22234         * DumpRenderTree/PixelDumpSupport.cpp:
22235         * DumpRenderTree/config.h:
22236         * DumpRenderTree/win/PixelDumpSupportWin.cpp:
22237         (createBitmapContextFromWebView):
22238         * WebKitTestRunner/config.h:
22239
22240 2011-04-16  Ojan Vafai  <ojan@chromium.org>
22241
22242         Reviewed by Eric Seidel.
22243
22244         show diffs+test inline in the results page
22245         https://bugs.webkit.org/show_bug.cgi?id=58723
22246
22247         Put each row in it's own tbody and sort based on tbodies. That way
22248         we can append the results to the tbody inline as another table row.
22249
22250         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
22251
22252 2011-04-16  Ojan Vafai  <ojan@chromium.org>
22253
22254         Reviewed by Eric Seidel.
22255
22256         handle new tests in the new results html file
22257         https://bugs.webkit.org/show_bug.cgi?id=58715
22258
22259         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
22260
22261 2011-04-15  Dirk Pranke  <dpranke@chromium.org>
22262
22263         Reviewed by Ojan Vafai.
22264
22265         old-run-webkit-tests: save the list of tests actually run to a file
22266         https://bugs.webkit.org/show_bug.cgi?id=58692
22267
22268         * Scripts/old-run-webkit-tests:
22269
22270 2011-04-15  Dirk Pranke  <dpranke@chromium.org>
22271
22272         Reviewed by Ojan Vafai.
22273
22274         fix MockFileSystem.abspath() on windows
22275         https://bugs.webkit.org/show_bug.cgi?id=58713
22276
22277         The existing code was borked on Windows if a path was passed
22278         in using a drive letter (e.g., "d:\foo.txt"), and would go
22279         into an infinite recursive loop. While win paths shouldn't
22280         normally be passed into the mock filesystem, this just
22281         bulletproofs things a bit (note that "/foo.txt" is considered
22282         an absolute path on windows python as well.
22283
22284         * Scripts/webkitpy/common/system/filesystem_mock.py:
22285
22286 2011-04-15  Ojan Vafai  <ojan@chromium.org>
22287
22288         Reviewed by Eric Seidel.
22289
22290         show stderr files and httpd log files
22291         https://bugs.webkit.org/show_bug.cgi?id=58714
22292
22293         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
22294         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
22295
22296 2011-04-15  Shishir Agrawal  <shishir@chromium.org>
22297
22298         Reviewed by James Robinson.
22299
22300         Add a flag to guard Page Visibility API changes.
22301         https://bugs.webkit.org/show_bug.cgi?id=58464
22302
22303         * Scripts/build-webkit:
22304
22305 2011-04-15  Chang Shu  <cshu@webkit.org>
22306
22307         Reviewed by Alexey Proskuryakov.
22308
22309         When a message with url embedded is added to console, the "file:" scheme
22310         and path should be stripped.
22311         https://bugs.webkit.org/show_bug.cgi?id=58665
22312
22313         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
22314         (WTR::lastFileURLPathComponent):
22315         (WTR::InjectedBundlePage::willAddMessageToConsole):
22316
22317 2011-04-15  Jeff Miller  <jeffm@apple.com>
22318
22319         Reviewed by Sam Weinig.
22320
22321         Add takeFocus callback to WKPageUIClient
22322         https://bugs.webkit.org/show_bug.cgi?id=58686
22323
22324         Add empty WKPageUIClient entries for the takeFocus callback in the MiniBrowser and WebKitTestRunner.
22325
22326         * MiniBrowser/mac/BrowserWindowController.m:
22327         (-[BrowserWindowController awakeFromNib]): Added empty takeFocus entry.
22328         * MiniBrowser/win/BrowserView.cpp:
22329         (BrowserView::create): Added empty takeFocus entry.
22330         * WebKitTestRunner/TestController.cpp:
22331         (WTR::TestController::createOtherPage): Added empty takeFocus entry.
22332         (WTR::TestController::initialize): Added empty takeFocus entry.
22333
22334 2011-04-15  Ojan Vafai  <ojan@chromium.org>
22335
22336         Reviewed by Eric Seidel.
22337
22338         only show pretty-diff/wdiff if they were generated
22339         https://bugs.webkit.org/show_bug.cgi?id=58682
22340
22341         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
22342         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
22343         * Scripts/webkitpy/layout_tests/port/base.py:
22344
22345 2011-04-15  Andreas Kling  <kling@webkit.org>
22346
22347         Reviewed by Antonio Gomes.
22348
22349         [Qt] DRT: Default custom policy delegate decision should be Ignore.
22350
22351         If layoutTestController.setCustomPolicyDelegate() is called with only
22352         one argument, the second ("permissive") should default to false.
22353
22354         * DumpRenderTree/qt/LayoutTestControllerQt.h:
22355
22356 2011-04-15  Ojan Vafai  <ojan@chromium.org>
22357
22358         Reviewed by Eric Seidel.
22359
22360         fix path to layout test in new results file
22361         https://bugs.webkit.org/show_bug.cgi?id=58618
22362
22363         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
22364         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
22365
22366 2011-04-15  Ojan Vafai  <ojan@chromium.org>
22367
22368         Reviewed by Eric Seidel.
22369
22370         fix sorting in new results file
22371         https://bugs.webkit.org/show_bug.cgi?id=58616
22372
22373         -Fix the custom sort to return -1, 0, 1 instead of true/false.
22374         -Secondary sort by test name when sort values are equal.
22375         -Sort by test name by default.
22376
22377         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
22378
22379 2011-04-15  Dmitry Lomov  <dslomov@google.com>
22380
22381         Reviewed by David Levin.
22382
22383         check-webkit-style shouldn't complain about not including a primary header file
22384         if none exists
22385         https://bugs.webkit.org/show_bug.cgi?id=39514
22386
22387         * Scripts/webkitpy/style/checkers/cpp.py:
22388         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
22389
22390 2011-04-15  Eric Seidel  <eric@webkit.org>
22391
22392         Reviewed by Mihai Parparita.
22393
22394         queues.webkit.org should display when a bot last rebooted
22395         https://bugs.webkit.org/show_bug.cgi?id=58562
22396
22397         There is more repeated code here than I would like.  I fear
22398         my django-fu isn't quite up to snuff.
22399
22400         * QueueStatusServer/handlers/queuestatus.py:
22401         * QueueStatusServer/templates/queuestatus.html:
22402
22403 2011-04-14  Alok Priyadarshi  <alokp@chromium.org>
22404
22405         Reviewed by Adam Barth.
22406
22407         Remove dependency on chromium skia::PlatformCanvas
22408         https://bugs.webkit.org/show_bug.cgi?id=57563
22409
22410         Fixed compile error on Mac when using Core Graphics.
22411
22412         * DumpRenderTree/chromium/WebViewHost.cpp:
22413         (WebViewHost::paintRect):
22414
22415 2011-04-14  Alok Priyadarshi  <alokp@chromium.org>
22416
22417         Reviewed by James Robinson.
22418
22419         Remove dependency on chromium skia::PlatformCanvas
22420         https://bugs.webkit.org/show_bug.cgi?id=57563
22421
22422         * DumpRenderTree/chromium/TestShell.cpp:
22423         (makeCanvasOpaque):
22424         (TestShell::dumpImage):
22425         * DumpRenderTree/chromium/TestShell.h:
22426         * DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
22427         (WebThemeControlDRTWin::WebThemeControlDRTWin):
22428         (WebThemeControlDRTWin::draw):
22429         (WebThemeControlDRTWin::drawTextField):
22430         (WebThemeControlDRTWin::drawProgressBar):
22431         * DumpRenderTree/chromium/WebThemeControlDRTWin.h:
22432         * DumpRenderTree/chromium/WebViewHost.cpp:
22433         (WebViewHost::canvas):
22434         * DumpRenderTree/chromium/WebViewHost.h:
22435
22436 2011-04-14  Keith Kyzivat  <keith.kyzivat@nokia.com>
22437
22438         Reviewed by Laszlo Gombos.
22439
22440         [Qt] Add -maximize flag to QtTestBrowser and MiniBrowser
22441         https://bugs.webkit.org/show_bug.cgi?id=58007
22442
22443         Adds flag to start the test browsers maximized.
22444         Symbian started maximized on QtTestBrowser - made MiniBrowser follow
22445         suit.
22446         Also fixes some ToggleFullScreen issues so that prior window state is
22447         restored.
22448
22449         * MiniBrowser/qt/BrowserWindow.cpp:
22450         (BrowserWindow::BrowserWindow):
22451         (BrowserWindow::toggleFullScreenMode):
22452         * MiniBrowser/qt/MiniBrowserApplication.cpp:
22453         (MiniBrowserApplication::handleUserOptions):
22454         * MiniBrowser/qt/MiniBrowserApplication.h:
22455         (WindowOptions::WindowOptions):
22456         * QtTestBrowser/launcherwindow.cpp:
22457         (LauncherWindow::init):
22458         (LauncherWindow::toggleFullScreenMode):
22459         * QtTestBrowser/launcherwindow.h:
22460         (WindowOptions::WindowOptions):
22461         * QtTestBrowser/main.cpp:
22462         (LauncherApplication::handleUserOptions):
22463
22464 2011-04-14  Ojan Vafai  <ojan@chromium.org>
22465
22466         Reviewed by Eric Seidel.
22467
22468         add checkbox to only show unexpected results to new results file
22469         https://bugs.webkit.org/show_bug.cgi?id=58606
22470
22471         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
22472
22473 2011-04-13  Brent Fulgham  <bfulgham@webkit.org>
22474
22475         Reviewed by Martin Robinson.
22476
22477         [WinCairo] Implement ImageDiff Logic.
22478         http://bugs.webkit.org/show_bug.cgi?id=58486
22479
22480         * DumpRenderTree/win/ImageDiffCairo.cpp: Added.
22481         (strtof):
22482         (readFromData):
22483         (createImageFromStdin):
22484         (releaseMallocBuffer):
22485         (createDifferenceImage):
22486         (imageHasAlpha):
22487         (writeToData):
22488         (main):
22489         * DumpRenderTree/win/ImageDiff.vcproj: Updated for new
22490           ImageDiffCairo.cpp (excluding from Apple builds).
22491         * DumpRenderTree/win/ImageDiffDebugCairoCFLite.vsprops: Use
22492           new ImageDiffWinCairoCommon.vsprops file.
22493         * DumpRenderTree/win/ImageDiffReleaseCairoCFLite.vsprops: Use
22494           new ImageDiffWinCairoCommon.vsprops file.
22495         * DumpRenderTree/win/ImageDiffWinCairoCommon.vsprops: Added.
22496         * Scripts/old-run-webkit-tests: Make sure proper executable
22497           is built and executed for WinCairo port.
22498
22499 2011-04-14  Ojan Vafai  <ojan@chromium.org>
22500
22501         Reviewed by Eric Seidel.
22502
22503         only show expected failure type for ports that use an expectations file
22504         https://bugs.webkit.org/show_bug.cgi?id=58588
22505
22506         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
22507         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
22508
22509 2011-04-14  Eric Seidel  <eric@webkit.org>
22510
22511         Reviewed by Adam Barth.
22512
22513         commit-queue fails to catch IOError when results.html is missing
22514         https://bugs.webkit.org/show_bug.cgi?id=58589
22515
22516         * Scripts/webkitpy/tool/commands/queues.py:
22517         * Scripts/webkitpy/tool/commands/queues_unittest.py:
22518
22519 2011-04-14  Kevin Ollivier  <kevino@theolliviers.com>
22520
22521         [wx] Unreviewed build fix for Mac installation script.
22522
22523         * wx/packaging/build-mac-installer.py:
22524
22525 2011-04-14  Ojan Vafai  <ojan@chromium.org>
22526
22527         Reviewed by Tony Chang.
22528
22529         sort columns in the new json_results.html file
22530         https://bugs.webkit.org/show_bug.cgi?id=58581
22531
22532         Also fix padding on TDs and the path to layout tests.
22533
22534         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
22535
22536 2011-04-14  Eric Seidel  <eric@webkit.org>
22537
22538         Reviewed by Dimitri Glazkov.
22539
22540         queues.webkit.org should display when a bot last passed a patch
22541         https://bugs.webkit.org/show_bug.cgi?id=58546
22542
22543         I am suspicious that some of our commit-queue instances are never
22544         actually passing anything.
22545
22546         This will let us know if those bots are never passing patches.
22547
22548         * QueueStatusServer/handlers/queuestatus.py:
22549         * QueueStatusServer/index.yaml:
22550         * QueueStatusServer/templates/queuestatus.html:
22551
22552 2011-04-14  Dirk Pranke  <dpranke@chromium.org>
22553
22554         Reviewed by Adam Roben.
22555
22556         REGRESSION (r83764): webkitpy.layout_tests.port.chromium_win_unittest.ChromiumWinTest.test_default_worker_model fails on apple-windows-13
22557         https://bugs.webkit.org/show_bug.cgi?id=58545
22558
22559         Remove platform-specific overrides; the default behavior should
22560         be working everywhere now, I think.
22561
22562         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
22563         * Scripts/webkitpy/layout_tests/port/mac.py:
22564
22565 2011-04-14  Eric Seidel  <eric@webkit.org>
22566
22567         Reviewed by Adam Barth.
22568
22569         commit-queues are hitting an exception trying to land when the tree is red
22570         https://bugs.webkit.org/show_bug.cgi?id=58558
22571
22572         Just an un-tested (and thus broken) code path from yesterday's changes.
22573
22574         * Scripts/webkitpy/common/net/layouttestresults.py:
22575         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
22576         * Scripts/webkitpy/tool/commands/queues.py:
22577         * Scripts/webkitpy/tool/commands/queues_unittest.py:
22578
22579 2011-04-14  Kevin Ollivier  <kevino@theolliviers.com>
22580
22581         [wx] Unreviewed build fixes for wxMSW and wx 2.9.1.1.
22582
22583         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
22584         (LayoutTestController::shadowRoot):
22585         * wx/build/wxpresets.py:
22586
22587 2011-04-14  Victoria Kirst  <vrk@google.com>
22588
22589         Reviewed by James Robinson.
22590
22591         [chromium] Failing GPU video tests
22592         https://bugs.webkit.org/show_bug.cgi?id=57926
22593
22594         Added/updated GPU directory names for rebaseline tool.
22595
22596         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
22597
22598 2011-04-13  Daniel Bates  <dbates@rim.com>
22599
22600         Reviewed by Adam Barth.
22601
22602         Perl unit test removeEOL.pl should print test case name on failure
22603         https://bugs.webkit.org/show_bug.cgi?id=58513
22604
22605         * Scripts/webkitperl/VCSUtils_unittest/removeEOL.pl: Pass $title as
22606           second argument of Test::Simple::ok().
22607
22608 2011-04-13  Kent Tamura  <tkent@chromium.org>
22609
22610         Reviewed by Eric Seidel.
22611
22612         gdb-safari should pass arguments to safari.
22613         https://bugs.webkit.org/show_bug.cgi?id=58439
22614
22615         We can do:
22616           gdb-safari --debug LayoutTests/.../foo.html
22617
22618         * Scripts/gdb-safari: Passing @ARGV to Safari via gdb --arg.
22619
22620 2011-04-13  Dirk Pranke  <dpranke@chromium.org>
22621
22622         Unreviewed, build fix.
22623         https://bugs.webkit.org/show_bug.cgi?id=52763
22624
22625         r83799 broke test-webkitpy because it assumed we always have
22626         access to a real filesystem (we don't during the unit tests).
22627         Stubbing out the copy for now but we might need a better
22628         solution at some point.
22629
22630         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
22631
22632 2011-04-13  Dirk Pranke  <dpranke@chromium.org>
22633
22634         Reviewed by Tony Chang.
22635
22636         new-run-webkit-tests --run-singly option is busted
22637         https://bugs.webkit.org/show_bug.cgi?id=55909
22638
22639         Python's broken lexical scoping strikes again :(.
22640
22641         * Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py:
22642         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
22643
22644 2011-04-13  Ojan Vafai  <ojan@chromium.org>
22645
22646         Reviewed by Eric Seidel.
22647
22648         generated unexpected_results.html from unexpected_results.json
22649         https://bugs.webkit.org/show_bug.cgi?id=52763
22650
22651         Eventually, we'll merge this with results.html and have a single richer results page.
22652         For now, I just want to get something checked in that we can iterate on.
22653
22654         * Scripts/webkitpy/layout_tests/layout_package/json_results.html: Added.
22655         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
22656
22657 2011-04-13  Eric Seidel  <eric@webkit.org>
22658
22659         Reviewed by Adam Barth.
22660
22661         Up the failure limit to allow the commit-queue to actually land when the tree is red
22662         https://bugs.webkit.org/show_bug.cgi?id=58499
22663
22664         This change increases the --exit-after-N-failures limit used by the commit-queue
22665         from 1 to 10.  This will cause the code added in bug 58494 to actually get
22666         exercised and the queues should start being able to land when the trees are red.
22667
22668         When testing I found that test_runtests_leopard_commit_queue_hack_command was the
22669         only unit test to actually detect this change!  Since we don't actually
22670         run on leopard anymore, I removed the hack code and repurposed the test.
22671
22672         * Scripts/webkitpy/tool/bot/commitqueuetask.py:
22673         * Scripts/webkitpy/tool/steps/runtests.py:
22674         * Scripts/webkitpy/tool/steps/steps_unittest.py:
22675
22676 2011-04-13  Eric Seidel  <eric@webkit.org>
22677
22678         Reviewed by Adam Barth.
22679
22680         commit-queue should be able to land when tree is red
22681         https://bugs.webkit.org/show_bug.cgi?id=58494
22682
22683         There is some yak hair on my hands, I will admit.
22684
22685         This change is mostly about adding an ExpectedFailures
22686         class to track when the bots are red and we should be
22687         ignoring failures when landing from the commit-queue.
22688
22689         However, to make intelligent decisions about patches we
22690         need to know whether the run hit the --exit-after-N-failures limit
22691         or not.  Right now that information is not saved off in results.html
22692         so we have to pull the information from RunTests.
22693
22694         I've plumbed the --exit-after-N-failures information into
22695         LayoutTestResults for now to make the ExpectedFailures code cleaner.
22696
22697         As a result of adding all these additional calls to delegate.layout_test_results()
22698         I broke some of our flaky test detection tests and had to re-write them
22699         to not depend on the number of layout_test_results code.
22700
22701         At the same time I updated the commit-queue to use the newer filesystem
22702         API (to allow us to use MockFileSystem) which required further changes
22703         to the layout tests.  Changes were required in either case, since
22704         we're now calling layout_test_results() in more cases, which previously
22705         would try and hit the disk (until I moved it to use tool.filesystem).
22706
22707         I should note that *all* of this code is disabled for now, since our
22708         --exit-after-N-failures limit is currently 1!  (Thus were always in the
22709         case where we can't actually tell if the layout test results are legit.)
22710         I will up that limit in a second patch (which may require a couple more unit test tweaks).
22711
22712         * Scripts/webkitpy/common/net/layouttestresults.py:
22713         * Scripts/webkitpy/tool/bot/commitqueuetask.py:
22714         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
22715         * Scripts/webkitpy/tool/bot/expectedfailures.py: Added.
22716         * Scripts/webkitpy/tool/bot/expectedfailures_unittest.py: Added.
22717         * Scripts/webkitpy/tool/commands/queues.py:
22718         * Scripts/webkitpy/tool/commands/queues_unittest.py:
22719         * Scripts/webkitpy/tool/commands/queuestest.py:
22720         * Scripts/webkitpy/tool/steps/runtests.py:
22721
22722 2011-04-13  Brent Fulgham  <bfulgham@webkit.org>
22723
22724         Unreviewed build correction.
22725
22726         * TestWebKitAPI/PlatformWebView.h: Limit include of
22727           <CoreGraphics/CGGeometry> to PLATFORM(CG) users.
22728
22729 2011-04-13  Mihai Parparita  <mihaip@chromium.org>
22730
22731         Reviewed by Eric Seidel.
22732
22733         [NRWT] Rename --baseline-search-patch to --additional-platform-directory to match ORWT
22734         https://bugs.webkit.org/show_bug.cgi?id=58489
22735
22736         r83743 added --additional-platform-directory to old-run-webkit-tests,
22737         rename the equivalent flag in new-run-webkit-tests to be consistent.
22738
22739         * Scripts/webkitpy/layout_tests/port/base.py:
22740         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
22741         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
22742         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
22743
22744 2011-04-13  Mihai Parparita  <mihaip@chromium.org>
22745
22746         Reviewed by Eric Seidel.
22747
22748         Rebaseline queue server
22749         https://bugs.webkit.org/show_bug.cgi?id=57891
22750
22751         Adds a simple rebaseline queue server (meant to run at
22752         http://rebaseline-queue.appspot.com/). It presents a list of all (NRWT)
22753         builders and their failing tests, and allows those tests to be added to
22754         a per-builer rebaseline queue.
22755
22756         A subsequent change will add a mode to NRWT (and/or a wrapper script)
22757         for pulling the tests to be rebaselined from the server and running them
22758         with --reset-results (the initial use-case is for the GPU hardware bots,
22759         which do not have checked-in baselines, therefore there is no need for a
22760         check-in step yet).
22761
22762         * RebaselineQueueServer/app.yaml: Added.
22763         * RebaselineQueueServer/handlers/__init__.py: Added.
22764         * RebaselineQueueServer/handlers/builderqueue.py: Added.
22765         * RebaselineQueueServer/handlers/pages.py: Added.
22766         * RebaselineQueueServer/index.yaml: Added.
22767         * RebaselineQueueServer/main.py: Added.
22768         * RebaselineQueueServer/model/__init__.py: Added.
22769         * RebaselineQueueServer/model/queueentry.py: Added.
22770         * RebaselineQueueServer/static/builder-frame-empty.html: Added.
22771         * RebaselineQueueServer/static/styles.css: Added.
22772         * RebaselineQueueServer/templates/builder-picker.html: Added.
22773         * RebaselineQueueServer/templates/builder-queue-edit.html: Added.
22774         * RebaselineQueueServer/templates/builder-queue-list.html: Added.
22775         * RebaselineQueueServer/templates/home.html: Added.
22776
22777 2011-04-13  Dirk Pranke  <dpranke@chromium.org>
22778
22779         Reviewed by Mihai Parparita.
22780
22781         nrwt: enable multiple processes by default on Chromium Win.
22782         https://bugs.webkit.org/show_bug.cgi?id=55163
22783
22784         Re-land r79268; it should be more stable now and will hopefully
22785         work.
22786
22787         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
22788         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
22789
22790 2011-04-13  Dirk Pranke  <dpranke@chromium.org>
22791
22792         Reviewed by Ojan Vafai.
22793
22794         new-run-webkit-tests: configure logging in child processes properly
22795         https://bugs.webkit.org/show_bug.cgi?id=58296
22796
22797         Logging propagated properly from manager to worker processes on
22798         Unix but not on Windows; this fixes that.
22799
22800         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
22801         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
22802
22803 2011-04-13  Dirk Pranke  <dpranke@chromium.org>
22804
22805         Reviewed by Ojan Vafai.
22806
22807         new-run-webkit-tests: suppress extraneous pretty patch warnings
22808         https://bugs.webkit.org/show_bug.cgi?id=58295
22809
22810         NRWT wasn't being careful about when it logged messages from
22811         PrettyPatch not being available; it should only log during
22812         check_build(), and be silent otherwise. This was causing us to
22813         get multiple errors at runtime, which was confusing.
22814
22815         * Scripts/webkitpy/layout_tests/port/base.py:
22816         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
22817         * Scripts/webkitpy/layout_tests/port/chromium.py:
22818
22819 2011-04-13  Dirk Pranke  <dpranke@chromium.org>
22820
22821         Reviewed by Jian Li.
22822
22823         add chromium-linux-x86_64 to rebaseline-chromium-webkit-tests
22824         https://bugs.webkit.org/show_bug.cgi?id=58461
22825
22826         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
22827
22828 2011-04-13  Tony Chang  <tony@chromium.org>
22829
22830         Unreviewed, fix NRWT.  Stack:
22831
22832             return super(ChromiumPort, self).results_directory(self)
22833         TypeError: results_directory() takes exactly 1 argument (2 given)
22834
22835         * Scripts/webkitpy/layout_tests/port/chromium.py:
22836
22837 2011-04-13  Adam Roben  <aroben@apple.com>
22838
22839         Teach ORWT about extra platform-specific directories to be searched before the default ones
22840
22841         Fixes <http://webkit.org/b/58446> ORWT needs a way to look in a user-specified directory for
22842         platform-specific results
22843
22844         Reviewed by Anders Carlsson.
22845
22846         * Scripts/old-run-webkit-tests:
22847         (top-level): Added @additionalPlatformDirectories and hooked up
22848         --additional-platform-directories to populate it.
22849         (buildPlatformResultHierarchy): Prepend the additional platform directories to the default
22850         list (after filtering out any of the additional directories that don't exist).
22851
22852 2011-04-13  Lucas Forschler  <lforschler@apple.com>
22853
22854         Reviewed by Adam Roben.
22855
22856         Add the /T command to taskkill.exe.
22857         This will kill spawned child processes.
22858
22859         * BuildSlaveSupport/win/kill-old-processes:
22860
22861 2011-04-13  Lucas Forschler  <lforschler@apple.com>
22862
22863         Reviewed by Adam Roben.
22864
22865         Sort kill list alphabeticaly.
22866         Fixed capitalization on VcBuildHelper.
22867
22868         * BuildSlaveSupport/win/kill-old-processes:
22869
22870 2011-04-13  Lucas Forschler  <lforschler@apple.com>
22871
22872         Reviewed by Adam Roben.
22873
22874         https://bugs.webkit.org/show_bug.cgi?id=58384
22875         Update kill-old-processes to include a few new ones.
22876
22877         * BuildSlaveSupport/win/kill-old-processes:
22878
22879 2011-04-13  Gustavo Noronha Silva  <gns@gnome.org>
22880
22881         Reviewed by Martin Robinson.
22882
22883         [GTK] PluginPackage should check whether a plugin mixes GTK+ 2 and 3 symbols itself
22884         https://bugs.webkit.org/show_bug.cgi?id=58297
22885
22886         Do not disable flash specifically when built with GTK+ 3 -
22887         PluginPackage should do the right thing while loading plugins.
22888
22889         * GtkLauncher/main.c:
22890         (main):
22891
22892 2011-04-12  Kent Tamura  <tkent@chromium.org>
22893
22894         Reviewed by Dimitri Glazkov.
22895
22896         [Chromium] Add WebSettings::setValidationMessageTimerMagnification()
22897         https://bugs.webkit.org/show_bug.cgi?id=57426
22898
22899         * DumpRenderTree/chromium/WebPreferences.cpp:
22900         (WebPreferences::applyTo): Disable the validation message timer.
22901
22902 2011-04-12  James Kozianski  <koz@chromium.org>
22903
22904         Unreviewed.
22905
22906         Adding myself to committers.py.
22907
22908         * Scripts/webkitpy/common/config/committers.py:
22909
22910 2011-04-12  Brent Fulgham  <bfulgham@webkit.org>
22911
22912         Unreviewed build correction for WinCairo bot after r83639.
22913         https://bugs.webkit.org/show_bug.cgi?id=51790
22914
22915         Correct typo for the new "update-webkit-wincairo-libs" script.
22916
22917         * Scripts/build-webkit: Incorrectly calling the
22918           update-webkit-cairo-libs which does not exist.
22919
22920 2011-04-12  Alice Liu  <alice.liu@apple.com>
22921
22922         Reviewed by Jon Honeycutt.
22923
22924         A change that should have been part of http://trac.webkit.org/changeset/83628
22925
22926         * DumpRenderTree/win/DumpRenderTree.cpp:
22927         (resetDefaultsToConsistentValues): Setting should be off for most tests. Reset to false.
22928
22929 2011-04-12  Dirk Pranke  <dpranke@chromium.org>
22930
22931         Reviewed by Tony Chang.
22932
22933         new-run-webkit-tests: --results-directory is relative to builddir, not $PWD
22934         https://bugs.webkit.org/show_bug.cgi?id=58272
22935
22936         NRWT was interpreting the --results-directory cmd line arg as
22937         relative to the build directory, not the current working
22938         directory (ORWT uses the latter, which is much more intuitive).
22939
22940         This patch fixes the base case, but includes an override for
22941         Chromium that is needed until the bots can be updated.
22942
22943         * Scripts/webkitpy/layout_tests/port/base.py:
22944         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
22945         * Scripts/webkitpy/layout_tests/port/chromium.py:
22946         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
22947         * Scripts/webkitpy/layout_tests/port/test.py:
22948         * Scripts/webkitpy/layout_tests/port/webkit.py:
22949         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
22950         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
22951
22952 2011-04-12  Philippe Normand  <pnormand@igalia.com>
22953
22954         Unreviewed, roll out r83621 as it broke the GTK build.
22955
22956         [GTK] arguments passed to build-jsc and build-webkit scripts are not taken into account
22957         https://bugs.webkit.org/show_bug.cgi?id=58333
22958
22959         * Scripts/build-jsc:
22960         * Scripts/build-webkit:
22961
22962 2011-04-12  Carl Lobo  <carllobo@gmail.com> and Brent Fulgham <bfulgham@webkit.org>
22963
22964         Reviewed by Adam Roben.
22965
22966         Patch to download the WinCairo dependancies as part of build-webkit.
22967         https://bugs.webkit.org/show_bug.cgi?id=51790
22968
22969         This patch downloads the dependancies from
22970             http://idisk.mac.com/bfulgham-Public/WinCairoRequirements.zip
22971         checking the modification timestamp etc.
22972
22973         * Tools/Scripts/build-webkit
22974         * Tools/Scripts/update-webkit: Modified to understand the --wincairo
22975           flag, and to update WinCairo dependencies when used.
22976         * Tools/Scripts/update-webkit-cairo-libs: added
22977         * Tools/Scripts/update-webkit-dependency: added
22978         * Tools/Scripts/update-webkit-auxiliary-libs
22979
22980 2011-04-12  Jeff Miller  <jeffm@apple.com>
22981
22982         Reviewed by Adam Roben.
22983
22984         WebKit2: Pressing Tab in Web Inspector's console does not cycle through completion options
22985         https://bugs.webkit.org/show_bug.cgi?id=56020
22986
22987         Don't call TranslateMessage() in the MiniBrowser or TestWebKitAPI for key messages destined for a WebKit2 view,
22988         since WebKit will do this for us.  If we didn't do this, TranslateMessage() would be called twice,
22989         which would generate two characters for every keypress (for example).  I didn't bother doing this for
22990         WebKitTestRunner, since it doesn't get any WM_KEYDOWN events.
22991
22992         Add new WebKit2/TranslateMessageGeneratesWMChar test to test expected TranslateMessage() behavior.
22993
22994         * MiniBrowser/win/main.cpp:
22995         (shouldTranslateMessage): Added.
22996         (_tWinMain): Don't call TranslateMessage() unless shouldTranslateMessage() says to.
22997         * TestWebKitAPI/PlatformUtilities.h: Added shouldTranslateMessage() on Windows.
22998         * TestWebKitAPI/PlatformWebView.h: Added simulateAKeyDown().
22999         * TestWebKitAPI/Tests/WebKit2/win/TranslateMessageGeneratesWMChar.cpp: Added.
23000         (TestWebKitAPI::didNotHandleKeyEventCallback): Added.
23001         (TestWebKitAPI::runAndWatchForWMChar): Added.
23002         (TestWebKitAPI::TEST): Added.
23003         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp:
23004         (TestWebKitAPI::Util::run): Don't call TranslateMessage() unless shouldTranslateMessage() says to.
23005         (TestWebKitAPI::Util::shouldTranslateMessage): Added.
23006         * TestWebKitAPI/win/PlatformWebViewWin.cpp:
23007         (TestWebKitAPI::PlatformWebView::simulateAKeyDown): Added.
23008         * TestWebKitAPI/win/TestWebKitAPI.vcproj: Added TranslateMessageGeneratesWMChar.cpp.
23009
23010 2011-04-12  Dirk Pranke  <dpranke@chromium.org>
23011
23012         Reviewed by Tony Chang.
23013
23014         new-run-webkit-tests: update code to use port.results_directory() instead of options.results_directory
23015         https://bugs.webkit.org/show_bug.cgi?id=58290
23016
23017         This is some preliminary refactoring for bug 58272.
23018
23019         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
23020         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
23021         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
23022         * Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py:
23023         * Scripts/webkitpy/layout_tests/port/test.py:
23024         * Scripts/webkitpy/layout_tests/port/chromium.py:
23025         * Scripts/webkitpy/layout_tests/port/base.py:
23026         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
23027
23028 2011-04-12  Dirk Pranke  <dpranke@chromium.org>
23029
23030         Reviewed by Eric Seidel.
23031
23032         webkitpy: fix mock_filesystem abspath to handle relative paths
23033         and add filesystem.chdir() and filesystem.getcwd() to be able
23034         to test this and mock it out.
23035
23036         https://bugs.webkit.org/show_bug.cgi?id=58288
23037
23038         * Scripts/webkitpy/common/system/filesystem.py:
23039         * Scripts/webkitpy/common/system/filesystem_mock.py:
23040         * Scripts/webkitpy/common/system/filesystem_unittest.py:
23041
23042 2011-04-12  Alice Liu  <alice.liu@apple.com>
23043
23044         Reviewed by Sam Weinig.
23045
23046         https://bugs.webkit.org/show_bug.cgi?id=58292
23047         Provide new setting to allow site icon loading despite disabling automatic image loading in general.
23048
23049         * DumpRenderTree/mac/DumpRenderTree.mm:
23050         (resetDefaultsToConsistentValues): Setting should be off for most tests. Reset to false.
23051
23052 2011-04-12  Thouraya ANDOLSI  <thouraya.andolsi@st.com>
23053
23054         Reviewed by Eric Seidel.
23055
23056         [Qt] Enable JIT build for SH4 platforms.
23057         https://bugs.webkit.org/show_bug.cgi?id=58317
23058
23059
23060         * DumpRenderTree/qt/DumpRenderTree.pro:
23061         * QtTestBrowser/QtTestBrowser.pro:
23062         * Scripts/webkitdirs.pm:
23063
23064 2011-04-12  Tony Chang  <tony@chromium.org>
23065
23066         Reviewed by Andreas Kling.
23067
23068         [Qt] embed checksums in PNGs written by Qt-DRT
23069         https://bugs.webkit.org/show_bug.cgi?id=58173
23070
23071         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
23072         (WebCore::DumpRenderTree::dump):
23073
23074 2011-04-12  Eric Seidel  <eric@webkit.org>
23075
23076         Reviewed by Adam Barth.
23077
23078         commit-queue should upload failure diffs when tests fail
23079         https://bugs.webkit.org/show_bug.cgi?id=58348
23080
23081         This change was mostly just plumbing.  We were already saving
23082         this information for flaky test reporting.  I just made it possible
23083         for normal failures to report archives as well.
23084
23085         I did a little abstraction work to try and share some code between
23086         flakytestreporter.py and this new code.  There is still more we could do.
23087
23088         In making this change I also went through and updated the various
23089         places we have urls hard-coded in our python and pointed them at
23090         common.config.urls.
23091
23092         * Scripts/webkitpy/common/checkout/scm.py:
23093         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
23094         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
23095         * Scripts/webkitpy/common/net/statusserver.py:
23096         * Scripts/webkitpy/layout_tests/port/base.py:
23097         * Scripts/webkitpy/tool/bot/botinfo.py: Added.
23098         * Scripts/webkitpy/tool/bot/botinfo_unittest.py: Added.
23099         * Scripts/webkitpy/tool/bot/commitqueuetask.py:
23100         * Scripts/webkitpy/tool/bot/flakytestreporter.py:
23101         * Scripts/webkitpy/tool/bot/flakytestreporter_unittest.py:
23102         * Scripts/webkitpy/tool/commands/queues.py:
23103
23104 2011-04-12  Daniel Bates  <dbates@rim.com>
23105
23106         Rubber-stamped by Eric Seidel.
23107
23108         svn-apply and/or patch(1) has trouble applying patches that makes changes to files
23109         with Windows line endings
23110         https://bugs.webkit.org/show_bug.cgi?id=53625
23111
23112         Fixes an issue where a patch made on a SCM checkout on a Unix file system cannot be
23113         applied to an SCM checkout on a Windows file system and vice-versa.
23114
23115         This issue manifests itself due to custom SCM settings as well as differences in
23116         gitattributes support in older versions of Git (e.g. 1.7.0.3). We implement support
23117         into svn-apply/unapply to compensate for differences in line endings by converting
23118         the line endings in a patch to match the line endings in the target file (if it exists).
23119
23120         * Scripts/VCSUtils.pm:
23121           - Added parseFirstEOL(), which returns the first end-of-line character(s) that appear in the
23122             content read from the specified file handle. Obviously, this heuristic will not work
23123             for files with mixed line endings, but I don't envision such files to be the norm. If
23124             this turns out to be an issue then we can improve the heuristic.
23125           - Added firstEOLInFile().
23126           - Modified parsePatch() and parseDiff() to take an optional hash reference to an options hash.
23127             In particular, added the hash key shouldNotUseIndexPathEOL to control whether to use the line
23128             endings in the diff instead of the line endings in the target file.
23129           - Extracted local variable $chunkRangeRegEx from fixChangeLogPatch() so that it can be
23130             re-used in parseDiff().
23131         * Scripts/webkitperl/VCSUtils_unittest/parseDiff.pl: Modified to pass {shouldNotUseIndexPathEOL => 1}
23132           to parsePatch() since these unit tests don't create mock files. Instead, created unit tests that use
23133           mock files in file VCSUtils_unittest/parseDiffWithMockFiles.pl.
23134         * Scripts/webkitperl/VCSUtils_unittest/parseDiffWithMockFiles.pl: Added.
23135         * Scripts/webkitperl/VCSUtils_unittest/parseFirstEOL.pl: Added.
23136
23137 2011-04-12  Sam Weinig  <sam@webkit.org>
23138
23139         Reviewed by Adam Roben.
23140
23141         REGRESSION (r83550): Multiple http tests crashing in the web process in SQLite beneath CFNetwork on Windows 7 Release (WebKit2 Tests)
23142         https://bugs.webkit.org/show_bug.cgi?id=58336
23143
23144         * WebKitTestRunner/TestController.cpp:
23145         (WTR::TestController::resetStateToConsistentValues):
23146         Remove cache clearing as that is causing tests to crash on Windows 7.
23147
23148 2011-04-11  Sam Weinig  <sam@webkit.org>
23149
23150         Reviewed by Anders Carlsson.
23151
23152         REGRESSION(r77974): http/tests/security/aboutBlank/security-context-window-open.html is failing
23153         <rdar://problem/8981346>
23154         https://bugs.webkit.org/show_bug.cgi?id=54159
23155
23156         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
23157         (WTR::InjectedBundlePage::InjectedBundlePage):
23158         (WTR::InjectedBundlePage::decidePolicyForNavigationAction):
23159         (WTR::InjectedBundlePage::decidePolicyForNewWindowAction):
23160         (WTR::InjectedBundlePage::decidePolicyForResponse):
23161         (WTR::InjectedBundlePage::unableToImplementPolicy):
23162         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
23163         Add short-circuited policy client which matches WebKit1.
23164
23165         * WebKitTestRunner/TestController.cpp:
23166         (WTR::TestController::resetStateToConsistentValues):
23167         Clear the cache between tests to ensure consistent results.
23168
23169 2011-04-11  George Guo  <George.Guo@nokia.com>
23170
23171         Reviewed by Eric Seidel.
23172
23173         prepare-ChangeLog support email input but -h did not show the option
23174         https://bugs.webkit.org/show_bug.cgi?id=58164
23175
23176         Add email option to -h to improve the usability
23177
23178         * Scripts/prepare-ChangeLog:
23179
23180 2011-04-11  Dimitri Glazkov  <dglazkov@chromium.org>
23181
23182         Reviewed by Eric Carlson.
23183
23184         Rename MediaControls to MediaControlRootElement.
23185         https://bugs.webkit.org/show_bug.cgi?id=58250
23186
23187         * Scripts/do-webcore-rename: Documented the renaming.
23188
23189 2011-04-11  Sam Weinig  <sam@webkit.org>
23190
23191         Fix Windows build.
23192
23193         * MiniBrowser/win/BrowserView.cpp:
23194         (BrowserView::create):
23195
23196 2011-04-11  Sam Weinig  <sam@webkit.org>
23197
23198         Reviewed by Anders Carlsson.
23199
23200         Move focus management out of WebKit via  the UIClient
23201         <rdar://problem/8784068>
23202         https://bugs.webkit.org/show_bug.cgi?id=58278
23203
23204         * MiniBrowser/mac/BrowserWindowController.m:
23205         (-[BrowserWindowController awakeFromNib]):
23206         * WebKitTestRunner/TestController.cpp:
23207         (WTR::TestController::createOtherPage):
23208         (WTR::TestController::initialize):
23209         Add stubs for new UIClient functions.
23210
23211 2011-02-03  Martin Robinson  <mrobinson@igalia.com>
23212
23213         Reviewed by Eric Seidel.
23214
23215         [GTK] The GTK+ DRT needs an implementation of the PlainTextController
23216         https://bugs.webkit.org/show_bug.cgi?id=53605
23217
23218         Add an implementation of the PlainTextController for the GTK+ DRT.
23219
23220         * DumpRenderTree/gtk/DumpRenderTree.cpp:
23221         (addControllerToWindow): Added this helper method.
23222         (webViewWindowObjectCleared): Use the new helper to add EventSender
23223         and the PlainTextController to the window object.
23224         * GNUmakefile.am: Added PlainTextController source files.
23225         * DumpRenderTree/gtk/PlainTextController.cpp: Added. This will rely
23226         on DumpRenderTreeSupportGtk to convert a JSValueRef into a WebKitDOMRange
23227         until that functionality exists somewhere in the WebKitGTK+ stack.
23228         * DumpRenderTree/gtk/PlainTextController.h: Added.
23229
23230 2011-04-11  Yael Aharon  <yael.aharon@nokia.com>
23231
23232         Unreviewed.
23233
23234         Adding my IRC nick.
23235
23236         * Scripts/webkitpy/common/config/committers.py:
23237
23238 2011-04-11  Dirk Pranke  <dpranke@chromium.org>
23239
23240         Reviewed by Adam Roben.
23241
23242         new-run-webkit-tests: fix crash when running under windows cmd.exe
23243         https://bugs.webkit.org/show_bug.cgi?id=58197
23244
23245         We would intermittently crash when running NRWT directly under
23246         cmd.exe on windows because we were sharing the stdin file
23247         descriptor between the python process and the http server.
23248         cmd.exe really didn't like that, and there was no reason to
23249         share the descriptor, so we now use a PIPE instead.
23250
23251         * Scripts/webkitpy/layout_tests/port/http_server.py:
23252
23253 2011-04-11  Dirk Pranke  <dpranke@chromium.org>
23254
23255         Reviewed by Ojan Vafai.
23256
23257         new-run-webkit-tests: add unit tests for Port.diff_image()
23258         https://bugs.webkit.org/show_bug.cgi?id=58196
23259
23260         Add the unit tests for the fix in bug 58195.
23261
23262         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
23263         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
23264         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
23265
23266 2011-04-11  Dirk Pranke  <dpranke@chromium.org>
23267
23268         Reviewed by Tony Chang.
23269
23270         new-run-webkit-tests: implement support for audio, take two
23271         https://bugs.webkit.org/show_bug.cgi?id=58195
23272
23273         Attempt to re-land the fix for bug 58101 (which was initially
23274         landed in r83330, but rolled out). This is the same patch but
23275         fixes crashes in Port.diff_image() caused by the change for
23276         empty image files being None instead of ''.
23277
23278         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
23279         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
23280         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
23281         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
23282         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
23283         * Scripts/webkitpy/layout_tests/port/base.py:
23284         * Scripts/webkitpy/layout_tests/port/chromium.py:
23285         * Scripts/webkitpy/layout_tests/port/dryrun.py:
23286         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
23287         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
23288         * Scripts/webkitpy/layout_tests/port/test.py:
23289         * Scripts/webkitpy/layout_tests/port/webkit.py:
23290
23291 2011-04-11  Mario Sanchez Prada  <msanchez@igalia.com>
23292
23293         Reviewed by Xan Lopez.
23294
23295         [GTK] Implement AccessibilityUIElement::{row|column}IndexRange in DRT
23296         https://bugs.webkit.org/show_bug.cgi?id=57854
23297
23298         Implement missing features in GTK's DRT.
23299
23300         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
23301         (indexRangeInTable): New helper function to get the range string
23302         for the current cell inside the parent table, either from the
23303         point of view of rows or columns.
23304         (AccessibilityUIElement::rowIndexRange): Implemented by relying on
23305         the new helper function indexRangeInTable().
23306         (AccessibilityUIElement::columnIndexRange): Ditto.
23307
23308 2011-04-11  Mario Sanchez Prada  <msanchez@igalia.com>
23309
23310         Reviewed by Chris Fleizach.
23311
23312         [GTK] Implement AccessibilityUIElement::cellForColumnAndRow in DRT
23313         https://bugs.webkit.org/show_bug.cgi?id=57826
23314
23315         Implement missing feature in GTK's DRT.
23316
23317         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
23318         (AccessibilityUIElement::cellForColumnAndRow): Implemented, by
23319         relying on the analogous function from the AtkTable interface.
23320
23321 2011-04-05  Timothy Hatcher  <timothy@apple.com>
23322
23323         Fix the extract-localizable-strings script errors that started after
23324         WEB_UI_STRING was added to LocalizedStrings.h in WebCore.
23325
23326         * Scripts/extract-localizable-strings: Skip LocalizedStrings.h.
23327
23328 2011-04-11  Tony Chang  <tony@chromium.org>
23329
23330         Reviewed by Eric Seidel.
23331
23332         fix crc computation on 64 bit machines
23333         https://bugs.webkit.org/show_bug.cgi?id=58243
23334
23335         unsigned long is dependenct on CPU architecture, so use unsigned which is always 32 bits
23336
23337         * DumpRenderTree/CyclicRedundancyCheck.cpp:
23338         (makeCrcTable):
23339         (computeCrc):
23340         * DumpRenderTree/CyclicRedundancyCheck.h:
23341         * DumpRenderTree/PixelDumpSupport.cpp:
23342         (appendIntToVector):
23343         (convertChecksumToPNGComment):
23344
23345 2011-04-10  Kimmo Kinnunen  <kimmo.t.kinnunen@nokia.com>
23346
23347         Reviewed by Eric Seidel.
23348
23349         Require no undefined symbols during compilation.
23350
23351         [Qt] [WK2] WebKitTestRunner, QtWebProcess and WTRInjectBundle should fail to compile when there's undefined symbols
23352         https://bugs.webkit.org/show_bug.cgi?id=54896
23353
23354         Add -Wl,--no-undefined to catch missing symbols early.
23355
23356         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
23357         * WebKitTestRunner/qt/WebKitTestRunner.pro:
23358
23359 2011-04-10  Kevin Ollivier  <kevino@theolliviers.com>
23360
23361         [wx] Unreviewed build fix. Remove collector dirs from the list of build dirs.
23362
23363         * wx/build/settings.py:
23364
23365 2011-04-10  Kevin Ollivier  <kevino@theolliviers.com>
23366
23367         [wx] Unreviewed build fix. Add Source/JavaScriptCore/heap to the list of directories.
23368
23369         * wx/build/settings.py:
23370
23371 2011-04-10  David Levin  <levin@chromium.org>
23372
23373         Reviewed by Dimitri Glazkov.
23374
23375         REGRESSION(r83384): Change to chromium.py is resulting in failures of the test framework on linux.
23376         https://bugs.webkit.org/show_bug.cgi?id=58201
23377
23378         * Scripts/webkitpy/layout_tests/port/chromium.py: Commented out the assert for now,
23379         and filed bug 58202.
23380
23381 2011-04-09  David Levin  <levin@chromium.org>
23382
23383         Unreviewed, rolling out r83394.
23384         http://trac.webkit.org/changeset/83394
23385         https://bugs.webkit.org/show_bug.cgi?id=53625
23386
23387         Patch was incorrect as noted in the bug.
23388
23389         * Scripts/svn-apply:
23390
23391 2011-04-09  David Levin  <levin@chromium.org>
23392
23393         Reviewed by Adam Barth.
23394
23395         svn-apply and/or patch(1) has trouble applying patches that makes changes to files with Windows line endings
23396         https://bugs.webkit.org/show_bug.cgi?id=53625
23397
23398         * Scripts/svn-apply: Ensure that the portions of patches
23399         which are for vcproj/vsprops files has DOS line endings.
23400
23401 2011-04-09  Dirk Pranke  <dpranke@chromium.org>
23402
23403         Reviewed by Tony Chang.
23404
23405         It looks like NRWT has not been stopping DRT/TestShell
23406         instances properly on windows, probably for a long time.
23407         This would go a long way to explaining why we often have
23408         processes lying around :)
23409
23410         https://bugs.webkit.org/show_bug.cgi?id=57807
23411
23412         * Scripts/webkitpy/layout_tests/port/chromium.py:
23413         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
23414
23415 2011-04-09  Keith Kyzivat  <keith.kyzivat@nokia.com>
23416
23417         Reviewed by Laszlo Gombos.
23418
23419         [Qt] Don't link against fontconfig or X11 if embedded
23420         https://bugs.webkit.org/show_bug.cgi?id=58104
23421
23422         * DumpRenderTree/qt/DumpRenderTree.pro: fontconfig !included in embedded
23423         * QtTestBrowser/QtTestBrowser.pro: ditto
23424         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro: ditto
23425         * WebKitTestRunner/qt/WebKitTestRunner.pro: ditto
23426
23427 2011-04-08  Jon Honeycutt  <jhoneycutt@apple.com>
23428
23429         http/tests/loading/preload-append-scan.php is failing on Windows
23430         https://bugs.webkit.org/show_bug.cgi?id=58178
23431
23432         Reviewed by Brian Weinstein.
23433
23434         * DumpRenderTree/win/DumpRenderTree.cpp:
23435         (lastPathComponent):
23436         Use CFURL API to find the last path component, rather than
23437         PathFindFileName(), because the latter will include the query string.
23438
23439 2011-04-08  Jian Li  <jianli@chromium.org>
23440
23441         Unreviewed, rolling out r83327, r83330 since these patches are very
23442         likely to break chromium webkit mac10.6 builders.
23443
23444         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
23445         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
23446         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
23447         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
23448         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
23449         * Scripts/webkitpy/layout_tests/port/base.py:
23450         * Scripts/webkitpy/layout_tests/port/chromium.py:
23451         * Scripts/webkitpy/layout_tests/port/dryrun.py:
23452         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
23453         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
23454         * Scripts/webkitpy/layout_tests/port/test.py:
23455         * Scripts/webkitpy/layout_tests/port/webkit.py:
23456
23457 2011-04-08  Vincent Scheib  <scheib@chromium.org>
23458
23459         Reviewed by Eric Seidel.
23460
23461         add chromium-gpu-linux-x86_64 to the recognized list of ports
23462         https://bugs.webkit.org/show_bug.cgi?id=58099
23463
23464         Add chromium-gpu-linux-x86_64 to fix assert.
23465
23466         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
23467
23468 2011-04-08  Alpha Lam  <hclam@chromium.org>
23469
23470         Unreviewed, rolling out r83335.
23471         http://trac.webkit.org/changeset/83335
23472         https://bugs.webkit.org/show_bug.cgi?id=53556
23473
23474         GTK and QT bots are broken
23475
23476         * Scripts/build-webkit:
23477
23478 2011-04-07  Anna Cavender  <annacc@chromium.org>
23479
23480         Reviewed by Eric Carlson.
23481
23482         Setup ENABLE(TRACK) feature define
23483         https://bugs.webkit.org/show_bug.cgi?id=53556
23484
23485         * Scripts/build-webkit:
23486
23487 2011-04-08  Dirk Pranke  <dpranke@chromium.org>
23488
23489         Reviewed by Adam Roben.
23490
23491         new-run-webkit-tests: configure a NRWT test bot for Mac SL
23492         https://bugs.webkit.org/show_bug.cgi?id=58114
23493
23494         This change removes the old new-run-webkit-tests entry and
23495         updates it for a new test-only bot.
23496
23497         * BuildSlaveSupport/build.webkit.org-config/config.json:
23498
23499 2011-04-08  Dominic Cooney  <dominicc@google.com>
23500
23501         Reviewed by Adam Roben.
23502
23503         Make layoutTestController.shadowRoot return null, not undefined,
23504         when its argument is invalid.
23505         https://bugs.webkit.org/show_bug.cgi?id=58121
23506
23507         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
23508         (LayoutTestController::shadowRoot):
23509         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
23510         (WTR::LayoutTestController::shadowRoot):
23511
23512 2011-04-08  Pere Martir  <pere.martir4@gmail.com>
23513
23514         Reviewed by Adam Roben.
23515
23516         Locate NSTD.EXE in 64-bit Windows
23517         https://bugs.webkit.org/show_bug.cgi?id=57847
23518
23519         * Scripts/old-run-webkit-tests:
23520
23521 2011-04-08  Adam Roben  <aroben@apple.com>
23522
23523         Qt build fix
23524
23525         * DumpRenderTree/TestNetscapePlugIn/PluginObject.h: Pull in stdarg.h for va_list.
23526
23527 2011-04-07  Adam Roben  <aroben@apple.com>
23528
23529         Test that NPP_SetWindow is passed a null window handle during plugin destruction on non-Mac platforms
23530
23531         Test for <http://webkit.org/b/47009> WebKit2 needs to call NPP_SetWindow when destroying a
23532         plugin
23533
23534         Reviewed by Anders Carlsson.
23535
23536         * DumpRenderTree/TestNetscapePlugIn/PluginObject.cpp:
23537         (pluginLogWithArguments): Moved code to format and log the message here...
23538         (pluginLog): ...from here.
23539
23540         * DumpRenderTree/TestNetscapePlugIn/PluginObject.h: Added pluginLogWithArguments.
23541
23542         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
23543         (PluginTest::log): Added. Calls through to pluginLogWithArguments.
23544
23545         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h: Added log.
23546
23547         * DumpRenderTree/TestNetscapePlugIn/Tests/NPPSetWindowCalledDuringDestruction.cpp: Added.
23548         (NPPSetWindowCalledDuringDestruction::setWillBeDestroyed): Records that destruction is about
23549         to begin.
23550         (NPPSetWindowCalledDuringDestruction::NPPSetWindowCalledDuringDestruction): Simple
23551         constructor.
23552         (NPPSetWindowCalledDuringDestruction::NPP_GetValue): Creates and returns a ScriptObject that
23553         can be used to invoke our setWillBeDestroyed function.
23554         (NPPSetWindowCalledDuringDestruction::NPP_SetWindow): Records what has happened (and logs if
23555         anything unexpected happens).
23556         (NPPSetWindowCalledDuringDestruction::NPP_Destroy): On Mac, logs a failure message if
23557         NPP_SetWindow was called during destruction. On other platforms, logs a failure message if
23558         NPP_SetWindow was *not* called during destruction.
23559         (NPPSetWindowCalledDuringDestruction::ScriptObject::hasMethod): Return true for our only
23560         method, setWillBeDestroyed.
23561         (NPPSetWindowCalledDuringDestruction::ScriptObject::invoke): Call through to the PluginTest
23562         object.
23563
23564         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
23565         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
23566         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
23567         * GNUmakefile.am:
23568         Added new test.
23569
23570         * Scripts/old-run-webkit-tests: Skip the new test when using out-of-process plugins with
23571         WebKit1 on Mac, since it can't work properly due to <http://webkit.org/b/58077>.
23572
23573 2011-04-08  Mario Sanchez Prada  <msanchez@igalia.com>
23574
23575         Reviewed by Martin Robinson.
23576
23577         [GTK] Implement increment() and decrement() functions in DRT's AccessibilityUIElement
23578         https://bugs.webkit.org/show_bug.cgi?id=58039
23579
23580         Implement missing functions in GTK's DRT.
23581
23582         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
23583         (AccessibilityUIElement::increment): Implemented.
23584         (AccessibilityUIElement::decrement): Implemented.
23585
23586 2011-04-08  Dominic Cooney  <dominicc@google.com>
23587
23588         Reviewed by Kent Tamura.
23589
23590         Make WK2 layoutTestController.shadowRoot return undefined, not
23591         null, when its argument is not an element.
23592         https://bugs.webkit.org/show_bug.cgi?id=58121
23593
23594         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
23595         (WTR::LayoutTestController::shadowRoot):
23596
23597 2011-04-07  Dirk Pranke  <dpranke@chromium.org>
23598
23599         Reviewed by Mihai Parparita.
23600
23601         Change reference port for Mac GPU baselines from
23602         Leopard to SnowLeopard, XP to Win7, and Linux-x86
23603         to Linux x86-64.
23604
23605         https://bugs.webkit.org/show_bug.cgi?id=58099
23606
23607         * Scripts/webkitpy/layout_tests/port/chromium.py:
23608
23609 2011-04-07  Andrew Scherkus  <scherkus@chromium.org>
23610
23611         Revert ENABLE_TRACK patch due to compile failures.
23612
23613         * Scripts/build-webkit:
23614
23615 2011-04-07  Tony Chang  <tony@chromium.org>
23616
23617         Reviewed by Eric Seidel.
23618
23619         update DRT to embed checksums in png files
23620         https://bugs.webkit.org/show_bug.cgi?id=57871
23621
23622         We insert the bytes for the comment in printPNG rather than at encode
23623         time because each platform does its own PNG encoding (either using CG
23624         or cairo).  Putting this in pringPNG avoids having to duplicate this
23625         code, although it's not as clean as doing it at encoding time.
23626
23627         We insert the comment right after the IHDR chunk of the PNG.
23628
23629         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
23630         * DumpRenderTree/CyclicRedundancyCheck.cpp: Added.
23631         (makeCrcTable): Ported from LayoutTests/fast/canvas/webgl/resources/pnglib.js
23632         (computeCrc): Ported from LayoutTests/fast/canvas/webgl/resources/pnglib.js
23633         * DumpRenderTree/CyclicRedundancyCheck.h: Added.
23634         * DumpRenderTree/PixelDumpSupport.cpp:
23635         (dumpWebViewAsPixelsAndCompareWithExpected):
23636         (convertChecksumToPNGComment): Generate the bytes to insert.
23637         (printPNG): Insert the png comment before the first IDAT section.
23638         * DumpRenderTree/PixelDumpSupport.h:
23639         * DumpRenderTree/cairo/PixelDumpSupportCairo.cpp:
23640         (printPNG):
23641         (dumpBitmap):
23642         * DumpRenderTree/cg/PixelDumpSupportCG.cpp:
23643         (printPNG):
23644         (dumpBitmap):
23645         * DumpRenderTree/win/DumpRenderTree.vcproj:
23646         * GNUmakefile.am:
23647
23648 2011-04-07  Jeff Miller  <jeffm@apple.com>
23649
23650         Reviewed by Adam Roben.
23651
23652         Replace WKStringGetCharactersPtr() with WKStringGetCharacters()
23653         https://bugs.webkit.org/show_bug.cgi?id=58058
23654
23655         * TestWebKitAPI/Tests/WebKit2/WKString.cpp:
23656         (TestWebKitAPI::TEST): Add tests for  WKStringGetLength() and WKStringGetCharactersPtr().
23657
23658 2011-04-07  Chang Shu  <cshu@webkit.org>
23659
23660         Reviewed by Darin Adler.
23661
23662         WebKitTestRunner needs layoutTestController.isPageBoxVisible
23663         https://bugs.webkit.org/show_bug.cgi?id=42695
23664
23665         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
23666         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
23667         (WTR::LayoutTestController::isPageBoxVisible):
23668         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
23669
23670 2011-04-07  Adam Roben  <aroben@apple.com>
23671
23672         Build fix
23673
23674         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Added a missing import.
23675
23676 2011-04-07  Adam Roben  <aroben@apple.com>
23677
23678         Move ExtractTestResults[AndLeaks]'s summarizing code back from commandComplete() to finished()
23679
23680         commandComplete() is never called for MasterShellCommands like ExtractTestResults[AndLeaks].
23681         (Unfortunately the buildbot documentation does not make this clear.) finished() is the only
23682         hook we have, so we have to do our work there. I added a new addCustomURLs method which can
23683         be overridden by subclasses to provide extra URLs before we call up to the base class (after
23684         which adding more URLs is no longer possible).
23685
23686         Really hopefully fixes <http://webkit.org/b/56032> Leaks viewer should be linked from leaks
23687         bot results page
23688
23689         Reviewed by John Sullivan.
23690
23691         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
23692         (ExtractTestResults.addCustomURLs): Moved code to add the "view results" URL here...
23693         (ExtractTestResults.finished): ...from here. This replaces commandComplete, since that
23694         method is never called for MasterShellCommands.
23695         (ExtractTestResultsAndLeaks.addCustomURLs): Replaced commandComplete (which is never called)
23696         with this method (which is).
23697
23698 2011-04-07  Adam Roben  <aroben@apple.com>
23699
23700         Move ExtractTestResults[AndLeaks]'s summarizing code from finished() to commandComplete()
23701
23702         This matches how most of our other build steps work, so is good just for improving
23703         consistency between build steps. It should also make it possible for
23704         ExtractTestResultsAndLeaks to successfully add a URL to Leaks Viewer. (Previously we were
23705         trying to do this in finished() after we had called up to the base class, but that was
23706         apparently too late to add more URLs.)
23707
23708         Hopefully fixes <http://webkit.org/b/56032> Leaks viewer should be linked from leaks bot
23709         results page
23710
23711         Reviewed by John Sullivan.
23712
23713         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
23714         (ExtractTestResults): Added a descriptionDone property so that we don't have to manually
23715         call setText to get the right text to show up for this step. This matches how most of our
23716         other build steps work.
23717         (ExtractTestResults.commandComplete): Replaced our override of finished with this function.
23718         This is how most of our other build steps work.
23719         (ExtractTestResultsAndLeaks.commandComplete): Replaced our override of finished with this
23720         function, to match the base class.
23721
23722 2011-04-07  Adam Barth  <abarth@webkit.org>
23723
23724         Reviewed by Maciej Stachowiak.
23725
23726         webkit-patch shouldn't state the whole working copy when calling check-webkit-style
23727         https://bugs.webkit.org/show_bug.cgi?id=58022
23728
23729         Previously, we were ignoring the args variable!  This patch fixes the
23730         regression introduced in http://trac.webkit.org/changeset/82771.
23731
23732         * Scripts/webkitpy/tool/commands/download_unittest.py:
23733         * Scripts/webkitpy/tool/mocktool.py:
23734         * Scripts/webkitpy/tool/steps/checkstyle.py:
23735
23736 2011-04-07  Maciej Stachowiak  <mjs@apple.com>
23737
23738         Reviewed by Adam Barth.
23739
23740         REGRESSION: "webkit-patch land" doesn't work correctly in SVN subdirectories
23741         https://bugs.webkit.org/show_bug.cgi?id=58017
23742
23743         * Scripts/webkitpy/common/checkout/scm.py:
23744         * Scripts/webkitpy/common/checkout/scm_unittest.py:
23745
23746 2011-04-06  Chang Shu  <cshu@webkit.org>
23747
23748         Reviewed by Darin Adler.
23749
23750         WebKitTestRunner needs layoutTestController.pageSizeAndMarginsInPixels
23751         https://bugs.webkit.org/show_bug.cgi?id=57984
23752
23753         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
23754         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
23755         (WTR::LayoutTestController::pageSizeAndMarginsInPixels):
23756         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
23757
23758 2011-04-06  Chris Rogers  <crogers@google.com>
23759
23760         Reviewed by Tony Chang.
23761
23762         Add web audio support to DumpRenderTree (mac port)
23763         https://bugs.webkit.org/show_bug.cgi?id=57969
23764
23765         * DumpRenderTree/LayoutTestController.cpp:
23766         (LayoutTestController::LayoutTestController):
23767         (setEncodedAudioDataCallback):
23768         (LayoutTestController::staticFunctions):
23769         * DumpRenderTree/LayoutTestController.h:
23770         (LayoutTestController::dumpAsAudio):
23771         (LayoutTestController::setDumpAsAudio):
23772         (LayoutTestController::encodedAudioData):
23773         (LayoutTestController::setEncodedAudioData):
23774         * DumpRenderTree/mac/DumpRenderTree.mm:
23775         (dumpAudio):
23776         (dump):
23777
23778 2011-04-06  Benjamin Poulain  <benjamin.poulain@nokia.com>
23779
23780         Reviewed by Kenneth Rohde Christiansen.
23781
23782         [Qt] LayoutTestController needs to implement numberOfPendingGeolocationPermissionRequests
23783         https://bugs.webkit.org/show_bug.cgi?id=56086
23784
23785         Add the missing method.
23786
23787         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
23788         (WebCore::DumpRenderTree::getAllPages): Add a method to return all the pages allocated for the current test.
23789         * DumpRenderTree/qt/DumpRenderTreeQt.h:
23790         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
23791         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
23792
23793 2011-04-06  Dai Mikurube  <dmikurube@chromium.org>
23794
23795         Reviewed by David Levin.
23796
23797         Add QUOTA build flag for unified quota API
23798         https://bugs.webkit.org/show_bug.cgi?id=57918
23799
23800         * Scripts/build-webkit: Added QUOTA build flag
23801
23802 2011-04-06  Dirk Pranke  <dpranke@chromium.org>
23803
23804         Reviewed by Tony Chang.
23805
23806         Adjust the apple webkit port's default timeout to match
23807         old-run-webkit-tests at 35 seconds.
23808
23809         https://bugs.webkit.org/show_bug.cgi?id=37738
23810
23811         * Scripts/webkitpy/layout_tests/port/mac.py:
23812
23813 2011-04-06  Scott Cameron  <sccameron@rim.com>
23814
23815         Reviewed by Eric Seidel.
23816
23817         Remove global variable $httpdPath and replace with the return value of
23818         getHTTPDPath().  Also remove unnecessary calls to getHTTPDPath().
23819         https://bugs.webkit.org/show_bug.cgi?id=53499
23820
23821         * Scripts/run-iexploder-tests:
23822         * Scripts/webkitperl/httpd.pm:
23823
23824 2011-04-06  Kevin Ollivier  <kevino@theolliviers.com>
23825
23826         Reviewed by Darin Adler.
23827
23828         Make sure JS_EXPORT_PRIVATE is an empty define when we aren't using the export macros.
23829
23830         https://bugs.webkit.org/show_bug.cgi?id=27551
23831
23832         * DumpRenderTree/config.h:
23833         * WebKitAPITest/config.h:
23834         * WebKitTestRunner/config.h:
23835
23836 2011-04-06  Tony Chang  <tony@chromium.org>
23837
23838         Reviewed by Darin Adler.
23839
23840         teach run-webkit-tests to read checksums from png files
23841         https://bugs.webkit.org/show_bug.cgi?id=57993
23842
23843         * Scripts/old-run-webkit-tests: Read the first 2k of a .png if there's
23844           no .checksum and look for the checksum in there.
23845
23846 2011-04-06  David Dorwin  <ddorwin@chromium.org>
23847
23848         Reviewed by David Levin.
23849
23850         Enable fullscreen layout tests for Chromium
23851         https://bugs.webkit.org/show_bug.cgi?id=55726
23852
23853         Make DumpRenderTree always run with fullscreen enabled (equivalent of --enable-fullscreen).
23854
23855         * DumpRenderTree/chromium/WebPreferences.cpp:
23856         (WebPreferences::applyTo):
23857
23858 2011-04-06  Adam Roben  <aroben@apple.com>
23859
23860         Add a "view leaks" link to builds on SnowLeopard Intel Leaks
23861
23862         Fixes <http://webkit.org/b/56032> Leaks viewer should be linked from leaks bot results page
23863
23864         Reviewed by David Kilzer.
23865
23866         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
23867         (ExtractTestResults.resultDirectoryURL): Added. Moved code to calculate the URL for the
23868         build's results directory here...
23869         (ExtractTestResults.finished): ...from here.
23870         (ExtractTestResultsAndLeaks): New class that's used by the leaks builder
23871         (ExtractTestResultsAndLeaks.finished): Calls up to the base class, but also adds a "view
23872         leaks" link to point to Leaks Viewer for this build.
23873         (TestFactory): Added ExtractTestResultsClass abstraction. This isn't overridden anywhere,
23874         but it seemed good to add for consistency with BuildAndTestFactory.
23875         (BuildAndTestFactory): Added ExtractTestResultsClass abstraction.
23876         (BuildAndTestLeaksFactory): Use ExtractTestResultsAndLeaks as our ExtractTestResultsClass so
23877         that we'll get a "view leaks" link.
23878
23879 2011-04-06  Zan Dobersek  <zandobersek@gmail.com>
23880
23881         Reviewed by Eric Seidel.
23882
23883         [Gtk] plugins/set-status.html fails
23884         https://bugs.webkit.org/show_bug.cgi?id=57844
23885
23886         Allow an empty status text to be dumped.
23887
23888         * DumpRenderTree/gtk/DumpRenderTree.cpp:
23889         (webViewStatusBarTextChanged):
23890
23891 2011-04-06  Chang Shu  <cshu@webkit.org>
23892
23893         Reviewed by Darin Adler.
23894
23895         WebKitTestRunner needs layoutTestController.pageNumberForElementById
23896         https://bugs.webkit.org/show_bug.cgi?id=42329
23897
23898         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
23899         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
23900         (WTR::LayoutTestController::pageNumberForElementById):
23901         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
23902
23903 2011-04-06  Sergio Villar Senin  <svillar@igalia.com>
23904
23905         Reviewed by Martin Robinson.
23906
23907         [GTK] DumpRenderTree: do not try to free NULL SoupURIs
23908         https://bugs.webkit.org/show_bug.cgi?id=57932
23909
23910         Some Layout tests have invalid URIs that do not generate valid
23911         SoupURI instances. Do not try to free those NULL SoupURIs.
23912
23913         * DumpRenderTree/gtk/DumpRenderTree.cpp:
23914         (willSendRequestCallback):
23915
23916 2011-04-06  Csaba Osztrogonác  <ossy@webkit.org>
23917
23918         Reviewed by Andreas Kling.
23919
23920         ORWT shouldn't generate diff files for tests without expected files
23921         https://bugs.webkit.org/show_bug.cgi?id=57846
23922
23923         * Scripts/old-run-webkit-tests:
23924
23925 2011-04-05  Kevin Ollivier  <kevino@theolliviers.com>
23926
23927         [wx] Mac build fix. Handle cases where 10.4 SDK is not installed, and also add x86_64 arch
23928         to deps.
23929
23930         * wx/install-unix-extras:
23931
23932 2011-04-05  Sheriff Bot  <webkit.review.bot@gmail.com>
23933
23934         Unreviewed, rolling out r82978, r82999, and r83001.
23935         http://trac.webkit.org/changeset/82978
23936         http://trac.webkit.org/changeset/82999
23937         http://trac.webkit.org/changeset/83001
23938         https://bugs.webkit.org/show_bug.cgi?id=57913
23939
23940         Does not work in Python 2.5 (Requested by abarth on #webkit).
23941
23942         * Scripts/webkitpy/common/system/executive.py:
23943         * Scripts/webkitpy/common/system/executive_unittest.py:
23944
23945 2011-04-05  Chang Shu  <cshu@webkit.org>
23946
23947         Reviewed by Darin Adler.
23948
23949         WebKitTestRunner needs layoutTestController.numberOfPages
23950         https://bugs.webkit.org/show_bug.cgi?id=42694
23951
23952         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
23953         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
23954         (WTR::LayoutTestController::numberOfPages):
23955         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
23956
23957 2011-04-05  Tony Chang  <tony@chromium.org>
23958
23959         Reviewed by Mihai Parparita.
23960
23961         [chromium] stop putting results downloaded from WebKit Linux in chromium-linux-x86_64
23962         https://bugs.webkit.org/show_bug.cgi?id=57889
23963
23964         * Scripts/webkitpy/layout_tests/port/chromium.py:
23965         * Scripts/webkitpy/layout_tests/port/test.py:
23966         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
23967
23968 2011-04-05  Adam Barth  <abarth@webkit.org>
23969
23970         Silly with statement, from the future!
23971
23972         * Scripts/webkitpy/common/system/executive.py:
23973
23974 2011-04-05  Adam Barth  <abarth@webkit.org>
23975
23976         Reviewed by Tony Chang.
23977
23978         Don't use Exception.message because it's deprecated
23979         https://bugs.webkit.org/show_bug.cgi?id=57892
23980
23981         Suppress the warning for now.  When we move to Python 3, we might need
23982         to something more dramatic.
23983
23984         * Scripts/webkitpy/common/system/executive.py:
23985
23986 2011-04-05  Mihai Parparita  <mihaip@chromium.org>
23987
23988         Reviewed by Ojan Vafai.
23989
23990         Add builders.js to dashboard file list
23991         https://bugs.webkit.org/show_bug.cgi?id=57899
23992
23993         Add file added by http://crrev.com/80538 to dashboard file list.
23994
23995         * TestResultServer/handlers/dashboardhandler.py:
23996
23997 2011-04-05  MORITA Hajime  <morrita@google.com>
23998
23999         Reviewed by Adam Barth.
24000
24001         webkit-patch should print git's stderr when git svn dcommit fail
24002         http://webkit.org/b/57861
24003
24004         * Scripts/webkitpy/common/checkout/scm.py:
24005
24006 2011-04-05  Mihai Parparita  <mihaip@chromium.org>
24007
24008         Reviewed by Tony Chang.
24009
24010         Add --baseline-search-path to NRWT
24011         https://bugs.webkit.org/show_bug.cgi?id=56233
24012
24013         Add NRWT option to specify additional directories to look for baselines
24014         (will be used by hardware GPU bots which will have local per-bot
24015         expectations for some tests)
24016
24017         * Scripts/webkitpy/common/system/filesystem_mock.py:
24018         * Scripts/webkitpy/layout_tests/port/base.py:
24019         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
24020         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
24021
24022 2011-04-05  Adam Roben  <aroben@apple.com>
24023
24024         Strip off /results.html from results URLs before trying to load leaks files from them
24025
24026         r82734 changed build.webkit.org's "view results" URLs to point straight to the results.html
24027         files, rather than pointing to the directory that contains them. This is more convenient for
24028         people browsing build.webkit.org, but confused Leaks Viewer.
24029
24030         Fixes <http://webkit.org/b/57869> REGRESSION (r82734): Links in Leaks Viewer's Recent Builds
24031         list don't work
24032
24033         Reviewed by Joseph Pecoraro.
24034
24035         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/RecentBuildsLoader.js:
24036         (RecentBuildsLoader.prototype.start): Strip off "/results.html" from the results URL.
24037
24038 2011-04-05  Carol Szabo  <carol@webkit.org>
24039
24040         Unreviewed.
24041
24042         Updated my info in committers.py
24043
24044         Scripts\webkitpy\common\config\committers.py
24045
24046 2011-04-05  Csaba Osztrogonác  <ossy@webkit.org>
24047
24048         Reviewed by Laszlo Gombos.
24049
24050         [Qt] Make WebKitLibraries optional for building QtWebKit
24051         https://bugs.webkit.org/show_bug.cgi?id=57542
24052
24053         * Scripts/build-webkit:
24054
24055 2011-04-05  Chang Shu  <cshu@webkit.org>
24056
24057         Reviewed by Laszlo Gombos.
24058
24059         [Qt] MiniBrowser defaultUrl does not work
24060         https://bugs.webkit.org/show_bug.cgi?id=57021
24061
24062         Match the behavior and coding of MiniBrowser to QtTestBrowser.
24063         * MiniBrowser/qt/main.cpp:
24064         (main):
24065
24066 2011-04-05  Jade Han  <jade.han@nokia.com>
24067
24068         Reviewed by Laszlo Gombos.
24069
24070         [Qt] [Symbian] Disable WebKitTestRunner for Symbian
24071         https://bugs.webkit.org/show_bug.cgi?id=54977
24072
24073         This change is a preparation to enable building webkit2 for Symbian.
24074         Bug 57834 is filed to fix and enable WebKitTestRunner for Symbian.
24075
24076         * Tools.pro:
24077
24078 2011-04-05  Zoltan Horvath  <zoltan@webkit.org>
24079
24080         Reviewed by Andreas Kling.
24081
24082         [Qt] Fix timeoutTimer of MiniBrowser's UrlLoader
24083         https://bugs.webkit.org/show_bug.cgi?id=57832
24084
24085         Only QWKPage has loadFinished signal so connect to it instead of BrowserWindow.
24086
24087         * MiniBrowser/qt/UrlLoader.cpp:
24088         (UrlLoader::UrlLoader):
24089
24090 2011-04-05  Zoltan Horvath  <zoltan@webkit.org>
24091
24092         [Qt] Linux Release minimal build fix after r82919.
24093
24094         * MiniBrowser/qt/BrowserWindow.cpp:
24095         (BrowserWindow::loadURLListFromFile):
24096         * QtTestBrowser/launcherwindow.cpp:
24097         (LauncherWindow::loadURLListFromFile):
24098
24099 2011-04-05  Zoltan Horvath  <zoltan@webkit.org>
24100
24101         Reviewed by Andreas Kling.
24102
24103         [Qt] Make URL loader accessible from testbrowser's menus
24104         https://bugs.webkit.org/show_bug.cgi?id=57823
24105
24106         Add "Load URLs from file" to QtTestBrowser's and to MiniBrowser's menu.
24107
24108         * MiniBrowser/qt/BrowserWindow.cpp:
24109         (BrowserWindow::BrowserWindow):
24110         (BrowserWindow::loadURLListFromFile):
24111         (BrowserWindow::~BrowserWindow):
24112         * MiniBrowser/qt/BrowserWindow.h:
24113         * QtTestBrowser/launcherwindow.cpp:
24114         (LauncherWindow::LauncherWindow):
24115         (LauncherWindow::~LauncherWindow):
24116         (LauncherWindow::createChrome):
24117         (LauncherWindow::loadURLListFromFile):
24118         * QtTestBrowser/launcherwindow.h:
24119
24120 2011-04-04  Sam Weinig  <sam@webkit.org>
24121
24122         Reviewed by Brian Weinstein.
24123
24124         Remove duplicate API from WKContext
24125         <rdar://problem/8727879>
24126         https://bugs.webkit.org/show_bug.cgi?id=57815
24127
24128         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
24129         * TestWebKitAPI/Tests/WebKit2/SendingMessagesToTheWebProcessBeforeItIsValid.cpp: Removed.
24130         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
24131         Remove SendingMessagesToTheWebProcessBeforeItIsValid since the API it was testing is now removed.
24132
24133 2011-04-04  Tony Chang  <tony@chromium.org>
24134
24135         Reviewed by Ojan Vafai.
24136
24137         [chromium] don't write .checksum files if a fallback platform has an embedded checksum
24138         https://bugs.webkit.org/show_bug.cgi?id=57783
24139
24140         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
24141         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
24142
24143 2011-04-04  Kevin Ollivier  <kevino@theolliviers.com>
24144
24145         [wx] Unreviewed build fix, run uninstall when cleaning to remove built files
24146         from WebKitBuild.
24147
24148         * Scripts/webkitdirs.pm:
24149
24150 2011-04-04  Martin Robinson  <mrobinson@igalia.com>
24151
24152         Reviewed by Gustavo Noronha Silva.
24153
24154         [GTK] WebGL support
24155         https://bugs.webkit.org/show_bug.cgi?id=31517
24156
24157         Add support to the DRT for turning on WebGL when a layout tests requests it.
24158
24159         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
24160         (LayoutTestController::overridePreference): Allow turning on WebGL from tests.
24161
24162 2011-04-04  Chang Shu  <cshu@webkit.org>
24163
24164         Reviewed by Darin Adler.
24165
24166         WebKitTestRunner needs layoutTestController.setAllowFileAccessFromFileURLs
24167         https://bugs.webkit.org/show_bug.cgi?id=57572
24168
24169         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
24170         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
24171         (WTR::LayoutTestController::setAllowFileAccessFromFileURLs):
24172         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
24173
24174 2011-04-04  Keith Kyzivat  <keith.kyzivat@nokia.com>
24175
24176         Reviewed by Csaba Osztrogonác.
24177
24178         [Qt] DumpRenderTree breaks compilation in some uClibc environments
24179         https://bugs.webkit.org/show_bug.cgi?id=57602
24180
24181         * DumpRenderTree/qt/main.cpp:
24182         (get_backtrace):
24183
24184 2011-04-04  Kevin Ollivier  <kevino@theolliviers.com>
24185
24186         [wx] Unreviewed build fix, add new LayoutTestController method stub to wx.
24187
24188         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
24189         (LayoutTestController::shadowRoot):
24190
24191 2011-04-04  Pavel Podivilov  <podivilov@chromium.org>
24192
24193         Unreviewed, fix exception in rebaseline tool.
24194
24195         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
24196
24197 2011-04-02  Beth Dakin  <bdakin@apple.com>
24198
24199         Rubber-stamped by Geoff Garen.
24200
24201         For Dan!
24202         * DumpRenderTree/mac/DumpRenderTree.mm:
24203         (resetWebViewToConsistentStateBeforeTesting):
24204
24205 2011-04-02  Beth Dakin  <bdakin@apple.com>
24206
24207         Rubber-stamped by Geoff Garen.
24208
24209         Need to reset the scale, much like zoom.
24210         * DumpRenderTree/mac/DumpRenderTree.mm:
24211         (resetWebViewToConsistentStateBeforeTesting):
24212
24213 2011-04-02  Sam Weinig  <sam@webkit.org>
24214
24215         Reviewed by Beth Dakin.
24216
24217         https://bugs.webkit.org/show_bug.cgi?id=57605
24218         Frame::pageScaleFactor() should not affect getBoundingClientRect() or
24219         getClientRects()
24220         -and corresponding-
24221         <rdar://problem/9194541>
24222
24223         Add DRT support for the scaleWebView SPI.
24224         * DumpRenderTree/mac/EventSendingController.mm:
24225         (+[EventSendingController isSelectorExcludedFromWebScript:]):
24226         (+[EventSendingController webScriptNameForSelector:]):
24227         (-[EventSendingController scalePageBy:atX:andY:]):
24228         * WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
24229         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
24230         (WTR::EventSendingController::scalePageBy):
24231         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
24232         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
24233         (WTR::InjectedBundlePage::reset):
24234
24235 2011-04-02  Dominic Cooney  <dominicc@google.com>
24236
24237         Reviewed by Martin Robinson.
24238
24239         Add layoutTestController.shadowRoot to GTK DumpRenderTree.
24240         https://bugs.webkit.org/show_bug.cgi?id=57551
24241
24242         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
24243         (LayoutTestController::shadowRoot):
24244
24245 2011-04-02  Patrick Gansterer  <paroga@webkit.org>
24246
24247         Reviewed by Eric Seidel.
24248
24249         Remove AbstractStep._run_script and move script names to ports.py
24250         https://bugs.webkit.org/show_bug.cgi?id=57704
24251
24252         Replace deprecated _run_script with _tool.executive.run_and_throw_if_fail.
24253
24254         * Scripts/webkitpy/common/config/ports.py:
24255         * Scripts/webkitpy/tool/commands/download_unittest.py:
24256         * Scripts/webkitpy/tool/commands/roll_unittest.py:
24257         * Scripts/webkitpy/tool/commands/upload_unittest.py:
24258         * Scripts/webkitpy/tool/steps/abstractstep.py:
24259         * Scripts/webkitpy/tool/steps/checkstyle.py:
24260         * Scripts/webkitpy/tool/steps/preparechangelog.py:
24261         * Scripts/webkitpy/tool/steps/preparechangelogfordepsroll.py:
24262         * Scripts/webkitpy/tool/steps/preparechangelogforrevert.py:
24263
24264 2011-04-02  Patrick Gansterer  <paroga@webkit.org>
24265
24266         Reviewed by Eric Seidel.
24267
24268         Emulate shebang on Win32
24269         https://bugs.webkit.org/show_bug.cgi?id=55927
24270
24271         Scripts on Windows work only if they are called with the explicit interpreter.
24272         Read the first line of scripts to detect the correct executable.
24273
24274         * Scripts/webkitpy/common/config/ports.py:
24275         * Scripts/webkitpy/common/system/executive.py: Added interpreter_for_script().
24276         * Scripts/webkitpy/common/system/executive_unittest.py:
24277
24278 2011-04-01  Adam Barth  <abarth@webkit.org>
24279
24280         Reviewed by Eric Seidel.
24281
24282         apos entities shouldn't show up in ChangeLogs when using webkit-patch
24283         https://bugs.webkit.org/show_bug.cgi?id=57692
24284
24285         Previously, we were using BeautifulSoup to process XML from
24286         bugs.webkit.org, but that's incorrect.  We should be using
24287         BeautifulStoneSoup to process the XML.  We were getting the &apos;
24288         entity wrong because &apos; is an XML entity but not an HTML entity.
24289
24290         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
24291         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
24292
24293 2011-04-01  Sheriff Bot  <webkit.review.bot@gmail.com>
24294
24295         Unreviewed, rolling out r82721.
24296         http://trac.webkit.org/changeset/82721
24297         https://bugs.webkit.org/show_bug.cgi?id=57687
24298
24299         This patch introduced assertion failures on the GTK+ bots.
24300         (Requested by mrobinson on #webkit).
24301
24302         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
24303         (LayoutTestController::shadowRoot):
24304
24305 2011-04-01  Csaba Osztrogonác  <ossy@webkit.org>
24306
24307         Unreviewed buildfix.
24308
24309         [Qt][WK2] Build Webkit2 using "-2" option on Qt
24310         https://bugs.webkit.org/show_bug.cgi?id=55074
24311
24312         * Scripts/build-webkit: Ensure that "-2" isn't passed to qmake.
24313         (The isWK2() function removes it from @ARGV, but not from @options.)
24314
24315 2011-04-01  Keith Kyzivat  <keith.kyzivat@nokia.com>
24316
24317         Reviewed by Csaba Osztrogonác.
24318
24319         [Qt] [WK2] MiniBrowser.qrc not found - regression from rev 82671
24320         https://bugs.webkit.org/show_bug.cgi?id=57666
24321
24322         * MiniBrowser/qt/MiniBrowser.qrc: Renamed from Tools/MiniBrowser/MiniBrowser.qrc.
24323
24324 2011-04-01  Dirk Pranke  <dpranke@chromium.org>
24325
24326         Reviewed by Adam Barth.
24327
24328         new-run-webkit-tests: fix feature detection, skipped platform lists on mac
24329
24330         We apparently never implemented the code to skip tests based on
24331         what was compiled into DRT. Also, change the logic used to skip
24332         platform directories to match what old-run-webkit-tests does:
24333         skip every test not in a directory in the baseline search path.
24334
24335         https://bugs.webkit.org/show_bug.cgi?id=57662
24336
24337         * Scripts/webkitpy/layout_tests/port/mac.py:
24338         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
24339         * Scripts/webkitpy/layout_tests/port/webkit.py:
24340
24341 2011-03-29  Mark Rowe  <mrowe@apple.com>
24342
24343         Reviewed by Jon Honeycutt.
24344
24345         <http://webkit.org/b/56730> new-run-webkit-tests fails on Lion seed
24346
24347         Teach new-run-webkit-tests about the concept of an unreleased version of Mac OS X.
24348
24349         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
24350         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
24351         * Scripts/webkitpy/layout_tests/port/mac.py:
24352         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
24353
24354 2011-04-01  Chang Shu  <cshu@webkit.org>
24355
24356         Reviewed by Csaba Osztrogonác.
24357
24358         [Qt][WK2] Build Webkit2 using "-2" option on Qt
24359         https://bugs.webkit.org/show_bug.cgi?id=55074
24360
24361         * Scripts/build-webkit:
24362
24363 2011-04-01  Dirk Pranke  <dpranke@chromium.org>
24364
24365         Reviewed by Tony Chang.
24366
24367         new-run-webkit-tests: remove spurious port version override in webkit.py
24368         base.py provides a default implementation so this is just
24369         breaking things.
24370
24371         https://bugs.webkit.org/show_bug.cgi?id=57667
24372
24373         * Scripts/webkitpy/layout_tests/port/webkit.py:
24374         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
24375
24376 2011-04-01  Csaba Osztrogonác  <ossy@webkit.org>
24377
24378         Reviewed by Adam Roben.
24379
24380         Make view results on waterfall direct link to results.html
24381         https://bugs.webkit.org/show_bug.cgi?id=57671
24382
24383         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Add "/results.html" to the URL.
24384         * Scripts/old-run-webkit-tests: Add links to httpd access and error logs.
24385
24386 2011-04-01  Sam Weinig  <sam@webkit.org>
24387
24388         Fix windows build.
24389
24390         * TestWebKitAPI/Tests/WebKit2/win/HideFindIndicator.cpp:
24391         (TestWebKitAPI::TEST):
24392         * TestWebKitAPI/Tests/WebKit2/win/ResizeViewWhileHidden.cpp:
24393         (TestWebKitAPI::flushMessages):
24394
24395 2011-04-01  Sam Weinig  <sam@webkit.org>
24396
24397         Reviewed by Adam Roben.
24398
24399         Add adoptWK to WKRetainPtr.h
24400         https://bugs.webkit.org/show_bug.cgi?id=57670
24401
24402         * TestWebKitAPI/PlatformUtilities.h:
24403         * TestWebKitAPI/Tests/WebKit2/AboutBlankLoad.cpp:
24404         * TestWebKitAPI/Tests/WebKit2/CanHandleRequest.cpp:
24405         * TestWebKitAPI/Tests/WebKit2/CanHandleRequest_Bundle.cpp:
24406         * TestWebKitAPI/Tests/WebKit2/CookieManager.cpp:
24407         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle.cpp:
24408         * TestWebKitAPI/Tests/WebKit2/PageLoadDidChangeLocationWithinPageForFrame.cpp:
24409         * TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly.cpp:
24410         * TestWebKitAPI/Tests/WebKit2/RestoreSessionStateContainingFormData.cpp:
24411         (TestWebKitAPI::createSessionStateContainingFormData):
24412         * WebKitTestRunner/StringFunctions.h:
24413         Replace custom versions of adoptWK with the API on in WebKit2/WKRetainPtr.h.
24414
24415 2011-04-01  Sam Weinig  <sam@webkit.org>
24416
24417         Reviewed by Adam Roben.
24418
24419         Fix leak noticed by Adam Roben in LayoutTestController::shadowRoot.
24420
24421         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
24422         (WTR::LayoutTestController::shadowRoot):
24423         Make judicious use of adoptWK().
24424
24425 2011-04-01  Sam Weinig  <sam@webkit.org>
24426
24427         Reviewed by Timothy Hatcher.
24428
24429         Fix extract-localizable-strings for macro change from UI_STRING -> WEB_UI_STRING.
24430
24431         * Scripts/extract-localizable-strings:
24432
24433 2011-04-01  Anders Carlsson  <andersca@apple.com>
24434
24435         Reviewed by Sam Weinig.
24436
24437         Shockwave plug-in doesn't accept mouse events
24438         https://bugs.webkit.org/show_bug.cgi?id=57653
24439         <rdar://problem/8483273>
24440
24441         Add a plug-in test.
24442
24443         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
24444         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
24445         (PluginTest::indicateTestFailure):
24446         Move code from NPDeallocateCalledBeforeNPShutdown::TestObject::~TestObject here.
24447
24448         (PluginTest::NPN_ConvertPoint):
24449         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
24450         * DumpRenderTree/TestNetscapePlugIn/Tests/NPDeallocateCalledBeforeNPShutdown.cpp:
24451         (NPDeallocateCalledBeforeNPShutdown::TestObject::~TestObject):
24452         Call indicateTestFailure.
24453
24454         * DumpRenderTree/TestNetscapePlugIn/Tests/mac: Added.
24455         * DumpRenderTree/TestNetscapePlugIn/Tests/mac/ConvertPoint.cpp: Added.
24456         (ConvertPoint::ConvertPoint):
24457         (ConvertPoint::testConvert):
24458         (ConvertPoint::NPP_New):
24459
24460 2011-04-01  Dominic Cooney  <dominicc@google.com>
24461
24462         Reviewed by Martin Robinson.
24463
24464         Add layoutTestController.shadowRoot to GTK DumpRenderTree.
24465         https://bugs.webkit.org/show_bug.cgi?id=57551
24466
24467         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
24468         (LayoutTestController::shadowRoot):
24469
24470 2011-04-01  Sam Weinig  <sam@webkit.org>
24471
24472         Reviewed by Anders Carlsson.
24473
24474         WebKitTestRunner needs layoutTestController.shadowRoot
24475         https://bugs.webkit.org/show_bug.cgi?id=57661
24476
24477         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
24478         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
24479         (WTR::LayoutTestController::shadowRoot):
24480         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
24481         Implement layoutTestController.shadowRoot for WebKit2.
24482
24483 2011-04-01  Chang Shu  <cshu@webkit.org>
24484
24485         Reviewed by Darin Adler.
24486
24487         WebKitTestRunner needs layoutTestController.setDatabaseQuota
24488         https://bugs.webkit.org/show_bug.cgi?id=57568
24489
24490         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
24491         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
24492         (WTR::LayoutTestController::setDatabaseQuota):
24493         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
24494
24495 2011-04-01  Dirk Pranke  <dpranke@chromium.org>
24496
24497         Reviewed by Tony Chang.
24498
24499         This patch completes the cleanup of
24500         rebaseline-chromium-webkit-tests to work with all of the
24501         variants of a platform (we can now rebaseline gpu- and non-gpu
24502         files at the same time).
24503
24504         When the rebaselining is complete, any lines declared as
24505         REBASELINE in the expectations file that matches a test that was
24506         actually rebaselined will be deleted, even if only one of the
24507         variants was actually rebaselined. This may cause odd problems,
24508         but is better than where we're at today.
24509
24510         This change removes the -g flag and deprecates -w. The -g flag is gone
24511         because GPU baselines are handled just like any other variant.
24512         The -w flag is deprecated because this tool now only works
24513         against the canaries, since that's the only place we have a full
24514         set of bots. It will be trivial to change this to
24515         build.webkit.org if we decide that's where we want them to be.
24516
24517         Also, this patch deletes a lot of cruft that is no longer needed
24518         in the test_expectations code and the port-specific code.
24519
24520         https://bugs.webkit.org/show_bug.cgi?id=55191
24521
24522         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
24523         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
24524         * Scripts/webkitpy/layout_tests/port/base.py:
24525         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
24526         * Scripts/webkitpy/layout_tests/port/chromium.py:
24527         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
24528         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
24529         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
24530         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
24531         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
24532         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
24533         * Scripts/webkitpy/layout_tests/port/mac.py:
24534         * Scripts/webkitpy/layout_tests/port/test.py:
24535         * Scripts/webkitpy/layout_tests/port/webkit.py:
24536         * Scripts/webkitpy/layout_tests/port/win.py:
24537         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
24538         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
24539
24540 2011-04-01  Dirk Pranke  <dpranke@chromium.org>
24541
24542         Reviewed by Adam Barth.
24543
24544         rebaseline-chromium-webkit-tests does not work correctly with
24545         version-specific baselines. This patch updates the tool to use
24546         all of the version-specific bots on the canaries, and will now
24547         attempt to rebaseline all of the versions by default, although
24548         it will not update both GPU and CPU versions.
24549
24550         Also, it will no longer modify the test_expectations.txt file
24551         *at all*. You will have to manually delete the REBASELINE lines
24552         after running the tool and determining that it did what you
24553         wanted it to do. This should be fixed in a separate bug - see
24554         webkit bug #55191.
24555
24556         https://bugs.webkit.org/show_bug.cgi?id=55608
24557
24558         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
24559         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
24560         * Scripts/webkitpy/layout_tests/port/chromium.py:
24561         * Scripts/webkitpy/layout_tests/port/test.py:
24562         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
24563         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
24564
24565 2011-03-31  Tony Chang  <tony@chromium.org>
24566
24567         Reviewed by Ojan Vafai.
24568
24569         [chromium] update the rebaseline tool to know about pngs with checksums
24570         https://bugs.webkit.org/show_bug.cgi?id=57481
24571
24572         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py: Skip over .checksum files if the checksum is already in the png
24573         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
24574
24575 2011-04-01  Adam Roben  <aroben@apple.com>
24576
24577         Retrieve revision numbers from the build's got_revision property in Leaks Viewer
24578
24579         Previously, we were getting the revision of the first revision that triggered a build.
24580         Choosing the last revision would have been more accurate. But got_revision is what is used
24581         everywhere else on build.webkit.org, and should work even when there were no changes that
24582         triggered a build (e.g., if someone clicked the Force Build button).
24583
24584         Fixes <http://webkit.org/b/57630> Leaks viewer gets some revision numbers wrong in the
24585         recent builds list
24586
24587         Reviewed by Anders Carlsson.
24588
24589         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/RecentBuildsLoader.js:
24590         (RecentBuildsLoader.prototype.start): Pull the revision number out of the got_revision
24591         property, rather than out of the first (i.e., earliest) change in the sourceStamp object.
24592
24593         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Utilities.js:
24594         (Array.prototype.first): Added this helper function to return the first element in an array
24595         that matches the given predicate, or null if no such element exists.
24596
24597 2011-03-31  Adam Roben  <aroben@apple.com>
24598
24599         Prefer (but don't require) bug URLs to be on their own line when parsing bug numbers from ChangeLogs
24600
24601         Fixes <http://webkit.org/b/57579> webkit-patch is too strict about bug URL formatting
24602
24603         Reviewed by Darin Adler.
24604
24605         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
24606         (parse_bug_id_from_changelog): Fall back to parse_bug_id if we weren't able to find a bug
24607         URL on its own line.
24608
24609         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
24610         (BugzillaTest.test_parse_bug_id_from_changelog): Updated expected results for test
24611         progression, and added a new test that uses a short bug URL while I was at it.
24612
24613 2011-04-01  Adam Roben  <aroben@apple.com>
24614
24615         Mark .vcproj/.vsprops/.sln files as being Windows-only
24616
24617         Fixes <http://webkit.org/b/57489> Mac builders built 82512, but shouldn't have
24618
24619         Reviewed by Anders Carlsson.
24620
24621         * Scripts/webkitpy/common/config/build.py:
24622         (_should_file_trigger_build): Added patterns to mark .vcproj/.vsprops/.sln files and .vcproj
24623         directories as Windows-only.
24624
24625         * Scripts/webkitpy/common/config/build_unittest.py:
24626         (ShoulBuildTest): Added test cases for the above.
24627
24628 2011-04-01  Keith Kyzivat  <keith.kyzivat@nokia.com>
24629
24630         Reviewed by Laszlo Gombos.
24631
24632         [Qt] Build MiniBrowser for Symbian
24633         https://bugs.webkit.org/show_bug.cgi?id=56319
24634
24635         Have MiniBrowser reference it's own copy of useragentlist.txt instead
24636         of copying QtTestBrowser's.
24637         Remove Tools/MiniBrowser/DerivedSources.pro
24638         This reduces complexity in the Tools scripts due to Symbian limitations.
24639
24640         * DerivedSources.pro:
24641         * MiniBrowser/DerivedSources.pro: Removed.
24642         * MiniBrowser/MiniBrowser.qrc:
24643         * MiniBrowser/qt/MiniBrowser.pro:
24644         * Scripts/webkitdirs.pm:
24645
24646 2011-03-31  Chang Shu  <cshu@webkit.org>
24647
24648         Reviewed by Darin Adler.
24649
24650         WebKitTestRunner needs layoutTestController.clearAllDatabases
24651         https://bugs.webkit.org/show_bug.cgi?id=42540
24652
24653         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
24654         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
24655         (WTR::InjectedBundle::beginTesting):
24656         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
24657         (WTR::LayoutTestController::clearAllDatabases):
24658         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
24659
24660 2011-03-31  Chang Shu  <cshu@webkit.org>
24661
24662         Reviewed by Darin Adler.
24663
24664         WebKitTestRunner needs layoutTestController.setAllowUniversalAccessFromFileURLs
24665         https://bugs.webkit.org/show_bug.cgi?id=42692
24666
24667         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
24668
24669 2011-03-31  Darin Adler  <darin@apple.com>
24670
24671         Reviewed by Sam Weinig.
24672
24673         Implement mouseDown, mouseUp, and mouseMoveTo in WebKitTestRunner
24674         https://bugs.webkit.org/show_bug.cgi?id=57573
24675
24676         * WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
24677         Added real definitions for mouseDown, mouseUp, mouseMoveTo and leapForward.
24678         Removed fake definitions of keyDown and contextClick.
24679
24680         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
24681         (WTR::operator==): Added. So we can compare two WKPoint structs.
24682         (WTR::parseModifier): Added.
24683         (WTR::parseModifierArray): Added.
24684         (WTR::EventSendingController::EventSendingController): Initialize the
24685         new data members.
24686         (WTR::EventSendingController::mouseDown): Added. Calls
24687         WKBundlePageSimulateMouseDown.
24688         (WTR::EventSendingController::mouseUp): Added. Calls
24689         WKBundlePageSimulateMouseUp.
24690         (WTR::EventSendingController::mouseMoveTo): Added. Calls
24691         WKBundlePageSimulateMouseMotion.
24692         (WTR::EventSendingController::leapForward): Added.
24693         (WTR::EventSendingController::updateClickCount): Added. Used by the
24694         mouseDown/Up functions to create a click count.
24695
24696         * WebKitTestRunner/InjectedBundle/EventSendingController.h: Updated
24697         for the changes above.
24698
24699 2011-03-31  Sergio Villar Senin  <svillar@igalia.com>
24700
24701         Reviewed by Martin Robinson.
24702
24703         [GTK] implement LayoutTestController::setWillSendRequestReturnsNull
24704         https://bugs.webkit.org/show_bug.cgi?id=57362
24705
24706         Do not generate DRT output if willSendRequestReturnsNull is set.
24707
24708         * DumpRenderTree/gtk/DumpRenderTree.cpp:
24709         (willSendRequestCallback):
24710
24711 2011-03-31  Vamshikrishna.Yellenki  <vamshi@motorola.com> and Alejandro G. Castro  <alex@igalia.com>
24712
24713         Reviewed by Martin Robinson.
24714
24715         Implement MiniBrowser for Gtk port.
24716         https://bugs.webkit.org/show_bug.cgi?id=48512
24717
24718         Initial implementation of the Gtk MiniBrowser.
24719
24720         * MiniBrowser/gtk/GNUmakefile.am: Added.
24721         * MiniBrowser/gtk/main.c: Added.
24722         (activateUriEntryCallback):
24723         (destroyCallback):
24724         (goBackCallback):
24725         (goForwardCallback):
24726         (createToolbar):
24727         (createWebView):
24728         (createWindow):
24729         (argumentToURL):
24730         (main):
24731
24732 2011-03-30  Dominic Cooney  <dominicc@google.com>
24733
24734         Reviewed by Dimitri Glazkov.
24735
24736         Adds layoutTestController.shadowRoot accessor to Mac DRT.
24737         https://bugs.webkit.org/show_bug.cgi?id=57415
24738
24739         * DumpRenderTree/LayoutTestController.cpp:
24740         (shadowRootCallback):
24741         (LayoutTestController::staticFunctions):
24742         * DumpRenderTree/LayoutTestController.h:
24743         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
24744         (LayoutTestController::shadowRoot):
24745         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
24746         (LayoutTestController::shadowRoot):
24747
24748 2011-03-30  Matthew Delaney  <mdelaney@apple.com>
24749
24750         Reviewed by Chris Marrin.
24751
24752         Update fast/canvas tests to avoid dumping the render tree when possible
24753         https://bugs.webkit.org/show_bug.cgi?id=57493
24754
24755         * DumpRenderTree/mac/DumpRenderTree.mm: Make DRT aware of new default values for
24756           accelerated drawing and accelerated drawing for canvas
24757
24758 2011-03-30  Adam Roben  <aroben@apple.com>
24759
24760         Stop ignoring leaks in CGGradientCreateWithColorStops
24761
24762         Fixes <rdar://problem/7888547>.
24763
24764         Rubber-stamped by John Sullivan.
24765
24766         * Scripts/old-run-webkit-tests:
24767         (countAndPrintLeaks): Removed some code to ignore those leaks.
24768
24769 2011-03-30  Timur Iskhodzhanov  <timurrrr@google.com>
24770
24771         Reviewed by Alexey Proskuryakov.
24772
24773         Add some dynamic annotations to JavaScriptCore/wtf
24774         https://bugs.webkit.org/show_bug.cgi?id=53747
24775
24776         By using these annotations we can improve the precision of finding
24777         WebKit errors using dynamic analysis tools like ThreadSanitizer and Valgrind.
24778         These annotations don't affect the compiled binaries unless USE(DYNAMIC_ANNOTATIONS) is "1".
24779
24780         These files don't add new functionality, so don't need extra tests.
24781
24782         * DumpRenderTree/ForwardingHeaders/wtf/DynamicAnnotations.h: Added.
24783
24784 2011-03-30  Steve Falkenburg  <sfalken@apple.com>
24785
24786         Reviewed by Adam Roben.
24787
24788         Share most vsprops between Release and Production builds in releaseproduction.vsprops
24789         https://bugs.webkit.org/show_bug.cgi?id=57508
24790
24791         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginProduction.vsprops:
24792         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginRelease.vsprops:
24793         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginReleaseCairoCFLite.vsprops:
24794         * DumpRenderTree/win/DumpRenderTreeProduction.vsprops:
24795         * DumpRenderTree/win/DumpRenderTreeRelease.vsprops:
24796         * DumpRenderTree/win/DumpRenderTreeReleaseCairoCFLite.vsprops:
24797         * DumpRenderTree/win/ImageDiffProduction.vsprops:
24798         * DumpRenderTree/win/ImageDiffRelease.vsprops:
24799         * DumpRenderTree/win/ImageDiffReleaseCairoCFLite.vsprops:
24800         * FindSafari/FindSafariProduction.vsprops:
24801         * FindSafari/FindSafariRelease.vsprops:
24802         * FindSafari/FindSafariReleaseCairoCFLite.vsprops:
24803         * FindSafari/FindSafariReleasePGO.vsprops:
24804         * MiniBrowser/Configurations/MiniBrowserProduction.vsprops:
24805         * MiniBrowser/Configurations/MiniBrowserRelease.vsprops:
24806         * MiniBrowser/Configurations/MiniBrowserReleaseCairoCFLite.vsprops:
24807         * TestWebKitAPI/Configurations/TestWebKitAPIRelease.vsprops:
24808         * TestWebKitAPI/Configurations/TestWebKitAPIReleaseCairoCFLite.vsprops:
24809         * WebKitAPITest/WebKitAPITestProduction.vsprops:
24810         * WebKitAPITest/WebKitAPITestRelease.vsprops:
24811         * WebKitAPITest/WebKitAPITestReleaseCairoCFLite.vsprops:
24812         * WebKitLauncherWin/WebKitLauncherWinProduction.vsprops:
24813         * WebKitLauncherWin/WebKitLauncherWinRelease.vsprops:
24814         * WebKitLauncherWin/WebKitLauncherWinReleaseCairoCFLite.vsprops:
24815         * WebKitTestRunner/win/InjectedBundleProduction.vsprops:
24816         * WebKitTestRunner/win/InjectedBundleRelease.vsprops:
24817         * WebKitTestRunner/win/InjectedBundleReleaseCairoCFLite.vsprops:
24818         * WebKitTestRunner/win/WebKitTestRunnerProduction.vsprops:
24819         * WebKitTestRunner/win/WebKitTestRunnerRelease.vsprops:
24820         * WebKitTestRunner/win/WebKitTestRunnerReleaseCairoCFLite.vsprops:
24821         * WinLauncher/WinLauncherProduction.vsprops:
24822         * WinLauncher/WinLauncherRelease.vsprops:
24823         * WinLauncher/WinLauncherReleaseCairoCFLite.vsprops:
24824         * record-memory-win/record-memory-winProduction.vsprops:
24825         * record-memory-win/record-memory-winRelease.vsprops:
24826         * record-memory-win/record-memory-winReleaseCairoCFLite.vsprops:
24827
24828 2011-03-30  Steve Falkenburg  <sfalken@apple.com>
24829
24830         Reviewed by Adam Roben.
24831
24832         Update Windows production build logic for new production configurations
24833         https://bugs.webkit.org/show_bug.cgi?id=57494
24834
24835         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginProduction.vsprops:
24836         * DumpRenderTree/win/DumpRenderTreeProduction.vsprops:
24837         * DumpRenderTree/win/ImageDiffProduction.vsprops:
24838         * FindSafari/FindSafariProduction.vsprops:
24839         * FindSafari/FindSafariReleasePGO.vsprops:
24840         * MiniBrowser/Configurations/MiniBrowserProduction.vsprops:
24841         * WebKitAPITest/WebKitAPITestProduction.vsprops:
24842         * WebKitLauncherWin/WebKitLauncherWinProduction.vsprops:
24843         * WebKitTestRunner/win/InjectedBundleProduction.vsprops:
24844         * WebKitTestRunner/win/WebKitTestRunnerProduction.vsprops:
24845         * WinLauncher/WinLauncherProduction.vsprops:
24846         * record-memory-win/record-memory-winProduction.vsprops:
24847
24848 2011-03-30  Robert Hogan  <robert@webkit.org>
24849
24850         Reviewed by Antonio Gomes.
24851
24852         [Qt] Fix LoadHTMLStringItem::invoke() after r75966
24853         Unskip http/tests/navigation/go-back-to-error-page.html
24854
24855         Also add the location of DumpRenderTreeSupportQt.h
24856         to DRT's include paths.
24857
24858         https://bugs.webkit.org/show_bug.cgi?id=52614
24859
24860         * DumpRenderTree/qt/DumpRenderTree.pro:
24861         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
24862         * DumpRenderTree/qt/DumpRenderTreeQt.h:
24863         * DumpRenderTree/qt/GCControllerQt.cpp:
24864         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
24865         (LayoutTestController::queueLoadHTMLString):
24866         * DumpRenderTree/qt/LayoutTestControllerQt.h:
24867         * DumpRenderTree/qt/PlainTextControllerQt.cpp:
24868         * DumpRenderTree/qt/TextInputControllerQt.cpp:
24869         * DumpRenderTree/qt/WorkQueueItemQt.cpp:
24870         (LoadAlternateHTMLStringItem::invoke):
24871         * DumpRenderTree/qt/WorkQueueItemQt.h:
24872         (LoadAlternateHTMLStringItem::LoadAlternateHTMLStringItem):
24873         * QtTestBrowser/QtTestBrowser.pro:
24874         * QtTestBrowser/launcherwindow.h:
24875
24876 2011-03-30  Adam Barth  <abarth@webkit.org>
24877
24878         Reviewed by Adam Roben.
24879
24880         SheriffBot rollouts take too long
24881         https://bugs.webkit.org/show_bug.cgi?id=57498
24882
24883         We used to build before landing rollouts via the commit-queue to
24884         prevent further breakage, but now that our individual commit-queue
24885         machines are slower, building takes too long.  I can't remember the
24886         last time a rollout broke compile.  It seems like just landing the
24887         patch is the better trade-off.
24888
24889         * Scripts/webkitpy/tool/bot/commitqueuetask.py:
24890         * Scripts/webkitpy/tool/commands/queues_unittest.py:
24891
24892 2011-03-30  Adam Barth  <abarth@webkit.org>
24893
24894         Remove platform/chromium-mac-snowleopard
24895         https://bugs.webkit.org/show_bug.cgi?id=57486
24896
24897         This directory no longer exists, so we can remove it from the fallback
24898         chain.
24899
24900         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
24901
24902 2011-03-30  Steve Falkenburg  <sfalken@apple.com>
24903
24904         Reviewed by Adam Roben.
24905
24906         Rename Windows configuration Release_LTCG to Production for clarity
24907         https://bugs.webkit.org/show_bug.cgi?id=57465
24908
24909         * DumpRenderTree/DumpRenderTree.sln:
24910         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
24911         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginProduction.vsprops: Copied from Tools/DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginReleaseLTCG.vsprops.
24912         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginReleaseLTCG.vsprops: Removed.
24913         * DumpRenderTree/win/DumpRenderTree.vcproj:
24914         * DumpRenderTree/win/DumpRenderTreeProduction.vsprops: Copied from Tools/DumpRenderTree/win/DumpRenderTreeReleaseLTCG.vsprops.
24915         * DumpRenderTree/win/DumpRenderTreeReleaseLTCG.vsprops: Removed.
24916         * DumpRenderTree/win/ImageDiff.vcproj:
24917         * DumpRenderTree/win/ImageDiffProduction.vsprops: Copied from Tools/DumpRenderTree/win/ImageDiffReleaseLTCG.vsprops.
24918         * DumpRenderTree/win/ImageDiffReleaseLTCG.vsprops: Removed.
24919         * FindSafari/FindSafari.vcproj:
24920         * FindSafari/FindSafariProduction.vsprops: Copied from Tools/FindSafari/FindSafariReleaseLTCG.vsprops.
24921         * FindSafari/FindSafariReleaseLTCG.vsprops: Removed.
24922         * MiniBrowser/Configurations/MiniBrowserProduction.vsprops: Copied from Tools/MiniBrowser/Configurations/MiniBrowserReleaseLTCG.vsprops.
24923         * MiniBrowser/Configurations/MiniBrowserReleaseLTCG.vsprops: Removed.
24924         * MiniBrowser/MiniBrowser.vcproj:
24925         * Scripts/webkitdirs.pm:
24926         * TestWebKitAPI/win/TestWebKitAPI.sln:
24927         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
24928         * TestWebKitAPI/win/TestWebKitAPIGenerated.vcproj:
24929         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
24930         * WebKitAPITest/WebKitAPITest.vcproj:
24931         * WebKitAPITest/WebKitAPITestProduction.vsprops: Copied from Tools/WebKitAPITest/WebKitAPITestReleaseLTCG.vsprops.
24932         * WebKitAPITest/WebKitAPITestReleaseLTCG.vsprops: Removed.
24933         * WebKitLauncherWin/WebKitLauncherWin.vcproj:
24934         * WebKitLauncherWin/WebKitLauncherWinProduction.vsprops: Copied from Tools/WebKitLauncherWin/WebKitLauncherWinReleaseLTCG.vsprops.
24935         * WebKitLauncherWin/WebKitLauncherWinReleaseLTCG.vsprops: Removed.
24936         * WebKitTestRunner/WebKitTestRunner.sln:
24937         * WebKitTestRunner/win/InjectedBundle.vcproj:
24938         * WebKitTestRunner/win/InjectedBundleGenerated.vcproj:
24939         * WebKitTestRunner/win/InjectedBundleProduction.vsprops: Copied from Tools/WebKitTestRunner/win/InjectedBundleReleaseLTCG.vsprops.
24940         * WebKitTestRunner/win/InjectedBundleReleaseLTCG.vsprops: Removed.
24941         * WebKitTestRunner/win/WebKitTestRunner.vcproj:
24942         * WebKitTestRunner/win/WebKitTestRunnerProduction.vsprops: Copied from Tools/WebKitTestRunner/win/WebKitTestRunnerReleaseLTCG.vsprops.
24943         * WebKitTestRunner/win/WebKitTestRunnerReleaseLTCG.vsprops: Removed.
24944         * WinLauncher/WinLauncher.vcproj:
24945         * WinLauncher/WinLauncherProduction.vsprops: Copied from Tools/WinLauncher/WinLauncherReleaseLTCG.vsprops.
24946         * WinLauncher/WinLauncherReleaseLTCG.vsprops: Removed.
24947         * record-memory-win/record-memory-win.vcproj:
24948         * record-memory-win/record-memory-winProduction.vsprops: Copied from Tools/record-memory-win/record-memory-winReleaseLTCG.vsprops.
24949         * record-memory-win/record-memory-winReleaseLTCG.vsprops: Removed.
24950
24951 2011-03-30  MORITA Hajime  <morrita@google.com>
24952
24953         Reviewed by Dimitri Glazkov.
24954
24955         [Chromium] Expose the shadow DOM to DumpRenderTree JS tests.
24956         https://bugs.webkit.org/show_bug.cgi?id=56573
24957
24958         Added LayoutTestController.shadowRoot() to Chromium DRT.
24959
24960         * DumpRenderTree/chromium/LayoutTestController.cpp:
24961         (LayoutTestController::LayoutTestController):
24962         (LayoutTestController::shadowRoot):
24963         * DumpRenderTree/chromium/LayoutTestController.h:
24964
24965 2011-03-30  Yuta Kitamura  <yutak@chromium.org>
24966
24967         Reviewed by Kent Tamura.
24968
24969         [Chromium] DumpRenderTree: Implement LayoutTestController::setPluginsEnabled
24970         https://bugs.webkit.org/show_bug.cgi?id=57430
24971
24972         * DumpRenderTree/chromium/LayoutTestController.cpp:
24973         (LayoutTestController::LayoutTestController):
24974         (LayoutTestController::setPluginsEnabled):
24975         * DumpRenderTree/chromium/LayoutTestController.h:
24976
24977 2011-03-30  Maciej Stachowiak  <mjs@apple.com>
24978
24979         Reviewed by Adam Barth.
24980
24981         make webkit-patch upload respect -d
24982         https://bugs.webkit.org/show_bug.cgi?id=57425
24983
24984         * Scripts/webkitpy/common/checkout/scm.py:
24985         * Scripts/webkitpy/tool/steps/commit.py:
24986
24987 2011-03-30  Maciej Stachowiak  <mjs@apple.com>
24988
24989         Reviewed by Adam Barth.
24990
24991         Make "webkit-patch --dry-run --verbose land" log the SVN command it's going to use
24992         https://bugs.webkit.org/show_bug.cgi?id=57429
24993
24994         * Scripts/webkitpy/common/checkout/scm.py:
24995
24996 2011-03-29  Sheriff Bot  <webkit.review.bot@gmail.com>
24997
24998         Unreviewed, rolling out r82383.
24999         http://trac.webkit.org/changeset/82383
25000         https://bugs.webkit.org/show_bug.cgi?id=57417
25001
25002         "nrwt isn't shutting down cleanly" (Requested by dpranke on
25003         #webkit).
25004
25005         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
25006         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
25007         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
25008
25009 2011-03-29  Kent Tamura  <tkent@chromium.org>
25010
25011         Reviewed by Dimitri Glazkov.
25012
25013         Make validation message bubble testable
25014         https://bugs.webkit.org/show_bug.cgi?id=57290
25015
25016         * DumpRenderTree/mac/DumpRenderTree.mm:
25017         (createWebViewAndOffscreenWindow): Change a setting so that validation
25018           bubbles isn't hidden automatically.
25019
25020 2011-03-29  Dirk Pranke  <dpranke@chromium.org>
25021
25022         Reviewed by Ojan Vafai.
25023
25024         new-run-webkit-tests: use 'threads' on win instead of 'old-threads'
25025
25026         It's a bit unclear whether multiple threads or multiple
25027         processes will be a better model on windows. This change will
25028         test threads for now while we're still working out the bugs in
25029         multiple-processes.
25030
25031         https://bugs.webkit.org/show_bug.cgi?id=57410
25032
25033         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
25034         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
25035         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
25036
25037 2011-03-29  David Levin  <levin@chromium.org>
25038
25039         Reviewed by Shinichiro Hamaji.
25040
25041         check-webkit-style confused by two ChangeLog entries in a row from same user
25042         https://bugs.webkit.org/show_bug.cgi?id=57250
25043
25044         * Scripts/webkitpy/style/checker.py: Add the line should be checked function to ChangeLogChecker.
25045         * Scripts/webkitpy/style/checker_unittest.py: Fix test due to that new function.
25046         * Scripts/webkitpy/style/checkers/changelog.py: Made this code aware of what lines were being checked.
25047           It basically assumes only one ChangeLog entry is being processed because that is the standard case and
25048           checking more than that would be very messey.
25049         * Scripts/webkitpy/style/checkers/changelog_unittest.py: Add testing to catch the broken case.
25050         * Scripts/webkitpy/style/error_handlers.py: Added should_line_be_checked.
25051
25052 2011-03-29  Kent Tamura  <tkent@chromium.org>
25053
25054         Reviewed by Dimitri Glazkov.
25055
25056         [Mac] Enable interactive-validation tests on Mac DRT
25057         https://bugs.webkit.org/show_bug.cgi?id=57308
25058
25059         * DumpRenderTree/mac/DumpRenderTree.mm:
25060         (createWebViewAndOffscreenWindow): Enable the form interactive validation feature.
25061
25062 2011-03-29  Chang Shu  <cshu@webkit.org>
25063
25064         Reviewed by Kenneth Rohde Christiansen.
25065
25066         WebKitTestRunner needs layoutTestController.setAllowUniversalAccessFromFileURLs
25067         https://bugs.webkit.org/show_bug.cgi?id=42692
25068
25069         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
25070         (WTR::InjectedBundle::beginTesting):
25071         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
25072         (WTR::LayoutTestController::setAllowUniversalAccessFromFileURLs):
25073         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
25074
25075 2011-03-24  Martin Robinson  <mrobinson@igalia.com>
25076
25077         Reviewed by Xan Lopez.
25078
25079         [GTK] [DRT] GtkScrolledWindow adds three pixels of padding between the WebView and the scrollbar
25080         https://bugs.webkit.org/show_bug.cgi?id=57067
25081
25082         Eliminate extra spacing between the DRT GtkScrolledWindow and the DRT WebView. This
25083         will allow WebKit1GTK+ and WebKit2GTK+ to share test results. Followup commits will
25084         update DRT and pixel results as hundreds of results will need to be updated.
25085
25086         * DumpRenderTree/gtk/DumpRenderTree.cpp:
25087         (setDefaultsToConsistentStateValuesForTesting): Eliminate spacing on the GtkScrolledWindow
25088         by overriding the theme RC/CSS files.
25089
25090 2011-03-29  Steve Falkenburg  <sfalken@apple.com>
25091
25092         Reviewed by Adam Roben.
25093
25094         Use per-configuration vsprops in tools projects to avoid WebKitVSPropsRedirectionDir removal by MSVC IDE
25095         https://bugs.webkit.org/show_bug.cgi?id=57394
25096
25097         Visual Studio's IDE was removing instances of $(WebKitVSPropsRedirectionDir) from
25098         InheritedPropertySheet rules in our vcproj files when the vcproj was edited from within
25099         the IDE. To avoid this, add a separate vsprops file for each project configuration that
25100         contains the required inherited property sheets.
25101
25102         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
25103         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginDebug.vsprops: Added.
25104         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginDebugAll.vsprops: Added.
25105         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginDebugCairoCFLite.vsprops: Added.
25106         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginRelease.vsprops: Added.
25107         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginReleaseCairoCFLite.vsprops: Added.
25108         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginReleaseLTCG.vsprops: Added.
25109         * DumpRenderTree/win/DumpRenderTree.vcproj:
25110         * DumpRenderTree/win/DumpRenderTreeDebug.vsprops: Added.
25111         * DumpRenderTree/win/DumpRenderTreeDebugAll.vsprops: Added.
25112         * DumpRenderTree/win/DumpRenderTreeDebugCairoCFLite.vsprops: Added.
25113         * DumpRenderTree/win/DumpRenderTreeRelease.vsprops: Added.
25114         * DumpRenderTree/win/DumpRenderTreeReleaseCairoCFLite.vsprops: Added.
25115         * DumpRenderTree/win/DumpRenderTreeReleaseLTCG.vsprops: Added.
25116         * DumpRenderTree/win/ImageDiff.vcproj:
25117         * DumpRenderTree/win/ImageDiffDebug.vsprops: Added.
25118         * DumpRenderTree/win/ImageDiffDebugAll.vsprops: Added.
25119         * DumpRenderTree/win/ImageDiffDebugCairoCFLite.vsprops: Added.
25120         * DumpRenderTree/win/ImageDiffRelease.vsprops: Added.
25121         * DumpRenderTree/win/ImageDiffReleaseCairoCFLite.vsprops: Added.
25122         * DumpRenderTree/win/ImageDiffReleaseLTCG.vsprops: Added.
25123         * FindSafari/FindSafari.vcproj:
25124         * FindSafari/FindSafariDebug.vsprops: Added.
25125         * FindSafari/FindSafariDebugAll.vsprops: Added.
25126         * FindSafari/FindSafariDebugCairoCFLite.vsprops: Added.
25127         * FindSafari/FindSafariRelease.vsprops: Added.
25128         * FindSafari/FindSafariReleaseCairoCFLite.vsprops: Added.
25129         * FindSafari/FindSafariReleaseLTCG.vsprops: Added.
25130         * FindSafari/FindSafariReleasePGO.vsprops: Added.
25131         * MiniBrowser/Configurations/MiniBrowserDebug.vsprops: Added.
25132         * MiniBrowser/Configurations/MiniBrowserDebugAll.vsprops: Added.
25133         * MiniBrowser/Configurations/MiniBrowserDebugCairoCFLite.vsprops: Added.
25134         * MiniBrowser/Configurations/MiniBrowserRelease.vsprops: Added.
25135         * MiniBrowser/Configurations/MiniBrowserReleaseCairoCFLite.vsprops: Added.
25136         * MiniBrowser/Configurations/MiniBrowserReleaseLTCG.vsprops: Added.
25137         * MiniBrowser/MiniBrowser.vcproj:
25138         * WebKitAPITest/WebKitAPITest.vcproj:
25139         * WebKitAPITest/WebKitAPITestDebug.vsprops: Added.
25140         * WebKitAPITest/WebKitAPITestDebugAll.vsprops: Added.
25141         * WebKitAPITest/WebKitAPITestDebugCairoCFLite.vsprops: Added.
25142         * WebKitAPITest/WebKitAPITestRelease.vsprops: Added.
25143         * WebKitAPITest/WebKitAPITestReleaseCairoCFLite.vsprops: Added.
25144         * WebKitAPITest/WebKitAPITestReleaseLTCG.vsprops: Added.
25145         * WebKitLauncherWin/WebKitLauncherWin.vcproj:
25146         * WebKitLauncherWin/WebKitLauncherWinDebug.vsprops: Added.
25147         * WebKitLauncherWin/WebKitLauncherWinDebugAll.vsprops: Added.
25148         * WebKitLauncherWin/WebKitLauncherWinDebugCairoCFLite.vsprops: Added.
25149         * WebKitLauncherWin/WebKitLauncherWinRelease.vsprops: Added.
25150         * WebKitLauncherWin/WebKitLauncherWinReleaseCairoCFLite.vsprops: Added.
25151         * WebKitLauncherWin/WebKitLauncherWinReleaseLTCG.vsprops: Added.
25152         * WebKitTestRunner/win/InjectedBundle.vcproj:
25153         * WebKitTestRunner/win/InjectedBundleDebug.vsprops: Added.
25154         * WebKitTestRunner/win/InjectedBundleDebugAll.vsprops: Added.
25155         * WebKitTestRunner/win/InjectedBundleDebugCairoCFLite.vsprops: Added.
25156         * WebKitTestRunner/win/InjectedBundleRelease.vsprops: Added.
25157         * WebKitTestRunner/win/InjectedBundleReleaseCairoCFLite.vsprops: Added.
25158         * WebKitTestRunner/win/InjectedBundleReleaseLTCG.vsprops: Added.
25159         * WebKitTestRunner/win/WebKitTestRunner.vcproj:
25160         * WebKitTestRunner/win/WebKitTestRunnerDebug.vsprops: Added.
25161         * WebKitTestRunner/win/WebKitTestRunnerDebugAll.vsprops: Added.
25162         * WebKitTestRunner/win/WebKitTestRunnerDebugCairoCFLite.vsprops: Added.
25163         * WebKitTestRunner/win/WebKitTestRunnerRelease.vsprops: Added.
25164         * WebKitTestRunner/win/WebKitTestRunnerReleaseCairoCFLite.vsprops: Added.
25165         * WebKitTestRunner/win/WebKitTestRunnerReleaseLTCG.vsprops: Added.
25166         * WinLauncher/WinLauncher.vcproj:
25167         * WinLauncher/WinLauncherDebug.vsprops: Added.
25168         * WinLauncher/WinLauncherDebugAll.vsprops: Added.
25169         * WinLauncher/WinLauncherDebugCairoCFLite.vsprops: Added.
25170         * WinLauncher/WinLauncherRelease.vsprops: Added.
25171         * WinLauncher/WinLauncherReleaseCairoCFLite.vsprops: Added.
25172         * WinLauncher/WinLauncherReleaseLTCG.vsprops: Added.
25173         * record-memory-win/record-memory-win.vcproj:
25174         * record-memory-win/record-memory-winDebug.vsprops: Added.
25175         * record-memory-win/record-memory-winDebugAll.vsprops: Added.
25176         * record-memory-win/record-memory-winDebugCairoCFLite.vsprops: Added.
25177         * record-memory-win/record-memory-winRelease.vsprops: Added.
25178         * record-memory-win/record-memory-winReleaseCairoCFLite.vsprops: Added.
25179         * record-memory-win/record-memory-winReleaseLTCG.vsprops: Added.
25180
25181 2011-03-29  Adam Barth  <abarth@webkit.org>
25182
25183         Reviewed by Darin Adler.
25184
25185         test_runner2.py can crash due to undefined variable
25186         https://bugs.webkit.org/show_bug.cgi?id=57356
25187
25188         Replace reference to undefined variable with a variable that's actually
25189         defined.  AFAICT, there's no way to test this code.
25190
25191         * Scripts/webkitpy/layout_tests/layout_package/test_runner2.py:
25192
25193 2011-03-29  Tony Chang  <tony@chromium.org>
25194
25195         Reviewed by Eric Seidel.
25196
25197         cleanup rebaseline-chromium-webkit-tests
25198         https://bugs.webkit.org/show_bug.cgi?id=57375
25199
25200         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
25201             - Split _extract_and_add_new_baselines into 3 functions
25202             - Reduce the amount of line wrapping (the file already has lines
25203               over 80 col, so may as well try to make the file consistent)
25204             - Remove unnecessary ()s
25205
25206 2011-03-29  Brent Fulgham  <bfulgham@webkit.org>
25207
25208         Reviewed by Adam Roben.
25209
25210         Make WinCairo a core builder.
25211         https://bugs.webkit.org/show_bug.cgi?id=57373
25212
25213         * Scripts/webkitpy/common/net/buildbot/buildbot.py: Add WinCairo.
25214         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py: Add WinCairo.
25215
25216 2011-03-29  Timothy Hatcher  <timothy@apple.com>
25217
25218         Update update-webkit-localizable-strings to understand that WebKit and
25219         WebKit2 strings go in WebCore.
25220
25221         Also make extract-localizable-strings optionally use an exclude file. No individual
25222         warnings are produced about unlocalized strings when there is no exclude file. This
25223         is needed for WebCore, since there are too many bare strings.
25224
25225         https://webkit.org/b/57354
25226
25227         Reviewed by Sam Weinig.
25228
25229         * Scripts/extract-localizable-strings: Support "-" to mean no exclude file.
25230         * Scripts/update-webkit-localizable-strings: Update paths to scan WebCore and WebKit2.
25231
25232 2011-03-29  Darin Adler  <darin@apple.com>
25233
25234         Reviewed by Adam Roben.
25235
25236         WebKit2 bundle page needs to offer generated file hooks
25237         https://bugs.webkit.org/show_bug.cgi?id=57279
25238
25239         (Land missing part of the original patch after r82289.)
25240
25241         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
25242         (WTR::InjectedBundlePage::InjectedBundlePage): Added new callbacks.
25243
25244 2011-03-29  Tony Chang  <tony@chromium.org>
25245
25246         Reviewed by Ojan Vafai.
25247
25248         [chromium] NRWT should be able to read checksums from png comments
25249         https://bugs.webkit.org/show_bug.cgi?id=57280
25250
25251         * Scripts/read-checksum-from-png: Added. Utility script to read the checksum
25252             from a file.
25253         * Scripts/webkitpy/common/system/filesystem.py: Add open_binary_file_for_reading
25254         * Scripts/webkitpy/common/system/filesystem_mock.py:
25255         * Scripts/webkitpy/layout_tests/port/base.py: When loading a checksum,
25256             if the -checksum.txt file doesn't exist, try looking in the png.
25257         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
25258         * Scripts/webkitpy/layout_tests/port/test.py:
25259         * Scripts/webkitpy/layout_tests/read_checksum_from_png.py: Added. Scan the first
25260             2k for a png comment with the checksum
25261         * Scripts/webkitpy/layout_tests/read_checksum_from_png_unittest.py: Added.
25262
25263 2011-03-29  Philippe Normand  <pnormand@igalia.com>
25264
25265         Rubber-stamped by Gustavo Noronha Silva.
25266
25267         [GTK] http/tests/uri/username-with-no-hostname.html fails
25268         https://bugs.webkit.org/show_bug.cgi?id=57325
25269
25270         * DumpRenderTree/gtk/DumpRenderTree.cpp:
25271         (willSendRequestCallback): URI host value check done case-insensitively.
25272
25273 2011-03-29  Philippe Normand  <pnormand@igalia.com>
25274
25275         Reviewed by Martin Robinson.
25276
25277         [GTK] http/tests/uri/username-with-no-hostname.html fails
25278         https://bugs.webkit.org/show_bug.cgi?id=57325
25279
25280         * DumpRenderTree/gtk/DumpRenderTree.cpp:
25281         (willSendRequestCallback): Check URI before submitting the request
25282         to WebCore.
25283
25284 2011-03-29  Noel Gordon  <noel.gordon@gmail.com>
25285
25286         Reviewed by Ojan Vafai.
25287
25288         [chromium] DRT EventSender: remove identity from dragTargetDragEnter() calls
25289         https://bugs.webkit.org/show_bug.cgi?id=57303
25290
25291         * DumpRenderTree/chromium/EventSender.cpp:
25292         (EventSender::doDragDrop):
25293         (EventSender::beginDragWithFiles):
25294
25295 2011-03-29  Zoltan Horvath  <zoltan@webkit.org>
25296
25297         Reviewed by Andreas Kling.
25298
25299         [Qt] Add -print-loaded-urls option to Qt's MiniBrowser and to QtTestBrowser
25300         https://bugs.webkit.org/show_bug.cgi?id=57314
25301
25302         Provide possibility to print loaded urls to the standard output.
25303
25304         * MiniBrowser/qt/BrowserWindow.cpp:
25305         (BrowserWindow::BrowserWindow):
25306         (BrowserWindow::printURL):
25307         * MiniBrowser/qt/BrowserWindow.h:
25308         * MiniBrowser/qt/MiniBrowserApplication.cpp:
25309         (MiniBrowserApplication::handleUserOptions):
25310         * MiniBrowser/qt/MiniBrowserApplication.h:
25311         (WindowOptions::WindowOptions):
25312         * QtTestBrowser/launcherwindow.cpp:
25313         (LauncherWindow::initializeView):
25314         (LauncherWindow::printURL):
25315         (LauncherWindow::cloneWindow):
25316         * QtTestBrowser/launcherwindow.h:
25317         (WindowOptions::WindowOptions):
25318         * QtTestBrowser/main.cpp:
25319         (LauncherApplication::handleUserOptions):
25320
25321 2011-03-29  Gabor Loki  <loki@webkit.org>
25322
25323         Rubber-stamped by Csaba Osztrogonác.
25324
25325         [Qt] Teach build-jsc how to build JavaScriptCore on Qt
25326         https://bugs.webkit.org/show_bug.cgi?id=56918
25327
25328         * Scripts/build-jsc: Remove duplicated --qt options to avoid passing them to qmake.
25329
25330 2011-03-28  Gabor Loki  <loki@webkit.org>
25331
25332         Reviewed by Csaba Osztrogonác.
25333
25334         [Qt] Teach build-jsc how to build JavaScriptCore on Qt
25335         https://bugs.webkit.org/show_bug.cgi?id=56918
25336
25337         * Scripts/build-jsc:
25338         * Scripts/webkitdirs.pm:
25339
25340 2011-03-28  Andrew Foster  <andrewf@chromium.org>
25341
25342         Reviewed by Ojan Vafai.
25343
25344         webkit-patch should be more intelligent about whether a bug applies to a patch
25345         Create a new function, parse_bug_id_from_changelog() which determines
25346         that a bug is related to a patch by parsing the output generated from
25347         prepare-ChangeLog, rather than arbitrarily matching a URL to a bug in
25348         the description of a patch.
25349         https://bugs.webkit.org/show_bug.cgi?id=56989
25350
25351         * Scripts/webkitpy/common/checkout/api.py:
25352         * Scripts/webkitpy/common/checkout/changelog.py:
25353         * Scripts/webkitpy/common/net/bugzilla/__init__.py:
25354         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
25355         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
25356         * Scripts/webkitpy/style/checkers/changelog.py:
25357         * Scripts/webkitpy/tool/commands/upload.py:
25358
25359 2011-03-28  Maciej Stachowiak  <mjs@apple.com>
25360
25361         Revert accidentally commited change.
25362
25363         * DumpRenderTree/mac/ResourceLoadDelegate.mm:
25364         (-[ResourceLoadDelegate webView:resource:willSendRequest:redirectResponse:fromDataSource:]):
25365
25366 2011-03-28  Tony Chang  <tony@chromium.org>
25367
25368         Reviewed by Adam Barth.
25369
25370         [chromium] have DRT write the png checksum into a png comment
25371         https://bugs.webkit.org/show_bug.cgi?id=57255
25372
25373         * DumpRenderTree/chromium/TestShell.cpp:
25374         (TestShell::dumpImage):
25375
25376 2011-03-28  Adele Peterson  <adele@apple.com>
25377
25378         Removing Qt stub, since it doesn't use the same header as the others.
25379
25380         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
25381
25382 2011-03-28  Adele Peterson  <adele@apple.com>
25383
25384         Added a few more stubs.
25385
25386         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
25387         (LayoutTestController::hasGrammarMarker):
25388         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
25389         (LayoutTestController::hasGrammarMarker):
25390         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
25391         (LayoutTestController::hasGrammarMarker):
25392
25393 2011-03-28  Adele Peterson  <adele@apple.com>
25394
25395         Build fix.
25396
25397         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
25398         (LayoutTestController::hasGrammarMarker):
25399
25400 2011-03-28  Adele Peterson  <adele@apple.com>
25401
25402         Reviewed by Eric Seidel.
25403
25404         Testing support for <rdar://problem/9112694> REGRESSION (r79411): "Check grammar with spelling" context menu doesn't check as you type
25405         https://bugs.webkit.org/show_bug.cgi?id=57173
25406
25407         * DumpRenderTree/LayoutTestController.cpp:
25408         (hasGrammarMarkerCallback): Added.
25409         (LayoutTestController::staticFunctions): Added case for hasGrammarMarker.
25410         * DumpRenderTree/LayoutTestController.h:
25411         * DumpRenderTree/mac/DumpRenderTree.mm: (createWebViewAndOffscreenWindow): Call setGrammarCheckingEnabled.
25412         * DumpRenderTree/mac/LayoutTestControllerMac.mm: (LayoutTestController::hasGrammarMarker): Added. Call new hasGrammarMarker method.
25413
25414 2011-03-28  Csaba Osztrogonác  <ossy@webkit.org>
25415
25416         Reviewed by Benjamin Poulain.
25417
25418         [Qt] QtWebKit will not compile with QT_ASCII_CAST_WARNINGS enabled
25419         https://bugs.webkit.org/show_bug.cgi?id=57087
25420
25421         Trivial typo fix after r82082.
25422
25423         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
25424         (WTR::activateFonts):
25425
25426 2011-03-28  Alexis Menard  <alexis.menard@openbossa.org>
25427
25428         Reviewed by Andreas Kling.
25429
25430         Add myself to the committers list.
25431
25432         * Scripts/webkitpy/common/config/committers.py:
25433
25434 2011-03-28  Simon Fraser  <simon.fraser@apple.com>
25435
25436         Reviewed by Sam Weinig.
25437
25438         WebKit2 unfortunately uses code from WebKit, so MiniBrowser needs
25439         to link against WebKit.framework for now.
25440
25441         * MiniBrowser/MiniBrowser.xcodeproj/project.pbxproj:
25442
25443 2011-03-28  Benjamin Poulain  <benjamin.poulain@nokia.com>
25444
25445         Reviewed by Andreas Kling.
25446
25447         [Qt] QtWebKit will not compile with QT_ASCII_CAST_WARNINGS enabled
25448         https://bugs.webkit.org/show_bug.cgi?id=57087
25449
25450         Disable QT_ASCII_CAST_WARNINGS for applications.
25451
25452         * DumpRenderTree/qt/DumpRenderTree.pro:
25453         * DumpRenderTree/qt/ImageDiff.pro:
25454         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
25455         * MiniBrowser/qt/MiniBrowser.pro:
25456         * QtTestBrowser/QtTestBrowser.pro:
25457         * WebKitTestRunner/qt/WebKitTestRunner.pro:
25458
25459 2011-03-28  Andreas Kling  <kling@webkit.org>
25460
25461         Reviewed by Kenneth Rohde Christiansen.
25462
25463         [Qt] QtTestBrowser doesn't reset its window title when opening about:blank
25464         https://bugs.webkit.org/show_bug.cgi?id=57224
25465
25466         * QtTestBrowser/mainwindow.cpp:
25467         (MainWindow::buildUI):
25468         (MainWindow::onTitleChanged):
25469         * QtTestBrowser/mainwindow.h:
25470
25471 2011-03-27  Jer Noble  <jer.noble@apple.com>
25472
25473         Reviewed by Maciej Stachowiak.
25474
25475         Full Screen: disable keyboard access by default
25476         https://bugs.webkit.org/show_bug.cgi?id=56684
25477
25478         Accept the withKeyboard parameter to supportsFullScreenForElement.
25479
25480         * DumpRenderTree/mac/UIDelegate.mm:
25481         (-[UIDelegate webView:supportsFullScreenForElement:withKeyboard:]):
25482
25483 2011-03-27  Patrick Gansterer  <paroga@webkit.org>
25484
25485         Reviewed by David Levin.
25486
25487         check-webkit-style should check ChangeLog for a valid bug number
25488         https://bugs.webkit.org/show_bug.cgi?id=57184
25489
25490         * Scripts/webkitpy/style/checker.py:
25491         * Scripts/webkitpy/style/checker_unittest.py:
25492         * Scripts/webkitpy/style/checkers/changelog.py: Added.
25493         * Scripts/webkitpy/style/checkers/changelog_unittest.py: Added.
25494
25495 2011-03-27  Benjamin Poulain  <benjamin.poulain@nokia.com>
25496
25497         Reviewed by Andreas Kling.
25498
25499         [Qt] QtWebKit will not compile with QT_ASCII_CAST_WARNINGS enabled
25500         https://bugs.webkit.org/show_bug.cgi?id=57087
25501
25502         Use explicit conversion for string to avoid depending on the default codec
25503         installed by the user code.
25504
25505         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
25506         (WTR::activateFonts):
25507
25508 2011-03-26  Maciej Stachowiak  <mjs@apple.com>
25509
25510         Revert inadvertantly committed changes.
25511
25512         * Scripts/old-run-webkit-tests:
25513         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
25514
25515 2011-03-26  Andreas Kling  <kling@webkit.org>
25516
25517         Reviewed by Kenneth Rohde Christiansen.
25518
25519         [Qt] Show page icons (favicons) in QtTestBrowser location bar.
25520         https://bugs.webkit.org/show_bug.cgi?id=57162
25521
25522         * QtTestBrowser/QtTestBrowser.qrc:
25523         * QtTestBrowser/favicon.png: Added.
25524         * QtTestBrowser/locationedit.cpp:
25525         (defaultPageIcon):
25526         (LocationEdit::LocationEdit):
25527         (LocationEdit::setPageIcon):
25528         (LocationEdit::resizeEvent):
25529         (LocationEdit::updateInternalGeometry):
25530         * QtTestBrowser/locationedit.h:
25531         * QtTestBrowser/mainwindow.cpp:
25532         (MainWindow::buildUI):
25533         (MainWindow::onIconChanged):
25534         (MainWindow::onLoadStarted):
25535         * QtTestBrowser/mainwindow.h:
25536
25537 2011-03-23  Martin Robinson  <mrobinson@igalia.com>
25538
25539         Reviewed by Xan Lopez.
25540
25541         [GTK] Force DumpRenderTree to use 96 DPI
25542         https://bugs.webkit.org/show_bug.cgi?id=56866
25543
25544         Make sure that GTK+ is set to use 96 DPI when running tests. This ensures
25545         broader compatibility with WebKit2 and makes setting font sizes more straightforward.
25546
25547         * DumpRenderTree/gtk/DumpRenderTree.cpp:
25548         (initializeGtkFontSettings): Set the screen resolution and GTK+ xft property to 96 DPI.
25549         (resetDefaultsToConsistentValues):Remove the call which repeatedly set the GDK screen resolution.
25550         (setDefaultsToConsistentStateValuesForTesting): Update the font sizes to reflect the new DPI.
25551
25552 2011-03-25  Andy Estes  <aestes@apple.com>
25553
25554         Reviewed by Adele Peterson.
25555
25556         REGRESSION (r70748): latest nightly builds kills AC_QuickTime.js
25557         https://bugs.webkit.org/show_bug.cgi?id=49016
25558
25559         TestNetscapePlugIn needs to register support for an image MIME type so
25560         we can test that the embed tag prefers plug-ins to render image MIME
25561         types.
25562
25563         * DumpRenderTree/TestNetscapePlugIn/mac/Info.plist: Register image/png.
25564         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
25565         (NP_GetMIMEDescription): Ditto.
25566         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.rc: Ditto.
25567
25568 2011-03-25  Jessie Berlin  <jberlin@apple.com>
25569
25570         Reviewed by Sam Weinig.
25571
25572         WebKit2: Need to be able to set and get the Cookie Storage Policy.
25573         https://bugs.webkit.org/show_bug.cgi?id=50780
25574
25575         Add a test for getting and setting the HTTP Cookie Accept Policy in WebKit2.
25576
25577         The test is named "CookieManager" so that we can eventually exercise more of the
25578         CookieManager functionality in the test.
25579
25580         * TestWebKitAPI/Tests/WebKit2/CookieManager.cpp: Added.
25581         (TestWebKitAPI::didGetTestHTTPCookieAcceptPolicy):
25582         Assert that the policy returned is equal to the policy set in
25583         didGetUserHTTPCookieAcceptPolicy, and then restore the user's policy.
25584         (TestWebKitAPI::didGetUserHTTPCookieAcceptPolicy):
25585         Set the policy to something different than the user's policy and get the policy again.
25586         (TestWebKitAPI::didFinishLoadForFrame):
25587         Get the user's policy so that it can be restored at the end of the test.
25588         (TestWebKitAPI::TEST):
25589         Load about:blank so that the Web Process gets instatiated (needed to get the
25590         CookieManager).
25591
25592         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
25593         Add CookieManager.cpp.
25594
25595 2011-03-24  Kevin Ollivier  <kevino@theolliviers.com>
25596
25597         [wx] Fix the build so that gcc-4.1 and up can be used to build on SnowLeopard.
25598
25599         * wx/build/settings.py:
25600
25601 2011-03-24  Ilya Sherman  <isherman@chromium.org>
25602
25603         Reviewed by Adam Roben.
25604
25605         Implement layoutTestController.setAutoFilled in DRT on Windows
25606         https://bugs.webkit.org/show_bug.cgi?id=56828
25607
25608         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
25609         (LayoutTestController::setAutofilled): Implemented.
25610
25611 2011-03-24  Brent Fulgham  <bfulgham@webkit.org>
25612
25613         Unreviewed build correction.
25614
25615         Correct vsprops file for Debug variant of the CFLite build.
25616         Disable a CFNETWORK-specific test when building for CFLite.
25617
25618         * TestWebKitAPI/Configurations/TestWebKitAPIDebugCairoCFLite.vsprops:
25619         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
25620
25621 2011-03-24  Dirk Pranke  <dpranke@chromium.org>
25622
25623         Reviewed by Adam Barth.
25624
25625         fix regressions introduced in 81908 - there were a couple of
25626         code paths that only fired under python 2.5 that I missed.
25627
25628         Also change a couple of 'python' references to sys.executable
25629         in order to handle runing test-webkitpy with a binary other than
25630         something called 'python'.
25631
25632         https://bugs.webkit.org/show_bug.cgi?id=57063
25633
25634         * Scripts/webkitpy/layout_tests/port/mac.py:
25635         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
25636         * Scripts/webkitpy/test/cat.py:
25637         * Scripts/webkitpy/test/echo.py:
25638
25639 2011-03-24  Brent Fulgham  <bfulgham@webkit.org>
25640
25641         Unreviewed build correction.
25642
25643         Add a stub implementation of the TestInvocation dumping logic.
25644         Update project files to reflect new file.
25645
25646         * WebKitTestRunner/cairo: Added.
25647         * WebKitTestRunner/cairo/TestInvocationCairo.cpp: Added.
25648         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
25649         * WebKitTestRunner/win/WebKitTestRunner.vcproj:
25650
25651 2011-03-24  Dirk Pranke  <dpranke@chromium.org>
25652
25653         Reviewed by Tony Chang.
25654
25655         new-run-webkit-tests: clean up worker model defaults. This
25656         changes the default behavior to 'threads' instead of
25657         'old-threads', and clarifies that chromium-win-* and
25658         chromium-mac-leopard are exceptions.
25659
25660         https://bugs.webkit.org/show_bug.cgi?id=56971
25661
25662         * Scripts/webkitpy/layout_tests/port/base.py:
25663         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
25664         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
25665         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
25666         * Scripts/webkitpy/layout_tests/port/test.py:
25667         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
25668
25669 2011-03-23  Leandro Gracia Gil  <leandrogracia@chromium.org>
25670
25671         Reviewed by Darin Adler.
25672
25673         Media Stream API: fix the order of the options in build-webkit.
25674         https://bugs.webkit.org/show_bug.cgi?id=56949
25675
25676         Fix the order of the options in build-webkit.
25677
25678         * Scripts/build-webkit:
25679
25680 2011-03-23  Adam Klein  <adamk@chromium.org>
25681
25682         Reviewed by Mihai Parparita.
25683
25684         Add an option to new-run-webkit-httpd to allow overriding the location of LayoutTests directory
25685         https://bugs.webkit.org/show_bug.cgi?id=56884
25686
25687         This option will be utilized by Chromium's ui_tests to allow
25688         referencing js-test-resources from a layout test over HTTP.
25689
25690         * Scripts/new-run-webkit-httpd:
25691         Added --layout_tests_dir option.
25692         * Scripts/webkitpy/layout_tests/port/http_server.py:
25693         Plumbed through as layout_tests_dir param.
25694
25695 2011-03-23  Dirk Pranke  <dpranke@chromium.org>
25696
25697         Reviewed by Mihai Parparita.
25698
25699         new-run-webkit-tests: run tests in ascending alphabetical order per
25700         dir. This is a second attempt at the fix; the first attempt, in
25701         r81597, broke the behavior on the old-inline and old-threads
25702         worker models.
25703
25704         https://bugs.webkit.org/show_bug.cgi?id=56760
25705
25706         * Scripts/webkitpy/layout_tests/layout_package/dump_render_tree_thread.py:
25707         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
25708         * Scripts/webkitpy/layout_tests/port/test.py:
25709         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
25710
25711 2011-03-23  Mario Sanchez Prada  <msanchez@igalia.com>
25712
25713         Reviewed by Martin Robinson.
25714
25715         [GTK] GTK's DRT to ouput detail's string for the 'property-change' signal
25716         https://bugs.webkit.org/show_bug.cgi?id=56953
25717
25718         Print the string for the detail of 'property-change' signal.
25719
25720         * DumpRenderTree/gtk/AccessibilityCallbacks.cpp:
25721         (axObjectEventListener): Print the detail string.
25722
25723 2011-03-23  Yury Semikhatsky  <yurys@chromium.org>
25724
25725         Reviewed by Pavel Feldman.
25726
25727         [V8] Web Inspector: compile DebuggerScript.js into DebuggerScriptSource.h
25728         https://bugs.webkit.org/show_bug.cgi?id=56843
25729
25730         * DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
25731         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
25732
25733 2011-03-22  Eric Seidel  <eric@webkit.org>
25734
25735         Reviewed by Adam Barth.
25736
25737         Teach update-webkit about gyp
25738         https://bugs.webkit.org/show_bug.cgi?id=56881
25739
25740         This makes it possible for someone who builds from Xcode
25741         to just run update-webkit --gyp and have the project files
25742         generated after update automatically.
25743
25744         I also added a stand-alone "generate-project-files" script
25745         wrapping Source/gyp/configure.  This makes running gyp simpler
25746         since most people will not have gyp or Source/gyp in their path.
25747
25748         * Scripts/build-webkit:
25749         * Scripts/generate-project-files: Added.
25750         * Scripts/update-webkit:
25751
25752 2011-03-22  Anton D'Auria  <adauria@apple.com>
25753
25754         Reviewed by Alexey Proskuryakov.
25755
25756         Add +[WebApplicationCache getOriginsWithCache]
25757         https://bugs.webkit.org/show_bug.cgi?id=56722
25758
25759         * DumpRenderTree/LayoutTestController.cpp:
25760         (originsWithApplicationCacheCallback):
25761         (LayoutTestController::staticFunctions):
25762         * DumpRenderTree/LayoutTestController.h:
25763         * DumpRenderTree/chromium/LayoutTestController.cpp: Added stub.
25764         (LayoutTestController::originsWithApplicationCache):
25765         * DumpRenderTree/chromium/LayoutTestController.h:
25766         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: Added stub.
25767         (LayoutTestController::originsWithApplicationCache):
25768         * DumpRenderTree/mac/LayoutTestControllerMac.mm: Pass array of origin's database identifiers.
25769         (originsArrayToJS): Helper function for converting array of WebSecurityOrigins to a JS array of origin identifiers.
25770         (LayoutTestController::originsWithApplicationCache):
25771         (LayoutTestController::originsWithLocalStorage):
25772         * DumpRenderTree/qt/LayoutTestControllerQt.cpp: Added stub.
25773         (LayoutTestController::originsWithApplicationCache):
25774         * DumpRenderTree/qt/LayoutTestControllerQt.h:
25775         * DumpRenderTree/win/LayoutTestControllerWin.cpp: Added stub.
25776         (LayoutTestController::originsWithApplicationCache):
25777         * DumpRenderTree/wx/LayoutTestControllerWx.cpp: Added stub.
25778         (LayoutTestController::originsWithApplicationCache):
25779
25780 2011-03-22  Eric Seidel  <eric@webkit.org>
25781
25782         Reviewed by Adam Barth.
25783
25784         Add support to build-webkit for building with gyp-generated project files
25785         https://bugs.webkit.org/show_bug.cgi?id=56877
25786
25787         build-webkit is a cesspool.  This change just hacks in --gyp support for Mac.
25788         Eventually we'll want to support gyp-generated projects in a more general manner.
25789
25790         * Scripts/build-webkit:
25791
25792 2011-03-22  David Kilzer  <ddkilzer@apple.com>
25793
25794         <http://webkit.org/b/56781> Add --dsym switch to enable dsym generation when building with Xcode
25795
25796         Reviewed by Joseph Pecoraro.
25797
25798         * Scripts/build-webkit: Updated usage statement to include
25799         --dsym.
25800         * Scripts/webkitdirs.pm: Added $generateDsym variable.
25801         (generateDsym): Added. Call determineGenerateDsym() and
25802         return $generateDsym.
25803         (determineGenerateDsym): Added.  Parse @ARGV for --dsym and set
25804         $generateDsym.
25805         (argumentsForXcode): Added.  Returns additional arguments for
25806         xcodebuild based on command-line switches.
25807         (XcodeOptions): Updated to include argumentsForXcode() in the
25808         array returned.
25809
25810 2011-03-22  Adam Roben  <aroben@apple.com>
25811
25812         Make Leopard Debug bots and Windows XP Debug (Tests) core builders
25813
25814         These bots are consistently green again.
25815
25816         Fixes <http://webkit.org/b/56830> Leopard Debug bots and Windows XP Debug (Tests) should be
25817         core builders
25818
25819         Rubber-stamped by Antti Koivisto.
25820
25821         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
25822         (BuildBot.__init__): Loosened the regexps for Leopard and Windows.
25823
25824         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
25825         (BuildBotTest.test_builder_name_regexps): Updated expectations.
25826
25827 2011-03-21  Sheriff Bot  <webkit.review.bot@gmail.com>
25828
25829         Unreviewed, rolling out r81597.
25830         http://trac.webkit.org/changeset/81597
25831         https://bugs.webkit.org/show_bug.cgi?id=56801
25832
25833         Change results in lots of unexpected flaky on the dashboard.
25834         (Requested by pfeldman on #webkit).
25835
25836         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
25837         * Scripts/webkitpy/layout_tests/port/test.py:
25838         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
25839
25840 2011-03-21  Sam Weinig  <sam@webkit.org>
25841
25842         Fix failing tests on the WebKit2 bots.
25843
25844         * WebKitTestRunner/TestController.cpp:
25845         (WTR::TestController::resetStateToConsistentValues):
25846
25847 2011-03-21  Daniel Sievers  <sievers@google.com>
25848
25849         Reviewed by Simon Fraser.
25850
25851         [Chromium] Make RenderAsTextBehavior and LayerTreeAsTextBehavior tweakable from the DumpRenderTree commandline
25852         https://bugs.webkit.org/show_bug.cgi?id=56139
25853
25854         * DumpRenderTree/chromium/DumpRenderTree.cpp:
25855         (main):
25856         * DumpRenderTree/chromium/LayoutTestController.cpp:
25857         (LayoutTestController::LayoutTestController):
25858         (LayoutTestController::layerTreeAsText):
25859         * DumpRenderTree/chromium/LayoutTestController.h:
25860         (LayoutTestController::setShowDebugLayerTree):
25861         * DumpRenderTree/chromium/TestShell.cpp:
25862         (TestShell::runFileTest):
25863         (TestShell::dump):
25864         * DumpRenderTree/chromium/TestShell.h:
25865         (TestParams::TestParams):
25866
25867 2011-03-21  Adam Roben  <aroben@apple.com>
25868
25869         Try again to get buildbot to show run-javascriptcore-tests's actual.html as an HTML file
25870
25871         Fixes <http://webkit.org/b/56746> build.webkit.org shows run-javascriptcore-tests's HTML
25872         output as plain text
25873
25874         Reviewed by David Kilzer.
25875
25876         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
25877         (RunJavaScriptCoreTests): List actual.html as one of our log files so it will get uploaded
25878         to the master.
25879         (RunJavaScriptCoreTests.commandComplete): Reading the actual.html file here won't work,
25880         since we're running on the build master, not the slave. Instead, turn the raw source of
25881         actual.html which the build slave uploaded into an HTML log.
25882
25883 2011-03-21  Dirk Pranke  <dpranke@chromium.org>
25884
25885         Reviewed by Mihai Parparita.
25886
25887         new-run-webkit-tests: run tests in ascending alphabetical order per dir
25888         https://bugs.webkit.org/show_bug.cgi?id=56760
25889
25890         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
25891         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
25892         * Scripts/webkitpy/layout_tests/port/test.py:
25893
25894 2011-03-21  Leandro Gracia Gil  <leandrogracia@chromium.org>
25895
25896         Reviewed by Steve Block.
25897
25898         Media Stream API patch 0: adding compilation guards.
25899         https://bugs.webkit.org/show_bug.cgi?id=56458
25900
25901         Adding options to enable the media stream API feature in Chromium.
25902
25903         * Scripts/build-webkit:
25904
25905 2011-03-21  Qi Zhang  <qi.2.zhang@nokia.com>
25906
25907         Reviewed by Laszlo Gombos.
25908
25909         [Qt] Add a command line option to capture stdout and stderr for DumpRenderTree
25910         https://bugs.webkit.org/show_bug.cgi?id=56323
25911
25912         Using freopen to redirect STDOUT and STDERR when DumpRenderTree command line provide
25913         "--stdout" or "--stderr" option.
25914
25915         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
25916         (WebCore::DumpRenderTree::~DumpRenderTree):
25917         (WebCore::DumpRenderTree::processArgsLine):
25918         (WebCore::DumpRenderTree::loadNextTestInStandAloneMode):
25919         * DumpRenderTree/qt/DumpRenderTreeQt.h:
25920         (WebCore::DumpRenderTree::setRedirectOutputFileName):
25921         (WebCore::DumpRenderTree::setRedirectErrorFileName):
25922         * DumpRenderTree/qt/main.cpp:
25923         (isOption):
25924         (takeOptionValue):
25925         (printUsage):
25926         (main):
25927
25928 2011-03-21  Adam Roben  <aroben@apple.com>
25929
25930         Fix exceptions on the buildbot due to my last change
25931
25932         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
25933         (RunJavaScriptCoreTests.commandComplete): Catch exceptions due to actual.html not existing.
25934
25935 2011-03-21  Adam Roben  <aroben@apple.com>
25936
25937         Teach buildbot to treat run-javascriptcore-tests's actual.html as an HTML file
25938
25939         Fixes <http://webkit.org/b/56746> build.webkit.org shows run-javascriptcore-tests's HTML
25940         output as plain text
25941
25942         Reviewed by Darin Adler.
25943
25944         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
25945         (RunJavaScriptCoreTests): Removed the logfiles variable, which can only handle plaintext
25946         logs.
25947         (RunJavaScriptCoreTests.commandComplete): Use the addHTMLLog method to upload actual.html to
25948         the build master. This will correctly treat it as HTML.
25949
25950 2011-03-20  Bill Budge  <bbudge@chromium.org>
25951
25952         Reviewed by Adam Barth.
25953
25954         Rename ThreadSafeShared to ThreadSafeRefCounted
25955         https://bugs.webkit.org/show_bug.cgi?id=56714
25956
25957         No new tests. Exposes no new functionality.
25958
25959         * DumpRenderTree/ForwardingHeaders/wtf/ThreadSafeRefCounted.h: Copied from DumpRenderTree/ForwardingHeaders/wtf/ThreadSafeShared.h.
25960         * DumpRenderTree/ForwardingHeaders/wtf/ThreadSafeShared.h: Removed.
25961         * Scripts/do-webcore-rename:
25962
25963 2011-03-19  Anton D'Auria  <adauria@apple.com>
25964
25965         Reviewed by Dan Bernstein.
25966
25967         Add stub to LayoutTestControllerGtk.cpp to fix GTK build
25968         https://bugs.webkit.org/show_bug.cgi?id=56719
25969
25970         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: Adding stub.
25971         (LayoutTestController::clearApplicationCacheForOrigin):
25972
25973 2011-03-19  Anton D'Auria  <adauria@apple.com>
25974
25975         Reviewed by Alexey Proskuryakov.
25976
25977         ApplicationCacheGroup is not obsolete after being deleted via ApplicationCacheStorage::deleteEntriesForOrigin
25978         https://bugs.webkit.org/show_bug.cgi?id=56415
25979
25980         * DumpRenderTree/LayoutTestController.cpp:
25981         (clearApplicationCacheForOriginCallback):
25982         (LayoutTestController::staticFunctions):
25983         * DumpRenderTree/LayoutTestController.h:
25984         * DumpRenderTree/chromium/LayoutTestController.cpp: Adding stubs.
25985         (LayoutTestController::LayoutTestController):
25986         (LayoutTestController::clearAllApplicationCaches):
25987         (LayoutTestController::clearApplicationCacheForOrigin):
25988         * DumpRenderTree/chromium/LayoutTestController.h:
25989         * DumpRenderTree/mac/LayoutTestControllerMac.mm: Adding stub.
25990         (LayoutTestController::clearApplicationCacheForOrigin):
25991         * DumpRenderTree/qt/LayoutTestControllerQt.cpp: Adding stub.
25992         (LayoutTestController::clearApplicationCacheForOrigin):
25993         * DumpRenderTree/qt/LayoutTestControllerQt.h:
25994         * DumpRenderTree/win/LayoutTestControllerWin.cpp: Adding stub.
25995         (LayoutTestController::clearApplicationCacheForOrigin):
25996         * DumpRenderTree/wx/LayoutTestControllerWx.cpp: Adding stub.
25997         (LayoutTestController::clearApplicationCacheForOrigin):
25998
25999 2011-03-18  Dan Bernstein  <mitz@apple.com>
26000
26001         Revised build fix for r81135.
26002
26003         * WebKitTestRunner/Configurations/InjectedBundle.xcconfig:
26004         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
26005
26006 2011-03-18  Adam Roben  <aroben@apple.com>
26007
26008         Windows build fix
26009
26010         * WebKitTestRunner/config.h: Add a missing #endif.
26011
26012 2011-03-18  Andreas Kling  <kling@webkit.org>
26013
26014         Unbreak Qt build after r81494.
26015
26016         * WebKitTestRunner/qt/TestInvocationQt.cpp:
26017         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
26018
26019 2011-03-18  Sam Weinig  <sam@webkit.org>
26020
26021         Fix windows build. This was really Adam Roben's fault. I swear!
26022
26023         * WebKitTestRunner/config.h:
26024
26025 2011-03-18  Sam Weinig  <sam@webkit.org>
26026
26027         Reviewed by Adam Roben.
26028
26029         Pixel tests don't work in WebKitTestRunner on Windows
26030         https://bugs.webkit.org/show_bug.cgi?id=56630
26031
26032         Cross-platformize (Mac and Windows CG) the pixel dumping code in WebKitTestRunner to use WebKit2
26033         built-in snapshotting instead of using the window server on the Mac.
26034
26035         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
26036         (WTR::InjectedBundle::InjectedBundle):
26037         (WTR::InjectedBundle::didReceiveMessage):
26038         (WTR::InjectedBundle::beginTesting):
26039         (WTR::InjectedBundle::done):
26040         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
26041         (WTR::InjectedBundle::setPixelResult):
26042         (WTR::InjectedBundle::shouldDumpPixels):
26043         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
26044         (WTR::InjectedBundlePage::dump):
26045         Take the snapshot in the bundle and send it to the UIProcess for dumping.
26046
26047         * WebKitTestRunner/TestInvocation.cpp:
26048         (WTR::TestInvocation::invoke):
26049         (WTR::TestInvocation::dump):
26050         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
26051         * WebKitTestRunner/TestInvocation.h:
26052         * WebKitTestRunner/cg: Added.
26053         * WebKitTestRunner/cg/TestInvocationCG.cpp: Copied from WebKitTestRunner/mac/TestInvocationMac.mm.
26054         (WTR::createCGContextFromImage):
26055         (WTR::computeMD5HashStringForContext):
26056         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
26057         * WebKitTestRunner/mac/TestInvocationMac.mm: Removed.
26058         Making pixel dumping code shared for CG ports. Also use the WTF/MD5 instead of platform specific code.
26059
26060         * WebKitTestRunner/win/TestInvocationWin.cpp:
26061         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
26062         Keep stub for Cairo build.
26063
26064         * WebKitTestRunner/Configurations/WebKitTestRunnerCoreGraphics.vsprops: Added.
26065         Add CG configuration.
26066
26067         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
26068         * WebKitTestRunner/win/WebKitTestRunner.vcproj:
26069         Add new files.
26070
26071 2011-03-18  Adam Roben  <aroben@apple.com>
26072
26073         Don't allow Leaks Viewer's vertical scrollbar to cover up long function names
26074
26075         Rubber-stamped by Simon Fraser.
26076
26077         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css:
26078         (.data-grid .data-container): Use overflow-y:auto instead of the Inspector's default
26079         overflow-y:overlay so that the scrollbar doesn't overlay the content.
26080
26081 2011-03-18  Adam Roben  <aroben@apple.com>
26082
26083         Don't ellipsize long or indented function names in Leaks Viewer
26084
26085         Fixes <http://webkit.org/b/56037> Leaks viewer needs horizontal
26086         scrollbar
26087
26088         Reviewed by Simon Fraser.
26089
26090         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css:
26091         (.data-grid .data-container):
26092         (.data-grid td > div, .data-grid th > div):
26093         Allow the contents to extend to the right.
26094
26095 2011-03-18  Adam Roben  <aroben@apple.com>
26096
26097         Fix some leaks in DRT seen on the leaks bot
26098
26099         Fixes <http://webkit.org/b/56638> addURLToRedirectCallback and
26100         setWillSendRequestClearHeaderCallback in DRT are leaky
26101
26102         Reviewed by Alexey Proskuryakov.
26103
26104         * DumpRenderTree/ForwardingHeaders/wtf/OwnArrayPtr.h: Added.
26105
26106         * DumpRenderTree/LayoutTestController.cpp:
26107         (addURLToRedirectCallback):
26108         (setWillSendRequestClearHeaderCallback):
26109         Use OwnArrayPtr to cause the strings to be deleted.
26110
26111 2011-03-18  Adam Roben  <aroben@apple.com>
26112
26113         Fix a NSURLCredential leak seen on the leaks bot
26114
26115         Fixes <http://webkit.org/b/56637> -[SynchronousLoader
26116         connection:didReceiveAuthenticationChallenge:] is leaky
26117
26118         Reviewed by John Sullivan.
26119
26120         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
26121         (-[SynchronousLoader connection:didReceiveAuthenticationChallenge:]): Use a RetainPtr to
26122         cause the NSURLCredential to be released.
26123
26124 2011-03-18  John Knottenbelt  <jknotten@chromium.org>
26125
26126         Unreviewed.
26127
26128         Remove unnecessary brackets around my email address.
26129
26130         * Scripts/webkitpy/common/config/committers.py:
26131
26132 2011-03-18  Adam Roben  <aroben@apple.com>
26133
26134         Remove some files I accidentally added in r81454
26135
26136         * WebKitTestRunner/Configurations/WebKitTestRunnerCoreGraphics.vsprops: Removed.
26137         * WebKitTestRunner/cg/TestInvocationCG.cpp: Removed.
26138
26139 2011-03-18  Adam Roben  <aroben@apple.com>
26140
26141         Ignore an NSNumberFormatter leak on SnowLeopard
26142
26143         * Scripts/old-run-webkit-tests:
26144         (countAndPrintLeaks): Ignore leaks in -[NSNumberFormatter
26145         getObjectValue:forString:errorDescription:].
26146
26147 2011-03-18  Adam Roben  <aroben@apple.com>
26148
26149         Fix a leak beneath LayoutTestController::authenticateSession seen on the bots
26150
26151         Rubber-stamped by Mark Rowe.
26152
26153         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
26154         (LayoutTestController::authenticateSession): Use a RetainPtr to cause the NSURLRequest to be
26155         released.
26156
26157 2011-03-17  Adam Roben  <aroben@apple.com>
26158
26159         Address a review comment I missed in r81445
26160
26161         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js:
26162         (LeaksParserWorker.prototype._parseLeaks): Anchor the RegExp used to find the number of
26163         leaked bytes to speed up searching on long lines.
26164
26165 2011-03-17  Adam Roben  <aroben@apple.com>
26166
26167         Mark leaky builds orange, even if all regression tests passed
26168
26169         Fixes <http://webkit.org/b/56609> SnowLeopard Intel Leaks bot appears green on
26170         build.webkit.org when all tests pass even though there are still leaks
26171
26172         Reviewed by Dan Bernstein.
26173
26174         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
26175         (RunWebKitLeakTests): Set the warnOnWarnings property to True so that the entire build will
26176         be marked orange if this build step generates warnings.
26177
26178 2011-03-17  Adam Roben  <aroben@apple.com>
26179
26180         Allow values in Leaks Viewer to be shown as percentages
26181
26182         Double-clicking on a column will toggle between showing percentages and showing real values.
26183
26184         Fixes <http://webkit.org/b/56628> Would like to view values as percentages of the total
26185         number of leaked bytes in Leaks Viewer
26186
26187         Reviewed by Joe Pecoraro.
26188
26189         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js:
26190         (LeaksParserWorker.prototype._incorporateLeaks): Calculate the (never-shown) top-level
26191         node's totalTime, which is used in calculating percentages.
26192
26193         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/WebInspectorShims.js:
26194         (monkeyPatchInspectorObjects): Only format values as bytes if we aren't supposed to be
26195         showing them as percentages.
26196
26197 2011-03-17  Adam Roben  <aroben@apple.com>
26198
26199         Show the number of leaked bytes, not just leaked allocations, in Leaks Viewer
26200
26201         The Self and Total columns now report the number of leaked bytes. The new Calls column shows
26202         the number of leaked allocations. And the new Average column shows the average number of
26203         leaked bytes per call.
26204
26205         Fixes <http://webkit.org/b/56344> Leaks Viewer should show how many bytes were leaked, not
26206         just how many allocations
26207
26208         Reviewed by Joe Pecoraro.
26209
26210         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js:
26211         (LeaksParserWorker.prototype._parseLeaks): Parse the number of leaked bytes from the "Leak:"
26212         line and store it with the leak stack.
26213         (LeaksParserWorker.prototype._createNode): Initialize the averageTime and numberOfCalls
26214         properties.
26215         (LeaksParserWorker.prototype._incorporateLeaks): Updated to get the stack from the .stack
26216         property, and changed to store the number of leaked bytes in .selfTime/.totalTime.
26217         .numberOfCalls now holds the number of leaked allocations.
26218
26219         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/WebInspectorShims.js:
26220         (Preferences): Set samplingCPUProfiler to false so that the Average and Calls columns will
26221         appear.
26222         (monkeyPatchInspectorObjects): Format the values as bytes.
26223
26224 2011-03-17  Adam Roben  <aroben@apple.com>
26225
26226         Make Leaks Viewer able to count leaks in builds that didn't fail any regression tests
26227
26228         Fixes <http://webkit.org/b/56626> Leaks Viewer omits recent builds which didn't fail any
26229         regression tests
26230
26231         Reviewed by James Robinson.
26232
26233         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/RecentBuildsLoader.js:
26234         (RecentBuildsLoader.prototype.start): Get the number of leaks out of the "text" array, which
26235         is always populated, even when all regression tests pass.
26236
26237 2011-03-10  Hayato Ito  <hayato@chromium.org>
26238
26239         Reviewed by Tony Chang.
26240
26241         Add a check logic for reftests in case that a reftest is marked AS REBASELINE.
26242
26243         https://bugs.webkit.org/show_bug.cgi?id=56076
26244
26245         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
26246
26247 2011-03-17  Adam Roben  <aroben@apple.com>
26248
26249         Skip a compositing-sensitive test when accelerated compositing is disabled
26250
26251         Reviewed by Sam Weinig.
26252
26253         * Scripts/old-run-webkit-tests: Skip media/media-document-audio-repaint.html when
26254         accelerated compositing is disabled.
26255
26256 2011-03-17  Adam Roben  <aroben@apple.com>
26257
26258         Start using PlatformSpecificScheduler for the Mac builders
26259
26260         The current set of file/directory patterns should work for Mac builders. We may discover
26261         that there are more files and directories we can exclude from the Mac builds, but they can
26262         be added later.
26263
26264         Fixes <http://webkit.org/b/56593> Mac builders do lots of unnecessary builds
26265
26266         Reviewed by Mark Rowe.
26267
26268         * BuildSlaveSupport/build.webkit.org-config/config.json: Moved the Mac builders into two
26269         PlatformSpecificSchedulers: one for Leopard, and one for SnowLeopard.
26270
26271         * Scripts/webkitpy/common/config/build.py:
26272         (_should_file_trigger_build): Allow "mac-leopard" and "mac-snowleopard" platforms, and
26273         cleaned up patterns a little bit.
26274
26275         * Scripts/webkitpy/common/config/build_unittest.py:
26276         (ShouldBuildTest): Fixed some expected results.
26277         (ShouldBuildTest.test_should_build): Start testing "mac-leopard" and "mac-snowleopard"
26278         platforms.
26279
26280 2011-03-17  Jeff Miller  <jeffm@apple.com>
26281
26282         Use a consistent set of file patterns in the svn:ignore property for all .xcodeproj directories, specifically:
26283
26284         *.mode*
26285         *.pbxuser
26286         *.perspective*
26287         project.xcworkspace
26288         xcuserdata
26289
26290         * DumpRenderTree/DumpRenderTree.xcodeproj: Modified property svn:ignore.
26291         * MiniBrowser/MiniBrowser.xcodeproj: Modified property svn:ignore.
26292         * TestWebKitAPI/TestWebKitAPI.xcodeproj: Modified property svn:ignore.
26293         * WebKitLauncher/WebKitLauncher.xcodeproj: Modified property svn:ignore.
26294         * WebKitTestRunner/WebKitTestRunner.xcodeproj: Modified property svn:ignore.
26295
26296 2011-03-17  Adam Roben  <aroben@apple.com>
26297
26298         Make old-run-webkit-tests pick up mac-wk2 results when running in WebKit2 mode on Windows
26299
26300         Fixes <http://webkit.org/b/56563> Running WebKit2 tests on Windows should use results from
26301         mac-wk2
26302
26303         Reviewed by Anders Carlsson.
26304
26305         * Scripts/old-run-webkit-tests:
26306         (expectedDirectoryForTest): Clean up the code to search in extra Mac platform result
26307         directories on Windows. Add mac-wk2 to the set of extra platforms we look in on Windows when
26308         we're running in WebKit2 mode.
26309
26310 2011-03-17  Mikhail Naganov  <mnaganov@chromium.org>
26311
26312         Reviewed by Yury Semikhatsky.
26313
26314         Web Inspector: Clean up Inspector strings.
26315         https://bugs.webkit.org/show_bug.cgi?id=56557
26316
26317         Modify the script to deal correctly with "\xNN" escapes, and to
26318         check strings in Chromium's DevTools.js.
26319
26320         * Scripts/check-inspector-strings:
26321
26322 2011-03-17  Keith Kyzivat  <keith.kyzivat@nokia.com>
26323
26324         Reviewed by Benjamin Poulain.
26325
26326         Update autoinstall to install mechanize 0.2.4
26327         https://bugs.webkit.org/show_bug.cgi?id=56499
26328
26329         Newer versions of mechanize support HTTPS proxies, which is necessary
26330         for people using webkit-patch behind firewalls.
26331
26332         * Scripts/webkitpy/common/system/autoinstall.py:
26333         * Scripts/webkitpy/thirdparty/__init__.py:
26334
26335 2011-03-16  Sam Weinig  <sam@webkit.org>
26336
26337         Reviewed by Adam Roben.
26338
26339         REGRESSION (r81084): media/invalid-media-url-crash.html and editing/execCommand/indent-paragraphs.html failing on WebKit2
26340         https://bugs.webkit.org/show_bug.cgi?id=56541
26341
26342         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
26343         (WTR::InjectedBundlePage::willSendRequestForFrame):
26344         Check for an empty host in addition to a null host.  This bug was introduced
26345         when porting off of the KURL code and onto the WKURLRef version.
26346
26347 2011-03-16  Sam Weinig  <sam@webkit.org>
26348
26349         Reviewed by Adam Roben.
26350
26351         Add Lucas Forschler to the list of WebKit committers.
26352
26353         * Scripts/webkitpy/common/config/committers.py:
26354
26355 2011-03-16  Sam Weinig  <sam@webkit.org>
26356
26357         Reviewed upon and dictated through Adam Roben.
26358
26359         Escape crashing function names in results.html. This fixes an issue
26360         when a crash involves a type with templates.
26361
26362         * Scripts/old-run-webkit-tests:
26363
26364 2011-03-16  Daniel Sievers  <sievers@google.com>
26365
26366         Reviewed by James Robinson.
26367
26368         Add setting to always force compositing mode
26369         https://bugs.webkit.org/show_bug.cgi?id=56156
26370
26371         * DumpRenderTree/chromium/DumpRenderTree.cpp:
26372         (main):
26373         * DumpRenderTree/chromium/TestShell.cpp:
26374         (TestShell::TestShell):
26375         (TestShell::resetWebSettings):
26376         * DumpRenderTree/chromium/TestShell.h:
26377         (TestShell::setForceCompositingMode):
26378         * DumpRenderTree/chromium/WebPreferences.cpp:
26379         (WebPreferences::reset):
26380         (WebPreferences::applyTo):
26381         * DumpRenderTree/chromium/WebPreferences.h:
26382
26383 2011-03-15  Hayato Ito  <hayato@chromium.org>
26384
26385         Reviewed by Ojan Vafai.
26386
26387         Ignore files that will be used by reftests in old-run-webkit-tests.
26388
26389         https://bugs.webkit.org/show_bug.cgi?id=55936
26390
26391         * Scripts/old-run-webkit-tests:
26392
26393 2011-03-15  Felipe Oliveira da Silva Netto  <f.dachshund@gmail.com>
26394
26395         Reviewed by Adam Roben.
26396
26397         Function relativeScriptsDir doesn't return the correct relative
26398         Scripts directory.
26399
26400         * Scripts/webkitdirs.pm:
26401         (relativeScriptsDir): Ensure we use the directory that contains
26402         webkitdirs.pm, not the one that contains the original script that
26403         was invoked.
26404
26405 2011-03-15  Kevin Ollivier  <kevino@theolliviers.com>
26406
26407         [wx] Build fixes for long linker commands on Win and method return error, plus a Mac
26408         fix to link against the Security framework after recent WebKitSystemInterface changes.
26409
26410         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
26411         (LayoutTestController::originsWithLocalStorage):
26412         * wx/build/settings.py:
26413         * wx/build/waf_extensions.py:
26414
26415 2011-03-15  Sam Weinig  <sam@webkit.org>
26416
26417         Reviewed by Darin Adler.
26418
26419         Add script to convert a patch to a prettypatch and show in the default browser
26420         https://bugs.webkit.org/show_bug.cgi?id=56416
26421
26422         This can be used either like
26423             $ svn-create-patch | show-pretty-diff
26424         or
26425             $ show-pretty-diff patch.diff
26426
26427         * Scripts/show-pretty-diff: Added.
26428
26429 2011-03-15  Brian Weinstein  <bweinstein@apple.com>
26430
26431         Rubber-stamped by Adam Roben.
26432
26433         Adding myself as a reviewer.
26434
26435         * Scripts/webkitpy/common/config/committers.py:
26436
26437 2011-03-15  Anders Carlsson  <andersca@apple.com>
26438
26439         Try to fix the Windows build.
26440
26441         * DumpRenderTree/TestNetscapePlugIn/Tests/NPDeallocateCalledBeforeNPShutdown.cpp:
26442         (NPDeallocateCalledBeforeNPShutdown::TestObject::~TestObject):
26443
26444 2011-03-15  Anders Carlsson  <andersca@apple.com>
26445
26446         Reviewed by Sam Weinig.
26447
26448         Make sure that NP_Shutdown is always the last NPP function called
26449         https://bugs.webkit.org/show_bug.cgi?id=56391
26450
26451         Add a test that times out if an NP_Deallocate object is called after NP_Shutdown.
26452
26453         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
26454         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
26455         (PluginTest::PluginTest):
26456         (PluginTest::NP_Shutdown):
26457         (PluginTest::registerNPShutdownFunction):
26458         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
26459         * DumpRenderTree/TestNetscapePlugIn/Tests/NPDeallocateCalledBeforeNPShutdown.cpp: Added.
26460         (NPDeallocateCalledBeforeNPShutdown::NPDeallocateCalledBeforeNPShutdown):
26461         (NPDeallocateCalledBeforeNPShutdown::TestObject::~TestObject):
26462         (NPDeallocateCalledBeforeNPShutdown::ScriptableObject::hasProperty):
26463         (NPDeallocateCalledBeforeNPShutdown::ScriptableObject::getProperty):
26464         (NPDeallocateCalledBeforeNPShutdown::NPP_New):
26465         (NPDeallocateCalledBeforeNPShutdown::NPP_GetValue):
26466         (NPDeallocateCalledBeforeNPShutdown::shutdown):
26467         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
26468         (NP_Shutdown):
26469         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
26470         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
26471         * GNUmakefile.am:
26472
26473 2011-03-15  Ilya Sherman  <isherman@chromium.org>
26474
26475         Reviewed by Tony Chang.
26476
26477         Autofilled form elements are assigned fixed background color but not text color
26478         https://bugs.webkit.org/show_bug.cgi?id=48382
26479
26480         * DumpRenderTree/LayoutTestController.cpp:
26481         (setAutofilledCallback): Added.
26482         (LayoutTestController::staticFunctions):
26483         * DumpRenderTree/LayoutTestController.h:
26484         * DumpRenderTree/chromium/LayoutTestController.cpp:
26485         (LayoutTestController::LayoutTestController):
26486         (LayoutTestController::setAutofilled): Added.
26487         * DumpRenderTree/chromium/LayoutTestController.h:
26488         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
26489         (LayoutTestController::setAutofilled): Added.
26490         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
26491         (LayoutTestController::setAutofilled): Added.
26492         (LayoutTestController::setValueForUser): Style tweak.
26493         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
26494         (LayoutTestController::setAutofilled): Added (stub).
26495         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
26496         (LayoutTestController::setAutofilled): Added (stub).
26497
26498 2011-03-15  Dan Bernstein  <mitz@apple.com>
26499
26500         Build fix after r81135.
26501
26502         * TestWebKitAPI/Configurations/InjectedBundle.xcconfig:
26503
26504 2011-03-15  Dan Bernstein  <mitz@apple.com>
26505
26506         Build fix after r81135.
26507
26508         * WebKitTestRunner/Configurations/InjectedBundle.xcconfig:
26509         * WebKitTestRunner/mac/TestControllerMac.mm:
26510         * WebKitTestRunner/mac/TestInvocationMac.mm:
26511         * WebKitTestRunner/mac/main.mm:
26512
26513 2011-03-15  Balazs Kelemen  <kbalazs@webkit.org>
26514
26515         Reviewed by Adam Barth.
26516
26517         REGRESSION(80977): breaks Qt ARMv7 Linux Release bot
26518         https://bugs.webkit.org/show_bug.cgi?id=56349
26519
26520         * QtTestBrowser/QtTestBrowser.pro: Link against fontconfig
26521         on platforms where we use it. Copied from DumpRenderTree.pro.
26522
26523 2011-03-15  Dan Bernstein  <mitz@apple.com>
26524
26525         Build fix after r81135.
26526
26527         * DumpRenderTree/cf/WebArchiveDumpSupport.cpp:
26528         * DumpRenderTree/mac/WebArchiveDumpSupportMac.mm:
26529
26530 2011-03-15  Kevin Ollivier  <kevino@theolliviers.com>
26531
26532         Reviewed by Darin Adler.
26533
26534         Introduce WTF_USE_EXPORT_MACROS, which will allow us to put shared library import/export
26535         info into the headers rather than in export symbol definition files, but disable it on
26536         all platforms initially so we can deal with port build issues one port at a time.
26537
26538         https://bugs.webkit.org/show_bug.cgi?id=27551
26539
26540         * DumpRenderTree/cg/ImageDiffCG.cpp:
26541         * DumpRenderTree/chromium/config.h:
26542         * DumpRenderTree/config.h:
26543         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
26544         * TestWebKitAPI/TestWebKitAPIPrefix.h:
26545         * WebKitAPITest/HostWindow.cpp:
26546         * WebKitAPITest/TestsController.cpp:
26547         * WebKitAPITest/config.h: Added.
26548         * WebKitAPITest/main.cpp:
26549         * WebKitAPITest/tests/WebViewDestruction.cpp:
26550         * WebKitTestRunner/InjectedBundle/Bindings/JSWrapper.cpp:
26551         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
26552         * WebKitTestRunner/InjectedBundle/GCController.cpp:
26553         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
26554         * WebKitTestRunner/InjectedBundle/InjectedBundleMain.cpp:
26555         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
26556         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
26557         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
26558         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
26559         * WebKitTestRunner/InjectedBundle/qt/LayoutTestControllerQt.cpp:
26560         * WebKitTestRunner/InjectedBundle/win/ActivateFonts.cpp:
26561         * WebKitTestRunner/InjectedBundle/win/InjectedBundleWin.cpp:
26562         * WebKitTestRunner/InjectedBundle/win/LayoutTestControllerWin.cpp:
26563         * WebKitTestRunner/TestController.cpp:
26564         * WebKitTestRunner/TestInvocation.cpp:
26565         * WebKitTestRunner/config.h: Added.
26566         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
26567         * WebKitTestRunner/qt/TestControllerQt.cpp:
26568         * WebKitTestRunner/qt/TestInvocationQt.cpp:
26569         * WebKitTestRunner/qt/main.cpp:
26570         * WebKitTestRunner/win/PlatformWebViewWin.cpp:
26571         * WebKitTestRunner/win/TestControllerWin.cpp:
26572         * WebKitTestRunner/win/TestInvocationWin.cpp:
26573         * WebKitTestRunner/win/main.cpp:
26574         * wx/build/settings.py:
26575
26576 2011-03-15  Hayato Ito  <hayato@chromium.org>
26577
26578         Reviewed by Shinichiro Hamaji.
26579
26580         [NRWT] Add support for reftests to new-run-webkit-tests.
26581
26582         https://bugs.webkit.org/show_bug.cgi?id=55457
26583
26584         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
26585         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
26586         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
26587         * Scripts/webkitpy/layout_tests/port/base.py:
26588         * Scripts/webkitpy/layout_tests/port/dryrun.py:
26589         * Scripts/webkitpy/layout_tests/port/test.py:
26590         * Scripts/webkitpy/layout_tests/port/test_files.py:
26591         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
26592
26593 2011-03-15  Alejandro G. Castro  <alex@igalia.com>
26594
26595         Reviewed by Martin Robinson.
26596
26597         [GTK] Fix compilation warning after r80892
26598         https://bugs.webkit.org/show_bug.cgi?id=56302
26599
26600         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
26601         (LayoutTestController::originsWithLocalStorage):
26602
26603 2011-03-15  Robert Hogan  <robert@webkit.org>
26604
26605         Reviewed by Adam Barth.
26606
26607         Document the qmakearg argument, and hint how to build webkit2 on Qt
26608         https://bugs.webkit.org/show_bug.cgi?id=56326
26609
26610         * Scripts/build-webkit: add --qmakearg to build-webkit --help
26611
26612 2011-03-15  Adam Roben  <aroben@apple.com>
26613
26614         Make Windows 7 Release (Tests) a core builder
26615
26616         We can probably make Windows XP Debug (Tests) a core builder, too, but we should probably
26617         give it a few more days of greenness first.
26618
26619         Fixes <http://webkit.org/b/55665> Windows 7 Release (Tests) should be a core builder
26620
26621         Reviewed by Sam Weinig.
26622
26623         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
26624         (BuildBot): Added a regexp to match Windows 7 Release (Tests).
26625
26626         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
26627         (BuildBotTest): Updated the list of builders to match the current list on build.webkit.org.
26628         Updated the core regexps to match the buildbot module. Updated the expected results to match
26629         the current set of core builders.
26630
26631 2011-03-14  Adam Roben  <aroben@apple.com>
26632
26633         Add forgotten file from r81116
26634
26635         * TestWebKitAPI/Configurations/TestWebKitAPICFNetwork.vsprops: Added.
26636
26637 2011-03-14  Sam Weinig  <sam@webkit.org>
26638
26639         Reviewed by Adam Roben.
26640
26641         about:blank fake responses don't get serialized when sent the UIProcess
26642         <rdar://problem/9108119>
26643         https://bugs.webkit.org/show_bug.cgi?id=56357
26644
26645         * TestWebKitAPI/Tests/WebKit2/AboutBlankLoad.cpp: Added.
26646         (TestWebKitAPI::decidePolicyForResponse):
26647         (TestWebKitAPI::TEST):
26648         Add test that loads of about:blank have a response with a MIMEType of text/html.
26649
26650         * TestWebKitAPI/PlatformUtilities.h:
26651         * TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
26652         (TestWebKitAPI::Util::MIMETypeForWKURLResponse):
26653         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp:
26654         (TestWebKitAPI::Util::MIMETypeForWKURLResponse):
26655         Add helper to get the MIMEType from a WKURLResponse.
26656
26657         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
26658         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
26659         Add new files.
26660
26661 2011-03-14  Jarkko Sakkinen  <jarkko.j.sakkinen@gmail.com>
26662
26663         Reviewed by Kenneth Rohde Christiansen.
26664
26665         [Qt] Command-line options for QGLWidget and WebGL to QtTestBrowser
26666         https://bugs.webkit.org/show_bug.cgi?id=56006
26667
26668         * QtTestBrowser/launcherwindow.cpp:
26669         (LauncherWindow::initializeView):
26670         * QtTestBrowser/main.cpp:
26671         (LauncherApplication::handleUserOptions):
26672
26673 2011-03-14  James Kozianski  <koz@chromium.org>
26674
26675         Reviewed by Ojan Vafai.
26676
26677         Add method to make BuildBot return test outputs
26678         https://bugs.webkit.org/show_bug.cgi?id=54374
26679
26680         The Builder.latest_cached_build will be used by the new rebaseline
26681         script to determine the build to use baselines from.
26682
26683         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
26684         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
26685
26686 2011-03-14  Adam Roben  <aroben@apple.com>
26687
26688         Test that WKBundlePageCanHandleRequest returns true for empty document URLs
26689
26690         Reviewed by Sam Weinig.
26691
26692         * TestWebKitAPI/Tests/WebKit2/CanHandleRequest.cpp: Added.
26693         (TestWebKitAPI::didReceiveMessageFromInjectedBundle): Store the result of the test.
26694         (TestWebKitAPI::setInjectedBundleClient): Hook up our callback.
26695         (TestWebKitAPI::TEST): Register "emptyscheme" as an empty document scheme, load a page to
26696         ensure the web process is initialized, then ask the bundle to run the test and assert that
26697         it succeeded.
26698
26699         * TestWebKitAPI/Tests/WebKit2/CanHandleRequest_Bundle.cpp: Added.
26700         (TestWebKitAPI::CanHandleRequestTest::CanHandleRequestTest): Just call up to the base class.
26701         (TestWebKitAPI::canHandleURL): Helper function to test whether WebKit2 claims to be able to
26702         handle a given URL.
26703         (TestWebKitAPI::runTest): Check that empty document URLs can be handled (and that an unknown
26704         URL type cannot be handled).
26705         (TestWebKitAPI::CanHandleRequestTest::didReceiveMessage): Run the test and return the
26706         result.
26707
26708         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
26709         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
26710         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
26711         Added new files to the project.
26712
26713 2011-03-14  Sheriff Bot  <webkit.review.bot@gmail.com>
26714
26715         Unreviewed, rolling out r81094.
26716         http://trac.webkit.org/changeset/81094
26717         https://bugs.webkit.org/show_bug.cgi?id=56355
26718
26719         Broke the chromium DRT related build. (Requested by dave_levin
26720         on #webkit).
26721
26722         * DumpRenderTree/chromium/DumpRenderTree.cpp:
26723         (main):
26724         * DumpRenderTree/chromium/LayoutTestController.cpp:
26725         (LayoutTestController::LayoutTestController):
26726         (LayoutTestController::layerTreeAsText):
26727         * DumpRenderTree/chromium/LayoutTestController.h:
26728         * DumpRenderTree/chromium/TestShell.cpp:
26729         (TestShell::runFileTest):
26730         (TestShell::dump):
26731         * DumpRenderTree/chromium/TestShell.h:
26732         (TestParams::TestParams):
26733
26734 2011-03-14  Daniel Sievers  <sievers@google.com>
26735
26736         Reviewed by Simon Fraser.
26737
26738         [Chromium] Make RenderAsTextBehavior and LayerTreeAsTextBehavior tweakable from the DumpRenderTree commandline
26739         https://bugs.webkit.org/show_bug.cgi?id=56139
26740
26741         * DumpRenderTree/chromium/DumpRenderTree.cpp:
26742         (main):
26743         * DumpRenderTree/chromium/LayoutTestController.cpp:
26744         (LayoutTestController::LayoutTestController):
26745         (LayoutTestController::layerTreeAsText):
26746         * DumpRenderTree/chromium/LayoutTestController.h:
26747         (LayoutTestController::setShowDebugLayerTree):
26748         * DumpRenderTree/chromium/TestShell.cpp:
26749         (TestShell::runFileTest):
26750         (TestShell::dump):
26751         * DumpRenderTree/chromium/TestShell.h:
26752         (TestParams::TestParams):
26753
26754 2011-03-14  Balazs Kelemen  <kbalazs@webkit.org>
26755
26756         Reviewed by Adam Roben.
26757
26758         [Qt][WK2]Unbreak InjectedBundle on Qt
26759         https://bugs.webkit.org/show_bug.cgi?id=54109
26760
26761         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
26762         (WTR::InjectedBundlePage::willSendRequestForFrame): Use new WebKit2
26763         API's instead of KURL to analyze the URL.
26764         * WebKitTestRunner/StringFunctions.h:
26765         (WTR::adoptWK): Moved from InjectedBundlePage.cpp to make it available globally
26766         in the WebKitTestRunner project.
26767         (WTR::toWK): Use adoptWK.
26768         (WTR::operator<<): Added helper for printing a WKURL.
26769         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj: Do not link
26770         against WebCore.
26771
26772 2011-03-14  Balazs Kelemen  <kbalazs@webkit.org>
26773
26774         Reviewed by Kenneth Rohde Christiansen.
26775
26776         check-webkit-style falsely should not complain about WebKit2 API headers in the Shared directory
26777         https://bugs.webkit.org/show_bug.cgi?id=56321
26778
26779         * Scripts/webkitpy/style/checker.py: Add the patch "Source/WebKit2/Shared/API/c"
26780         to the list of exceptions for WebKit2 C API headers.
26781
26782 2011-03-14  Adam Roben  <aroben@apple.com>
26783
26784         Show the list of recent leaky builds in reverse chronological order
26785
26786         Previously they were being shown in a deterministic but surprising order based on
26787         lexicographically sorting the build offsets (-1, -10, -2, -3...).
26788
26789         Fixes <http://webkit.org/b/56342> Leaks Viewer shows recent builds in a strange order
26790
26791         Reviewed by Darin Adler.
26792
26793         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/RecentBuildsLoader.js:
26794         (RecentBuildsLoader.prototype.start): Sort the builds by revision before passing them to our
26795         callback.
26796
26797 2011-03-14  Adam Roben  <aroben@apple.com>
26798
26799         Show a list of up to 10 recent leaky builds when Leaks Viewer loads
26800
26801         Fixes <http://webkit.org/b/56043> Leaks Viewer should present a list of recent builds to
26802         analyze
26803
26804         Reviewed by David Kilzer and Joe Pecoraro.
26805
26806         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css:
26807         (#url-prompt): Made this a bit taller to make room for the list of builds.
26808
26809         (#recent-builds-loading-indicator):
26810         (#recent-builds-list):
26811         Added styles for the build list and loading indicator.
26812
26813         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.js:
26814         (LeaksViewer.loaded): Pass the URL from the query through decodeURIComponent before loading
26815         it. This will unescape it.
26816         (LeaksViewer.urlPromptButtonClicked): Moved code from here to _urlChosenFromPrompt.
26817         (LeaksViewer._didLoadRecentBuilds): Added. Stores the builds and updates the prompt.
26818         (LeaksViewer._displayURLPrompt): Start loading the recent leaky builds.
26819         (LeaksViewer._updateURLPrompt): Create a list of the recent leaky builds and add them to the
26820         prompt UI.
26821         (LeaksViewer._urlChosenFromPrompt): Added. Code came from urlPromptButtonClicked. This is
26822         now also used when a recent build is clicked.
26823
26824         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/RecentBuildsLoader.js: Added.
26825         (RecentBuildsLoader): Store the callback.
26826         (RecentBuildsLoader.prototype.start): Load the N most recent builds from the specified
26827         builder. Find the revision number, leak count, and results URL for each one. Pass the
26828         resulting info to our callback.
26829
26830         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Utilities.js:
26831         (range): Added this helper function to return an array containing the integers [0, n).
26832
26833         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html: Added
26834         RecentBuildsLoader.js and added text to the prompt UI to show the recent builds.
26835
26836 2011-03-14  Mark Rowe  <mrowe@apple.com>
26837
26838         Reviewed by Adam Roben.
26839
26840         Fix two memory leaks within LayoutTestController's storage support.
26841
26842         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
26843         (LayoutTestController::originsWithLocalStorage): Don't leak the JSStringRef that is returned
26844         by JSStringCreateWithCFString.
26845         (LayoutTestController::deleteLocalStorageForOrigin): Don't leak the WebSecurityOrigin that we
26846         allocate.
26847
26848 2011-03-14  Mark Rowe  <mrowe@apple.com>
26849
26850         Reviewed by Oliver Hunt.
26851
26852         <http://webkit.org/b/56304> REGRESSION(r80892): 100,000+ leaks seen on the build bot
26853
26854         * DumpRenderTree/LayoutTestController.cpp:
26855         (LayoutTestController::staticFunctions): Remove duplicate entries.
26856
26857 2011-03-14  Alice Liu  <alice.liu@apple.com>
26858
26859         Reviewed by Dan Bernstein.
26860
26861         DumpRenderTree Perl Support fails to build with ARCHS="i386 x86_64"
26862
26863         * DumpRenderTree/mac/PerlSupport/Makefile:
26864         Add whitespace when joining "-arch" with the architectures passed in as arguments.
26865
26866 2011-03-14  Daniel Bates  <dbates@rim.com>
26867
26868         https://bugs.webkit.org/show_bug.cgi?id=55438
26869         Add double quotes around value of -DCMAKE_WINCE_SDK that I inadvertently did not add
26870         when landing changeset 81041 <http://trac.webkit.org/changeset/81041>.
26871
26872         * Scripts/build-webkit:
26873
26874 2011-03-14  Daniel Bates  <dbates@rim.com>
26875
26876         Reviewed by David Kilzer.
26877
26878         Cleanup: Separate port-specific implementation details from webkitdirs::buildCMakeProject()
26879         https://bugs.webkit.org/show_bug.cgi?id=55438
26880
26881         Separate out the EFL- and WinCE-specific logic from the port-independent logic in
26882         webkitdirs::buildCMakeProject(). Also, remove redundant code for generating the
26883         CMake arguments for feature defines.
26884
26885         * Scripts/build-webkit:
26886           - Added cMakeArgsFromFeatures() to convert the @features array to CMake arguments.
26887           - Modified EFL and WinCE-specific building logic to call buildCMakeProjectOrExit().
26888           - Sorted forward declarations.
26889           - Initialize $makeArgs with the empty string so as to simplify its use in string operations;
26890             Modified call sites as needed.
26891         * Scripts/webkitdirs.pm:
26892           - Added cleanCMakeGeneratedProject() to clean a CMake build.
26893           - Added buildCMakeGeneratedProject() to build using generated build system.
26894           - Added buildCMakeProjectOrExit() which is a facade for building a CMake project.
26895           - Added generateBuildSystemFromCMakeProject() to generate the build system from
26896             a CMake project.
26897           - Removed buildCMakeProject(). This functionality is in buildCMakeProjectOrExit().
26898
26899 2011-03-14  John Knottenbelt  <jknotten@chromium.org>
26900
26901         Reviewed by Steve Block.
26902
26903         Detach Geolocation from Frame when Page destroyed.
26904         https://bugs.webkit.org/show_bug.cgi?id=52877
26905
26906         Extend the layout test controller to expose the number of pending
26907         geolocation requests, so that we can test that the requests have
26908         been cancelled on page close.
26909
26910         * DumpRenderTree/LayoutTestController.cpp:
26911         (numberOfPendingGeolocationPermissionRequestsCallback):
26912         (LayoutTestController::staticFunctions):
26913         * DumpRenderTree/LayoutTestController.h:
26914         * DumpRenderTree/chromium/LayoutTestController.cpp:
26915         (LayoutTestController::LayoutTestController):
26916         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
26917         * DumpRenderTree/chromium/LayoutTestController.h:
26918         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
26919         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
26920         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
26921         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
26922         * DumpRenderTree/mac/UIDelegate.h:
26923         * DumpRenderTree/mac/UIDelegate.mm:
26924         (-[UIDelegate numberOfPendingGeolocationPermissionRequests]):
26925         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
26926         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
26927         * DumpRenderTree/qt/LayoutTestControllerQt.h:
26928         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
26929         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
26930         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
26931         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
26932
26933
26934 2011-03-14  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
26935
26936         Unreviewed. Modify my email address in committers.py.
26937
26938         * Scripts/webkitpy/common/config/committers.py:
26939
26940 2011-03-13  Joe Wild  <joseph.wild@nokia.com>
26941
26942         Reviewed by Kenneth Rohde Christiansen.
26943
26944         [Qt] QtLauncher does not load the same set of fonts as the DRT
26945         https://bugs.webkit.org/show_bug.cgi?id=34959
26946
26947         This patch adds the option "-use-test-fonts" to the QtTestBrowser.
26948         When this option is used the webkit fonts are loaded the same
26949         as they are in DumpRenderTree.  This option can be used on
26950         QtTestBrowser and run-launcher.  It can only be used
26951         on Linux systems with FcInit and is configured as such.
26952
26953         * QtTestBrowser/launcherwindow.h:
26954         (WindowOptions::WindowOptions):
26955         * QtTestBrowser/main.cpp:
26956         (initWebKitTestFonts):
26957         (launcherMain):
26958         (LauncherApplication::handleUserOptions):
26959
26960 2011-03-12  Dan Bernstein  <mitz@apple.com>
26961
26962         Reviewed by Mark Rowe.
26963
26964         <rdar://problem/8995822> Build DumpRenderTree perl support only for the architectures
26965         DumpRenderTree is being built for.
26966
26967         * DumpRenderTree/mac/PerlSupport/Makefile:
26968
26969 2011-03-12  Dan Bernstein  <mitz@apple.com>
26970
26971         Suggested by Mark Rowe.
26972
26973         * DumpRenderTree/mac/Configurations/Base.xcconfig: Leave C++ exceptions enabled
26974         as they were before r80915.
26975
26976 2011-03-12  Anders Carlsson  <andersca@apple.com>
26977
26978         Reviewed by Sam Weinig.
26979
26980         Should be able to download a PDF to the Downloads folder
26981         https://bugs.webkit.org/show_bug.cgi?id=56256
26982
26983         * MiniBrowser/mac/BrowserWindowController.m:
26984         (-[BrowserWindowController awakeFromNib]):
26985         * WebKitTestRunner/TestController.cpp:
26986         (WTR::TestController::createOtherPage):
26987         (WTR::TestController::initialize):
26988         Update for API changes.
26989
26990 2011-03-11  Anton D'Auria  <adauria@apple.com>
26991
26992         Reviewed and landed by Brady Eidson.
26993
26994         https://bugs.webkit.org/show_bug.cgi?id=51878
26995         Add WebKit1 API to view and delete local storage
26996
26997         Added tests that write to LocalStorage, delete one origin, get list of origins with local storage, delete all origins.
26998
26999         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
27000         * DumpRenderTree/LayoutTestController.cpp:
27001         (syncLocalStorageCallback):
27002         (observeStorageTrackerNotificationsCallback):
27003         (deleteAllLocalStorageCallback):
27004         (deleteLocalStorageForOriginCallback):
27005         (originsWithLocalStorageCallback):
27006         (LayoutTestController::staticFunctions):
27007         * DumpRenderTree/LayoutTestController.h:
27008         * DumpRenderTree/StorageTrackerDelegate.h: Added.
27009         * DumpRenderTree/StorageTrackerDelegate.mm: Added.
27010         (-[StorageTrackerDelegate init]):
27011         (-[StorageTrackerDelegate logNotifications:controller:]):
27012         (-[StorageTrackerDelegate originModified:]):
27013         (-[StorageTrackerDelegate dealloc]):
27014         (-[StorageTrackerDelegate setControllerToNotifyDone:]):
27015         * DumpRenderTree/chromium/LayoutTestController.cpp:
27016         (LayoutTestController::LayoutTestController):
27017         (LayoutTestController::deleteAllLocalStorage):
27018         (LayoutTestController::originsWithLocalStorage):
27019         (LayoutTestController::deleteLocalStorageForOrigin):
27020         (observeStorageTrackerNotifications):
27021         (syncLocalStorage):
27022         * DumpRenderTree/chromium/LayoutTestController.h:
27023         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
27024         (LayoutTestController::originsWithLocalStorage):
27025         (LayoutTestController::deleteAllLocalStorage):
27026         (LayoutTestController::deleteLocalStorageForOrigin):
27027         (LayoutTestController::observeStorageTrackerNotifications):
27028         (LayoutTestController::syncLocalStorage):
27029         * DumpRenderTree/mac/DumpRenderTree.mm:
27030         (resetDefaultsToConsistentValues):
27031         (allocateGlobalControllers):
27032         (releaseGlobalControllers):
27033         * DumpRenderTree/mac/DumpRenderTreeMac.h:
27034         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
27035         (LayoutTestController::syncLocalStorage):
27036         (LayoutTestController::observeStorageTrackerNotifications):
27037         (LayoutTestController::deleteAllLocalStorage):
27038         (LayoutTestController::originsWithLocalStorage):
27039         (LayoutTestController::deleteLocalStorageForOrigin):
27040         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
27041         (LayoutTestController::originsWithLocalStorage):
27042         (LayoutTestController::deleteAllLocalStorage):
27043         (LayoutTestController::deleteLocalStorageForOrigin):
27044         (LayoutTestController::observeStorageTrackerNotifications):
27045         (LayoutTestController::syncLocalStorage):
27046         * DumpRenderTree/qt/LayoutTestControllerQt.h:
27047         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
27048         (LayoutTestController::clearAllApplicationCaches):
27049         (LayoutTestController::syncLocalStorage):
27050         (LayoutTestController::observeStorageTrackerNotifications):
27051         (LayoutTestController::clearAllDatabases):
27052         (LayoutTestController::deleteAllLocalStorage):
27053         (LayoutTestController::originsWithLocalStorage):
27054         (LayoutTestController::deleteLocalStorageForOrigin):
27055         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
27056         (LayoutTestController::syncLocalStorage):
27057         (LayoutTestController::observeStorageTrackerNotifications):
27058         (LayoutTestController::clearAllDatabases):
27059         (LayoutTestController::deleteAllLocalStorage):
27060         (LayoutTestController::originsWithLocalStorage):
27061         (LayoutTestController::deleteLocalStorageForOrigin):
27062
27063 2011-03-11  Jessie Berlin  <jberlin@apple.com>
27064
27065         Reviewed by Adam Roben.
27066
27067         Crash calling WebContext::clearResourceCaches(InMemoryResourceCachesOnly) before the Web
27068         Process has finished launching.
27069         https://bugs.webkit.org/show_bug.cgi?id=56208
27070
27071         * TestWebKitAPI/Tests/WebKit2/SendingMessagesToTheWebProcessBeforeItIsValid.cpp: Added.
27072         (TestWebKitAPI::TEST):
27073         Test that calling WKContextClearResourceCaches works with both options when called before
27074         the Web Process is valid.
27075
27076         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
27077         Add the new test.
27078         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
27079         Ditto.
27080
27081 2011-03-11  Adam Roben  <aroben@apple.com>
27082
27083         Make it possible to view all leaks from a build at once in Leaks Viewer
27084
27085         You can now give Leaks Viewer the URL of a build results page (e.g.,
27086         <http://build.webkit.org/results/SnowLeopard%20Intel%20Leaks/r80847%20(15493)/>), and it
27087         will load all the leaks files from that build.
27088
27089         Fixes <http://webkit.org/b/56030> Leaks Viewer: Would like to be able to look at all leaks
27090         files from a particular build at once, rather than one at a time
27091
27092         Reviewed by David Kilzer.
27093
27094         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksLoader.js: Added.
27095         (LeaksLoader): Initialize our properties.
27096         (LeaksLoader.prototype.start): If the URL ends in .txt, assume it's a single leaks file and
27097         load it. Otherwise assume it's a build results page URL and start fetching the leaks files
27098         from that page.
27099         (LeaksLoader.prototype._loadLeaksFiles): Tell our client how many leaks files we're loading
27100         so it can provide feedback. Then load each one and pass it to the client.
27101         (LeaksLoader.prototype._loadLeaksFromResultsPage): Load the HTML of the results page, parse
27102         it into a DOM, pull out all the links to leaks files, and load each one.
27103
27104         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParser.js: Added.
27105         (LeaksParser): Initialize our worker.
27106         (LeaksParser.prototype.addLeaksFile): Tell the worker about the leaks file's contents.
27107
27108         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js: Added.
27109         This code mostly came from Worker.js.
27110         (LeaksParserWorker): Initialize our profile.
27111         (LeaksParserWorker.prototype.addLeaksFile): Parse the leaks file and incorporate it into our
27112         profile.
27113         (LeaksParserWorker.prototype._parseLeaks):
27114         (LeaksParserWorker.prototype._createNode):
27115         (LeaksParserWorker.prototype._incorporateLeaks):
27116         Moved these functions here from LeaksViewer. _incorporateLeaks is essentially just a renamed
27117         version of createProfile which adds to this.profile instead of making a new profile each
27118         time.
27119         (onmessage): Parse the file and send back the new profile.
27120
27121         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css:
27122         (#loading-indicator):
27123         (#spinner):
27124         (#loading-indicator-label):
27125         Tweaked styles to accomodate a longer label.
27126
27127         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.js:
27128         (LeaksViewer.loaded): Set up a loader and parser.
27129
27130         (LeaksViewer.get filesLeftToParse):
27131         (LeaksViewer.set filesLeftToParse):
27132         Added these simple accessors.
27133
27134         (LeaksViewer._didCountLeaksFiles): Callback from LeaksLoader. Stores the count.
27135         (LeaksViewer._didLoadLeaksFile): Callback from LeaksLoader. Passes the file contents off to
27136         the parser.
27137         (LeaksViewer._didParseLeaksFile): Callback from LeaksParser. If all files have been parsed,
27138         tell the ProfilerAgent and mark that we're done loading. (Code came from the old
27139         _loadLeaksFromURL function).
27140         (LeaksViewer._loadLeaksFromURL): Now just calls through to the loader.
27141         (LeaksViewer._loadingIndicatorText): Added. Returns the text that should show up in the
27142         loading indicator, including the number of files being loaded.
27143         (LeaksViewer._loadingStatusChanged): Update the loading indicator's label, too.
27144         (LeaksViewer._updateLoadingIndicatorLabel): Added. Just updates the label!
27145         (LeaksViewer._updateTitle): Moved code to compute the "Loading" text to
27146         _loadingIndicatorText.
27147
27148         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Utilities.js: Added.
27149         (getResource): Moved here from LeaksViewer.js.
27150
27151         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Worker.js: Subsumed by
27152         LeaksParserWorker.js.
27153
27154         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html: Added new JS
27155         files, added a #loading-indicator-label element, and tweaked the prompt wording.
27156
27157 2011-03-11  Adam Roben  <aroben@apple.com>
27158
27159         Don't trigger a build when build.webkit.org's HTML files are modified
27160
27161         Fixes <http://webkit.org/b/56190> Windows bots shouldn't have built r80848/r80851, but did
27162
27163         Reviewed by Dan Bernstein.
27164
27165         * Scripts/webkitpy/common/config/build.py:
27166         (_should_file_trigger_build): Added build.webkit.org's public_html directory to the list of
27167         directories that shouldn't trigger builds. Re-sorted the list.
27168
27169         * Scripts/webkitpy/common/config/build_unittest.py:
27170         (ShouldBuildTest): Added a test to cover the above change.
27171
27172 2011-03-11  Adam Roben  <aroben@apple.com>
27173
27174         Roll out r80848
27175
27176         It was accidentally committed.
27177
27178         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.js:
27179         (LeaksViewer._loadLeaksFromURL):
27180
27181 2011-03-11  Tony Chang  <tony@chromium.org>
27182
27183         Reviewed by Ojan Vafai.
27184
27185         kill hung lighttpd processes on the chromium win bot
27186         https://bugs.webkit.org/show_bug.cgi?id=56152
27187
27188         Sometimes this process doesn't exit and it holds on to
27189         a log file that we can't delete or copy, causing the tests
27190         to go red.
27191
27192         * BuildSlaveSupport/win/kill-old-processes:
27193
27194 2011-03-11  Ilya Sherman  <isherman@chromium.org>
27195
27196         Reviewed by Kent Tamura.
27197
27198         [Chromium] missing HTMLInputElement::setValueForUser() to tests
27199         https://bugs.webkit.org/show_bug.cgi?id=55854
27200
27201         * DumpRenderTree/chromium/LayoutTestController.cpp:
27202         (LayoutTestController::LayoutTestController):
27203         (LayoutTestController::setValueForUser): Added.
27204         * DumpRenderTree/chromium/LayoutTestController.h:
27205
27206 2011-03-10  Peter Kasting  <pkasting@google.com>
27207
27208         Reviewed by Dimitri Glazkov.
27209
27210         Clean up some gross code in TestShellWin.cpp.  No functional change.
27211         https://bugs.webkit.org/show_bug.cgi?id=56048
27212
27213         * DumpRenderTree/chromium/TestShellWin.cpp:
27214         (checkLayoutTestSystemDependencies):
27215
27216 2011-03-10  Keith Kyzivat  <keith.kyzivat@nokia.com>
27217
27218         Reviewed by Laszlo Gombos.
27219
27220         [Qt] Buildfix for platforms where X11 is not available (e.g. Symbian)
27221
27222         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
27223         (WTR::activateFonts): Guard X11-specific code with Q_WS_X11.
27224
27225 2011-03-10  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
27226
27227         Reviewed by Kenneth Rohde Christiansen.
27228
27229         [EFL] Make single backing store the default for EWebLauncher
27230         https://bugs.webkit.org/show_bug.cgi?id=55753
27231
27232         Set single backing store the default for EWebLauncher.
27233
27234         * EWebLauncher/main.c:
27235         (main):
27236
27237 2011-03-10  Adam Roben  <aroben@apple.com>
27238
27239         Ignore errors from files being held open when creating a test result archive
27240
27241         Fixes <http://webkit.org/b/55581> Test results don't get uploaded if test-result-archive
27242         fails to delete the results directory
27243
27244         Reviewed by Mark Rowe.
27245
27246         * BuildSlaveSupport/test-result-archive:
27247         (archiveTestResults): Ignore errno 2 when deleting the results directory, since it can be
27248         thrown when a process is holding a file open.
27249
27250 2011-03-10  Sheriff Bot  <webkit.review.bot@gmail.com>
27251
27252         Unreviewed, rolling out r80701.
27253         http://trac.webkit.org/changeset/80701
27254         https://bugs.webkit.org/show_bug.cgi?id=56126
27255
27256         fast/dom/Geolocation/window-close-crash.html fails on Snow
27257         Leopard release builds (Requested by mihaip on #webkit).
27258
27259         * DumpRenderTree/LayoutTestController.cpp:
27260         (LayoutTestController::staticFunctions):
27261         * DumpRenderTree/LayoutTestController.h:
27262         * DumpRenderTree/chromium/LayoutTestController.cpp:
27263         (LayoutTestController::LayoutTestController):
27264         * DumpRenderTree/chromium/LayoutTestController.h:
27265         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
27266         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
27267         * DumpRenderTree/mac/UIDelegate.h:
27268         * DumpRenderTree/mac/UIDelegate.mm:
27269         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
27270         * DumpRenderTree/qt/LayoutTestControllerQt.h:
27271         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
27272         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
27273
27274 2011-03-10  Philippe Normand  <pnormand@igalia.com>
27275
27276         Reviewed by Gustavo Noronha Silva.
27277
27278         [GTK] run-gtk-tests doesn't fail if a test crashes
27279         https://bugs.webkit.org/show_bug.cgi?id=56089
27280
27281         * Scripts/run-gtk-tests: gtester -k exits with a 0 exit status
27282         even if a test crashes. So run the tests individually and if a
27283         test fails remember its exit code so run-gtk-test fails too, if
27284         required.
27285
27286 2011-01-26  John Knottenbelt  <jknotten@chromium.org>
27287
27288         Reviewed by Dmitry Titov.
27289
27290         Detach Geolocation from Frame when Page destroyed.
27291         https://bugs.webkit.org/show_bug.cgi?id=52877
27292
27293         Extend the layout test controller to expose the number of pending
27294         geolocation requests, so that we can test that the requests have
27295         been cancelled on page close.
27296
27297         * DumpRenderTree/LayoutTestController.cpp:
27298         (numberOfPendingGeolocationPermissionRequestsCallback):
27299         (LayoutTestController::staticFunctions):
27300         * DumpRenderTree/LayoutTestController.h:
27301         * DumpRenderTree/chromium/LayoutTestController.cpp:
27302         (LayoutTestController::LayoutTestController):
27303         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
27304         * DumpRenderTree/chromium/LayoutTestController.h:
27305         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
27306         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
27307         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
27308         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
27309         * DumpRenderTree/mac/UIDelegate.h:
27310         * DumpRenderTree/mac/UIDelegate.mm:
27311         (-[UIDelegate numberOfPendingGeolocationPermissionRequests]):
27312         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
27313         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
27314         * DumpRenderTree/qt/LayoutTestControllerQt.h:
27315         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
27316         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
27317         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
27318         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
27319
27320
27321 2011-03-09  Adam Roben  <aroben@apple.com>
27322
27323         Hide Leaks Viewer's URL prompt by default
27324
27325         We will show it as needed. I think I left it visible by default by accident after doing some
27326         testing.
27327
27328         Fixes <http://webkit.org/b/56031> Leaks Viewer: URL prompt overlay shouldn't appear when a
27329         URL is passed as a query parameter, but does
27330
27331         Reviewed by Simon Fraser.
27332
27333         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html:
27334
27335 2011-03-09  Adam Roben  <aroben@apple.com>
27336
27337         Don't use Element.classList or box-shadow in Leaks Viewer
27338
27339         They aren't supported by Safari 5. Luckily we have substitutes for them!
27340
27341         Fixes <http://webkit.org/b/56035> Leaks viewer doesn't work in Safari 5
27342
27343         Reviewed by Simon Fraser.
27344
27345         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css:
27346         (#url-prompt): Use -webkit-box-shadow instead of box-shadow.
27347
27348         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.js:
27349         (LeaksViewer.urlPromptButtonClicked):
27350         (LeaksViewer._displayURLPrompt):
27351         (LeaksViewer._setLoadingIndicatorHidden):
27352         Use add/removeStyleClass (from utilities.js) instead of Element.classList.
27353
27354 2011-03-09  Adam Roben  <aroben@apple.com>
27355
27356         Add a link to the Leaks Viewer tool on the build.webkit.org homepage
27357
27358         Rubber-stamped by David Kilzer.
27359
27360         * BuildSlaveSupport/build.webkit.org-config/templates/root.html:
27361
27362 2011-03-09  Adam Roben  <aroben@apple.com>
27363
27364         Add a Leaks Viewer tool to build.webkit.org
27365
27366         This tool parses leaks from the files generated by the leaks bot and displays them in a
27367         view similar to the Web Inspector's Profiles pane (in fact, it uses the Web Inspector's
27368         code).
27369
27370         The URL of the leaks file can be entered in a prompt when the tools is first shown, or can
27371         be passed to the tool via the "url" query parameter.
27372
27373         Only one leaks file can be viewed at a time for now. Only the number of leaks, as opposed to
27374         the number of leaked bytes, is shown. And this tool has made apparent some bugs in the Web
27375         Inspector's Heavy profile view. But it's still pretty useful as-is.
27376
27377         Fixes <http://webkit.org/b/56028> Would like a Shark-like tool for viewing leaks from leak bots
27378
27379         Reviewed by Darin Adler.
27380
27381         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css: Added.
27382         (#url-prompt-container): This is the dark overlay that is shown behind the URL prompt.
27383         (#url-prompt): This is the prompt itself.
27384
27385         (#loading-indicator):
27386         (#loading-indicator > img):
27387         These show a spinner in the bottom-right when we're loading a leaks file.
27388
27389         (.percent-time-status-bar-item): Hide the Percent button since we never want to show leaks
27390         as percentages.
27391
27392         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.js: Added.
27393         (LeaksViewer.loaded): Initialize ourselves and start loading the leaks file (or prompt for a
27394         URL if one wasn't given to us as a query parameter).
27395
27396         (LeaksViewer.get loading):
27397         (LeaksViewer.set loading):
27398         (LeaksViewer.get url):
27399         (LeaksViewer.set url):
27400         Pretty simple accessors.
27401
27402         (LeaksViewer.urlPromptButtonClicked): Start loading the URL from the prompt and hide the
27403         prompt.
27404         (LeaksViewer._displayURLPrompt): Show the prompt and focus the URL input.
27405         (LeaksViewer._loadLeaksFromURL): Start fetching the leaks file. When we've retrieved the
27406         file, hand it off to a Web Worker that will parse it into a fake "profile", then hand off
27407         the profile to the Web Inspector code once it's been created.
27408         (LeaksViewer._loadingStatusChanged): Update our loading indicator and title.
27409         (LeaksViewer._setLoadingIndicatorHidden): Simply hide or show the indicator.
27410         (LeaksViewer._updateTitle): Set our title based on our URL and loading status.
27411         (getResource): Helper function to fetch a resource via XHR and pass it to a callback.
27412
27413         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/WebInspectorShims.js: Added.
27414         (WebInspector.UIString): Just a simple implementation since this tool isn't localized.
27415         (ProfilerAgent.getProfile): Store the callback.
27416         (ProfilerAgent.profileReady): Call the callback.
27417         (monkeyPatchInspectorObjects): Fix up WebInspector.ProfileDataGridNode to show the self leak
27418         count and total leak counts as real values, not percentages or milliseconds.
27419
27420         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Worker.js: Added.
27421         (onmessage): Create a profile based on the leaks file and pass it back to our caller.
27422         (parseLeaks): Extract call stacks out of a leaks file.
27423         (createNode): Helper function to create a new empty profile node.
27424         (createProfile): Take the leak stacks and turn them into a fake "profile" for consumption by
27425         WebInspector.ProfileView.
27426
27427         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html: Added.
27428
27429 2011-03-09  Dan Bernstein  <mitz@apple.com>
27430
27431         Reviewed by Adam Roben.
27432
27433         Combine multiple --ignore-tests values
27434         https://bugs.webkit.org/show_bug.cgi?id=56016
27435
27436         * Scripts/old-run-webkit-tests:
27437
27438 2011-03-08  Kenji Imasaki  <imasaki@chromium.org>
27439
27440         Reviewed by David Levin.
27441
27442         Adding myself to the committer file.
27443
27444         * Scripts/webkitpy/common/config/committers.py:
27445
27446 2011-03-08  Jessie Berlin  <jberlin@apple.com>
27447
27448         Rubber-stamped by Adam Roben.
27449
27450         Fix two leaks I introduced into the API tests.
27451
27452         * TestWebKitAPI/Tests/WebKit2/win/DoNotCopyANullCFURLResponse.cpp:
27453         (TestWebKitAPI::TEST):
27454
27455 2011-03-08  Jessie Berlin  <jberlin@apple.com>
27456
27457         Reviewed by Anders Carlsson.
27458
27459         Crash in CFNetwork visiting google.com
27460         https://bugs.webkit.org/show_bug.cgi?id=55958
27461
27462         Add a test for not trying to copy a null CFURLResponse.
27463
27464         * TestWebKitAPI/Tests/WebKit2/win/DoNotCopyANullCFURLResponse.cpp: Added.
27465         (TestWebKitAPI::TEST):
27466         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
27467
27468 2011-03-08  Hayato Ito  <hayato@chromium.org>
27469
27470         Reviewed by Ojan Vafai.
27471
27472         [NRWT] Renamed a FailureWithType class to a ComparisonTestFailure class.
27473
27474         FailureWithType is a confusing naming because we have removed test_type/* classes in r79840.
27475
27476         https://bugs.webkit.org/show_bug.cgi?id=55450
27477
27478         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
27479
27480 2011-03-08  Carlos Garcia Campos  <cgarcia@igalia.com>
27481
27482         Reviewed by Martin Robinson.
27483
27484         [GTK] Add layoutTestController.setValueForUser() after r80412
27485         https://bugs.webkit.org/show_bug.cgi?id=55862
27486
27487         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
27488         (LayoutTestController::setValueForUser): Implement setValueForUser()
27489
27490 2011-03-08  Patrick Gansterer  <paroga@webkit.org>
27491
27492         Reviewed by Adam Barth.
27493
27494         Add script_shell_command to WebKitPort
27495         https://bugs.webkit.org/show_bug.cgi?id=55925
27496
27497         Add a central place where we generate the shell command from the script name.
27498         This function will be used later to detect the correct script interpreter on Win32.
27499
27500         * Scripts/webkitpy/common/config/ports.py:
27501         * Scripts/webkitpy/layout_tests/port/base.py:
27502         * Scripts/webkitpy/tool/steps/abstractstep.py:
27503         * Scripts/webkitpy/tool/steps/preparechangelog.py:
27504
27505 2011-03-07  Ojan Vafai  <ojan@chromium.org>
27506
27507         Reviewed by Adam Barth.
27508
27509         add webkit-patch attach-to-bug for posting to bugzilla from the commandline
27510         https://bugs.webkit.org/show_bug.cgi?id=55749
27511
27512         * Scripts/webkitpy/tool/commands/upload.py:
27513         * Scripts/webkitpy/tool/commands/upload_unittest.py:
27514         * Scripts/webkitpy/tool/steps/__init__.py:
27515         * Scripts/webkitpy/tool/steps/attachtobug.py: Added.
27516         * Scripts/webkitpy/tool/steps/options.py:
27517
27518 2011-03-07  Daniel Cheng  <dcheng@chromium.org>
27519
27520         Reviewed by David Levin.
27521
27522         Add support for DataTransferItems
27523         https://bugs.webkit.org/show_bug.cgi?id=55115
27524
27525         Enable data transfer items in DRT.
27526
27527         * DumpRenderTree/chromium/TestShell.cpp:
27528         (TestShell::TestShell):
27529
27530 2011-03-06  Ojan Vafai  <ojan@chromium.org>
27531
27532         Reviewed by Tony Chang.
27533
27534         [test dashboard] pull treemap files
27535         https://bugs.webkit.org/show_bug.cgi?id=55852
27536
27537         * TestResultServer/handlers/dashboardhandler.py:
27538
27539 2011-03-03  Hayato Ito  <hayato@chromium.org>
27540
27541         Reviewed by Mihai Parparita.
27542
27543         [NRWT] Clear output image file used by DumpRenderTree to make sure the
27544         previous image should not be used in the ChromiumDriver.
27545
27546         Also make sure that output image from the ChromiumDriver should be ''
27547         (empty string) if a test crashes for consistency with the WebKitDriver
27548         implementation.
27549
27550         https://bugs.webkit.org/show_bug.cgi?id=55746
27551
27552         * Scripts/webkitpy/layout_tests/port/chromium.py:
27553
27554 2011-03-07  Zan Dobersek  <zandobersek@gmail.com>
27555
27556         Reviewed by Martin Robinson.
27557
27558         [Gtk] fast/js/navigator-language.html fails locally
27559         https://bugs.webkit.org/show_bug.cgi?id=55880
27560
27561         Add LANG variable to the clean environment when running
27562         old-run-webkit-tests script.
27563
27564         * Scripts/old-run-webkit-tests:
27565
27566 2011-03-07  Tony Chang  <tony@chromium.org>
27567
27568         Reviewed by Mihai Parparita.
27569
27570         [Chromium] Find .checksum files without .pngs and vice-versa
27571         https://bugs.webkit.org/show_bug.cgi?id=55236
27572
27573         * Scripts/find-mismatched-layout-test-results: Added. A script for finding
27574             mismatched results.
27575
27576 2011-03-07  Chris Fleizach  <cfleizach@apple.com>
27577
27578         Reviewed by Beth Dakin.
27579
27580         AX: WK1 needs to use the ScrollView attachment for AXScrollArea, WK2 should not use this element.
27581         https://bugs.webkit.org/show_bug.cgi?id=55706
27582
27583         The rootElement in DRT should point to the scroll view. In WK1, that will be the parent of the rootObject.
27584
27585         * DumpRenderTree/mac/AccessibilityControllerMac.mm:
27586         (AccessibilityController::focusedElement):
27587            Remove FIXME comment about caching the focusedElement. In fact, we shouldn't cache this
27588            because it could change at any time.
27589         (AccessibilityController::rootElement):
27590
27591 2011-03-07  Sam Weinig  <sam@webkit.org>
27592
27593         Reviewed by Anders Carlsson.
27594
27595         Add unableToImplementPolicy callback for WebKit2 policy client
27596         <rdar://problem/9071902>
27597         https://bugs.webkit.org/show_bug.cgi?id=55884
27598
27599         * MiniBrowser/mac/BrowserWindowController.m:
27600         (-[BrowserWindowController awakeFromNib]):
27601         Update policy client initialization struct for new member.
27602
27603 2011-03-07  Sam Weinig  <sam@webkit.org>
27604
27605         Reviewed by Anders Carlsson.
27606
27607         Replace WebKit2's decidePolicyForMIMEType with decidePolicyForResponse
27608         https://bugs.webkit.org/show_bug.cgi?id=55827
27609
27610         * MiniBrowser/mac/BrowserWindowController.m:
27611         (decidePolicyForResponse):
27612         (-[BrowserWindowController awakeFromNib]):
27613         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
27614         (TestWebKitAPI::decidePolicyForResponse):
27615         (TestWebKitAPI::TEST):
27616         Update tools for new name.
27617
27618 2011-03-06  Daniel Bates  <dbates@rim.com>
27619
27620         Reviewed by Eric Seidel and David Kilzer.
27621
27622         Fix misspelled word in build-webkit's checkForJavaSDK() and sort forward declarations
27623         https://bugs.webkit.org/show_bug.cgi?id=55503
27624
27625         Fix misspelled word "Dowloads" [sic] in the message printed to standard output
27626         in checkForJavaSDK().
27627
27628         Also, remove extraneous space in function prototype for unlinkZeroFiles.
27629
27630         * Scripts/build-webkit:
27631
27632 2011-03-05  Brian Weinstein  <bweinstein@apple.com>
27633
27634         Reviewed by Sam Weinig.
27635
27636         Fix Web Process crashes caused by https://bugs.webkit.org/show_bug.cgi?id=53919.
27637
27638         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
27639         (WTR::InjectedBundlePage::willSendRequestForFrame): Retain the WKURLRequestRef returned
27640             by willSendRequestForFrame. The API expects a retained reference to the URL request.
27641
27642 2011-03-05  Ilya Sherman  <isherman@chromium.org>
27643
27644         Reviewed by Darin Adler.
27645
27646         HTMLInputElement::setValue() should schedule change event when the element is focused
27647         In service of https://code.google.com/p/chromium/issues/detail?id=42716
27648         https://bugs.webkit.org/show_bug.cgi?id=53160
27649
27650         * DumpRenderTree/LayoutTestController.cpp:
27651         (setValueForUserCallback):
27652         (LayoutTestController::staticFunctions):
27653         * DumpRenderTree/LayoutTestController.h:
27654         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
27655         (LayoutTestController::setValueForUser): Stubbed out; need real implementation.
27656         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
27657         (LayoutTestController::setValueForUser): Expose HTMLInputElement::setValueForUser() to tests
27658         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
27659         (LayoutTestController::setValueForUser): Stubbed out; need real implementation.
27660         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
27661         (LayoutTestController::setValueForUser): Stubbed out; need real implementation.
27662         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
27663         (LayoutTestController::setValueForUser): Stubbed out; need real implementation.
27664
27665 2011-03-04  Mark Rowe  <mrowe@apple.com>
27666
27667         Reviewed by Alexey Proskuryakov.
27668
27669         <rdar://problem/9055252> DumpRenderTree is crashing with GuardMalloc
27670
27671         GuardMalloc does not expect the memory containing its malloc_zone_t to ever be read-only.
27672
27673         * DumpRenderTree/mac/CheckedMalloc.cpp:
27674         (protectionOfRegion): Retrieve the protection flags for the VM region containing the given address.
27675         (makeLargeMallocFailSilently): Save the initial protection flags, temporarily make the region writeable,
27676         and then restore the initial protection flags when we're done.
27677
27678 2011-03-04  Evan Martin  <evan@chromium.org>
27679
27680         Reviewed by Tony Chang.
27681
27682         [gdb] pretty-print KURL in gdb
27683         https://bugs.webkit.org/show_bug.cgi?id=55788
27684
27685         * gdb/webkit.py: extend the pretty-printers to print WTF::CString
27686         and KURLGooglePrivate.  As always, "p/r" can be used to sidestep
27687         pretty-printing.
27688
27689 2011-03-03  Timothy Hatcher  <timothy@apple.com>
27690
27691         Update the tests that use WKPageRunJavaScriptInMainFrame use the WKSerializedScriptValueRef.
27692
27693         Reviewed by Darin Adler.
27694
27695         * TestWebKitAPI/JavaScriptTest.cpp:
27696         (TestWebKitAPI::javaScriptCallback):
27697         * TestWebKitAPI/Tests/WebKit2/EvaluateJavaScript.cpp:
27698         (TestWebKitAPI::didRunJavaScript):
27699         * TestWebKitAPI/Tests/WebKit2/PageLoadDidChangeLocationWithinPageForFrame.cpp:
27700         (TestWebKitAPI::nullJavaScriptCallback):
27701         * TestWebKitAPI/Tests/WebKit2/PreventEmptyUserAgent.cpp:
27702         (TestWebKitAPI::didRunJavaScript):
27703
27704 2011-03-03  Timothy Hatcher  <timothy@apple.com>
27705
27706         Make run-api-tests show failed if a test fails.
27707
27708         https://webkit.org/b/55725
27709
27710         Reviewed by Sam Weinig.
27711
27712         * Scripts/run-api-tests:
27713         (runTest): Remove "my" from the nested $result assignment so the
27714         check outside the condition sees the real result.
27715
27716 2011-03-02  MORITA Hajime  <morrita@google.com>
27717
27718         Reviewed by Kent Tamura.
27719
27720         [Chromium][DRT] WebTask should use webkit_support::TaskAdaptor
27721         https://bugs.webkit.org/show_bug.cgi?id=55434
27722
27723         Changed WebTask to be a subclass of webkit_support::TaskAdaptor
27724         and passed the object webkit_support::PostDelayedTask directly.
27725
27726         Note that the change keeps postTask() as is because it has a
27727         semantics taht webkit_support::PostDelayedTask() has; It jumps to
27728         the main thread.
27729
27730         * DumpRenderTree/chromium/Task.cpp:
27731         (postDelayedTask):
27732         * DumpRenderTree/chromium/Task.h:
27733         (WebTask::Run):
27734
27735 2011-03-03  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
27736
27737         Reviewed by Kenneth Rohde Christiansen.
27738
27739         [QT] Implement mock client-based geolocation for layout testing
27740         https://bugs.webkit.org/show_bug.cgi?id=54334
27741
27742         Implement testing for Client-Based geolocation for QtWebkit.
27743
27744         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
27745         (WebCore::DumpRenderTree::DumpRenderTree):
27746         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
27747         (LayoutTestController::reset):
27748         (LayoutTestController::setGeolocationPermission):
27749         (LayoutTestController::setMockGeolocationError):
27750         (LayoutTestController::setMockGeolocationPosition):
27751
27752 2011-03-03  Sheriff Bot  <webkit.review.bot@gmail.com>
27753
27754         Unreviewed, rolling out r80237.
27755         http://trac.webkit.org/changeset/80237
27756         https://bugs.webkit.org/show_bug.cgi?id=55714
27757
27758         Windows bots aren't ready to be core (Requested by aroben on
27759         #webkit).
27760
27761         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
27762         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
27763
27764 2011-03-03  Mihai Parparita  <mihaip@chromium.org>
27765
27766         Reviewed by Tony Chang.
27767
27768         NRWT: AttributeError: TestRunner2 instance has no attribute '_cancel_workers'
27769         https://bugs.webkit.org/show_bug.cgi?id=55694
27770
27771         Fixes for interruptions in NRWT:
27772         - Make TestRunInterruptedException be pickleable correctly (the base
27773           Exception class defines a __reduce__ that does not include the reason)
27774         - Fix ordering of returned arguments from TestRunner2._run_tests
27775           (interrupted and keyboard_interrupted were reversed)
27776         - Fix cancel_workers callsites (was using old name).
27777         - In handle_exception re-raise actual exception instance that was thrown
27778
27779         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
27780         * Scripts/webkitpy/layout_tests/layout_package/test_runner2.py:
27781         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
27782         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
27783
27784 2011-03-03  Qi Zhang  <qi.2.zhang@nokia.com>
27785
27786         Unreviewed.
27787
27788         Adding myself as a committer.
27789
27790         * Scripts/webkitpy/common/config/committers.py:
27791
27792 2011-03-03  Tony Gentilcore  <tonyg@chromium.org>
27793
27794         Reviewed by Mihai Parparita.
27795
27796         Teach sheriffbot to answer whois commands
27797         https://bugs.webkit.org/show_bug.cgi?id=55687
27798
27799         * Scripts/webkitpy/tool/bot/irc_command.py:
27800         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
27801         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
27802
27803 2011-03-03  Victoria Kirst  <vrk@google.com>
27804
27805         Reviewed by Eric Carlson.
27806
27807         Enable media tests in chromium-gpu/test_expectations
27808         https://bugs.webkit.org/show_bug.cgi?id=53608
27809
27810         Add media folder to the list of GPU test paths.
27811
27812         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
27813
27814 2011-03-03  Adam Roben  <aroben@apple.com>
27815
27816         Work around a bug in KURL's parsing of Windows-style absolute file: URLs
27817
27818         Fixes <http://webkit.org/b/55674> Many tests fail in WebKit2 mode on Windows due to
27819         incorrect parsing of absolute Windows-style file: URLs
27820
27821         Reviewed by Tim Hatcher.
27822
27823         * WebKitTestRunner/TestInvocation.cpp:
27824         (WTR::createWKURL): Add a "localhost" host to file: URLs we create to work around
27825         <http://webkit.org/b/55683>.
27826
27827 2011-03-03  Tony Chang  <tony@chromium.org>
27828
27829         Reviewed by Dimitri Glazkov.
27830
27831         [chromium] Run kill-old-processes on the chromium-win bot too
27832         https://bugs.webkit.org/show_bug.cgi?id=55630
27833
27834         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
27835         * BuildSlaveSupport/win/kill-old-processes: Add wdiff.exe to the kill list.
27836
27837 2011-03-03  Carlos Garcia Campos  <cgarcia@igalia.com>
27838
27839         Reviewed by Martin Robinson.
27840
27841         [GTK] DRT needs implementation of EventSender.scheduleAsynchronousClick
27842         https://bugs.webkit.org/show_bug.cgi?id=53960
27843
27844         * DumpRenderTree/gtk/EventSender.cpp:
27845         (sendClick):
27846         (scheduleAsynchronousClickCallback):
27847
27848 2011-03-03  Adam Roben  <aroben@apple.com>
27849
27850         Look for Windows-style paths when checking whether a test is an Inspector test
27851
27852         WTR part of <http://webkit.org/b/55672> <rdar://problem/9080867> All inspector tests time
27853         out or crash on Windows 7 Release (WebKit2 Tests)
27854
27855         Reviewed by Darin Adler.
27856
27857         * WebKitTestRunner/TestInvocation.cpp:
27858         (WTR::shouldOpenWebInspector): Also look for backslashes, since that's what we'll get on
27859         Windows.
27860
27861 2011-03-03  Adam Roben  <aroben@apple.com>
27862
27863         Make Windows 7 Release (Tests) a core builder
27864
27865         We can probably make Windows XP Debug (Tests) a core builder, too, but we should probably
27866         give it a few more days of greenness first.
27867
27868         Fixes <http://webkit.org/b/55665> Windows 7 Release (Tests) should be a core builder
27869
27870         Reviewed by Andreas Kling.
27871
27872         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
27873         (BuildBot): Added a regexp to match Windows 7 Release (Tests).
27874
27875         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
27876         (BuildBotTest): Updated the list of builders to match the current list on build.webkit.org.
27877         Updated the core regexps to match the buildbot module. Updated the expected results to match
27878         the current set of core builders.
27879
27880 2011-03-03  Benjamin Poulain  <benjamin.poulain@nokia.com>
27881
27882         Reviewed by Andreas Kling.
27883
27884         Adding myself as a reviewer.
27885
27886         * Scripts/webkitpy/common/config/committers.py:
27887
27888 2011-03-03  Lukasz Slachciak  <l.slachciak@samsung.com>
27889
27890         Reviewed by Xan Lopez.
27891
27892         [GTK] Extended application cache database API and added unit tests file.
27893         https://bugs.webkit.org/show_bug.cgi?id=55335
27894
27895         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: removed external declaration
27896         of webkit_application_cache_set_maximum_size due to API exposal.
27897
27898 2011-03-02  Ojan Vafai  <ojan@chromium.org>
27899
27900         Reviewed by Adam Barth.
27901
27902         make webkit-patch upload work for security bugs
27903         https://bugs.webkit.org/show_bug.cgi?id=55655
27904
27905         The problem is that we shell out to prepare-ChangeLog,
27906         which does not authenticate with bugzilla. All it needs from
27907         bugzilla is the bug title though. We just pass it through from
27908         webkit-patch instead.
27909
27910         * Scripts/prepare-ChangeLog:
27911         * Scripts/webkitpy/common/net/bugzilla/bug.py:
27912         * Scripts/webkitpy/tool/steps/preparechangelog.py:
27913
27914 2011-03-01  Ojan Vafai  <ojan@chromium.org>
27915
27916         Reviewed by Tony Chang.
27917
27918         [test-dashboard] make gtest JSON match layout-tests JSON
27919         https://bugs.webkit.org/show_bug.cgi?id=55556
27920
27921         http://trac.webkit.org/changeset/80090 stopped converting time_ms to t.
27922         Make this code match.
27923
27924         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
27925
27926 2011-03-02  Dirk Pranke  <dpranke@chromium.org>
27927
27928         Reviewed by Tony Chang.
27929
27930         NRWT - implement Linux Hardy 64-bit port support. This involves
27931         adding support for non-'x86' architectures and updating the
27932         "deduplicate_tests" script to correctly understand that
27933         port.name() is not always equal to basename(port.baseline_path()).
27934
27935         This change introduces two new port names: 'chromium-linux-x86'
27936         and 'chromium-linux-x86_64'. Using the prior 'chromium-linux'
27937         flag will cause the code to determine which one to use at
27938         runtime, just like we currently do with 'chromium-win'.
27939
27940         https://bugs.webkit.org/show_bug.cgi?id=55535
27941
27942         * Scripts/webkitpy/common/system/executive_mock.py:
27943           Handle 'return_stderr' keyword arg to run_command()
27944         * Scripts/webkitpy/layout_tests/deduplicate_tests.py:
27945         * Scripts/webkitpy/layout_tests/deduplicate_tests_unittest.py:
27946         * Scripts/webkitpy/layout_tests/port/base.py:
27947         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
27948         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
27949         * Scripts/webkitpy/layout_tests/port/factory.py:
27950         * Scripts/webkitpy/layout_tests/port/google_chrome.py:
27951
27952 2011-03-02  Kevin Ollivier  <kevino@theolliviers.com>
27953
27954         [wx] Build fix, make sure we put the wxWebKit dylib in the right directory and
27955         update symlinks accordingly.
27956
27957         * wx/packaging/build-mac-installer.py:
27958
27959 2011-03-02  Sheriff Bot  <webkit.review.bot@gmail.com>
27960
27961         Unreviewed, rolling out r80139.
27962         http://trac.webkit.org/changeset/80139
27963         https://bugs.webkit.org/show_bug.cgi?id=55620
27964
27965         failing and crashing tests on Snow Leopard bot (Requested by
27966         mihaip1 on #webkit).
27967
27968         * DumpRenderTree/LayoutTestController.cpp:
27969         (LayoutTestController::staticFunctions):
27970         * DumpRenderTree/LayoutTestController.h:
27971         * DumpRenderTree/chromium/LayoutTestController.cpp:
27972         (LayoutTestController::LayoutTestController):
27973         * DumpRenderTree/chromium/LayoutTestController.h:
27974         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
27975         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
27976         * DumpRenderTree/mac/UIDelegate.h:
27977         * DumpRenderTree/mac/UIDelegate.mm:
27978         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
27979         * DumpRenderTree/qt/LayoutTestControllerQt.h:
27980         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
27981         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
27982
27983 2011-03-02  Adam Roben  <aroben@apple.com>
27984
27985         Wait for force a paint in WebKitTestRunner until we're actually ready to dump the output
27986
27987         This matches DumpRenderTree on Windows.
27988
27989         Fixes <http://webkit.org/b/55469> <rdar://problem/9068539> REGRESSION (r79863): Lots of
27990         dom/html/level2/html/HTMLFrameElement*.html tests crashing in FrameView::paintOverhangAreas
27991         in WebKit2
27992
27993         Reviewed by Beth Dakin.
27994
27995         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
27996         (WTR::InjectedBundlePage::dump): Moved code to force a paint here...
27997         (WTR::InjectedBundlePage::didFinishLoadForFrame): ...from here.
27998
27999 2011-03-02  Daniel Cheng  <dcheng@chromium.org>
28000
28001         Reviewed by David Levin.
28002
28003         Add feature define for data transfer items
28004         https://bugs.webkit.org/show_bug.cgi?id=55510
28005
28006         * Scripts/build-webkit:
28007
28008 2011-03-02  Adam Roben  <aroben@apple.com>
28009
28010         Look in the new-to-SnowLeopard location for crash logs on SnowLeopard and newer
28011
28012         ReportCrash saves logs to ~/Library/Logs/CrashReporter on Leopard, but to
28013         ~/Library/Logs/DiagnosticReports on SnowLeopard. old-run-webkit-tests was only looking in
28014         the former location, but was getting lucky on some SnowLeopard machines because symlinks
28015         were being created there pointing into DiagnosticReports. For machines without these
28016         symlinks, crash logs were not getting captured.
28017
28018         Fixes <http://webkit.org/b/55607> old-run-webkit-tests doesn't capture crash logs on some
28019         SnowLeopard machines
28020
28021         Reviewed by Eric Seidel.
28022
28023         * Scripts/old-run-webkit-tests:
28024         (captureSavedCrashLog): Look in ~/Library/Logs/DiagnosticReports on SnowLeopard and newer.
28025
28026 2011-03-02  Adam Roben  <aroben@apple.com>
28027
28028         Don't try to capture crash logs on platforms that haven't implemented that feature
28029
28030         Fixes <http://webkit.org/b/55504> Use of uninitialized value in numeric lt (<) at
28031         Tools/Scripts/old-run-webkit-tests line 1778 seen on Qt bots after a crash
28032
28033         Reviewed by Joseph Pecoraro.
28034
28035         * Scripts/old-run-webkit-tests:
28036         (captureSavedCrashLog): Bail if we don't have a crash log glob.
28037         (findNewestFileMatchingGlob): Fix a buggy test for when no paths match the glob. This isn't
28038         strictly required to fix bug 55504, but seems worthwhile.
28039
28040 2011-01-26  John Knottenbelt  <jknotten@chromium.org>
28041
28042         Reviewed by Dmitry Titov.
28043
28044         Detach Geolocation from Frame when Page destroyed.
28045         https://bugs.webkit.org/show_bug.cgi?id=52877
28046
28047         Extend the layout test controller to expose the number of pending
28048         geolocation requests, so that we can test that the requests have
28049         been cancelled on page close.
28050
28051         * DumpRenderTree/LayoutTestController.cpp:
28052         (numberOfPendingGeolocationPermissionRequestsCallback):
28053         (LayoutTestController::staticFunctions):
28054         * DumpRenderTree/LayoutTestController.h:
28055         * DumpRenderTree/chromium/LayoutTestController.cpp:
28056         (LayoutTestController::LayoutTestController):
28057         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
28058         * DumpRenderTree/chromium/LayoutTestController.h:
28059         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
28060         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
28061         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
28062         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
28063         * DumpRenderTree/mac/UIDelegate.h:
28064         * DumpRenderTree/mac/UIDelegate.mm:
28065         (-[UIDelegate numberOfPendingGeolocationPermissionRequests]):
28066         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
28067         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
28068         * DumpRenderTree/qt/LayoutTestControllerQt.h:
28069         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
28070         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
28071         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
28072         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
28073
28074
28075 2011-03-01  Kent Tamura  <tkent@chromium.org>
28076
28077         Reviewed by Dimitri Glazkov.
28078
28079         Assertion fails when a form validation bubble appears
28080         https://bugs.webkit.org/show_bug.cgi?id=55550
28081
28082         Test: fast/forms/interactive-validation-attach-assertion.html
28083
28084         * DumpRenderTree/chromium/WebPreferences.cpp:
28085         (WebPreferences::applyTo): Enable the interactive validation feature by
28086           default. This matches to Chromium browser.
28087
28088 2011-03-01  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
28089
28090         Reviewed by Kent Tamura.
28091
28092         Remove efl port style exception of Webcore/JavaScriptCore
28093         https://bugs.webkit.org/show_bug.cgi?id=55549
28094
28095         WebCore and JavaScriptCore need to adhere WebKit coding style.
28096
28097         * Scripts/webkitpy/style/checker.py:
28098
28099 2011-03-01  Tony Chang  <tony@chromium.org>
28100
28101         Reviewed by Ojan Vafai.
28102
28103         [chromium] clean up temp files on chromium-mac builder too
28104         https://bugs.webkit.org/show_bug.cgi?id=55525
28105
28106         * BuildSlaveSupport/chromium/remove-crash-logs:
28107
28108 2011-02-28  Ojan Vafai  <ojan@chromium.org>
28109
28110         Reviewed by Tony Chang.
28111
28112         wrap json in a function call to afford cross-domain loading
28113         https://bugs.webkit.org/show_bug.cgi?id=55353
28114
28115         Also delete the code that compacts the JSON. It turns out this
28116         breaks the rebaseline tool and is probably a premature optimization anyways.
28117
28118         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
28119         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
28120         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
28121         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
28122
28123 2011-03-01  Sheriff Bot  <webkit.review.bot@gmail.com>
28124
28125         Unreviewed, rolling out r80079.
28126         http://trac.webkit.org/changeset/80079
28127         https://bugs.webkit.org/show_bug.cgi?id=55547
28128
28129         "Broke the Win debug build?" (Requested by dcheng on #webkit).
28130
28131         * Scripts/build-webkit:
28132
28133 2011-03-01  Mark Rowe  <mrowe@apple.com>
28134
28135         Reviewed by Sam Weinig.
28136
28137         Update verification scripts to handle WebKit2.
28138
28139         * Scripts/check-for-global-initializers: Skip object files that use RefCountedLeakChecker.
28140         * Scripts/check-for-inappropriate-files-in-framework: Add some inappropriate file types for
28141         WebKit2.
28142
28143 2011-03-01  Daniel Cheng  <dcheng@chromium.org>
28144
28145         Reviewed by David Levin.
28146
28147         Add feature define for data transfer items
28148         https://bugs.webkit.org/show_bug.cgi?id=55510
28149
28150         * Scripts/build-webkit:
28151
28152 2011-02-23  Joseph Pecoraro  <joepeck@webkit.org>
28153
28154         Reviewed by Timothy Hatcher.
28155
28156         All Console Messages should be passed to ChromeClients.
28157         https://bugs.webkit.org/show_bug.cgi?id=54926
28158
28159         Update DumpRenderTree's UIDelegate to use the new console message delegate.
28160
28161         * DumpRenderTree/mac/UIDelegate.mm:
28162         (-[UIDelegate webView:addMessageToConsole:withSource:]):
28163
28164 2011-03-01  Dirk Pranke  <dpranke@chromium.org>
28165
28166         Reviewed by Adam Barth.
28167
28168         Tragically, we stopped running the python unit tests
28169         on Chromium Win just as I was getting them to be green.
28170         This patch turns them back on.
28171
28172         https://bugs.webkit.org/show_bug.cgi?id=55521
28173
28174         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
28175
28176 2011-03-01  Dirk Pranke  <dpranke@chromium.org>
28177
28178         Reviewed by Adam Roben.
28179
28180         Disable
28181         webkitpy.common.system.executive_unittest.ExecutiveTest.\
28182            test_kill_all fails on win until it can be properly triaged.
28183
28184         https://bugs.webkit.org/show_bug.cgi?id=54790
28185
28186         * Scripts/webkitpy/common/system/executive_unittest.py:
28187
28188     2011-03-01  Helder Correia  <helder@sencha.com>
28189
28190             Unreviewed. Add myself to committers.py.
28191
28192             * Scripts/webkitpy/common/config/committers.py:
28193
28194 2011-03-01  Adam Roben  <aroben@apple.com>
28195
28196         Give up if a crash log for the web process is taking too long to be saved on Windows
28197
28198         This should keep the bots from getting stuck while waiting for a crash log, if writing a
28199         crash log ever hangs or the UI process for some reason doesn't notice it's finished.
28200
28201         Fixes <http://webkit.org/b/55499> WebKitTestRunner can hang forever waiting for a crash log
28202         to be saved for the web process
28203
28204         Reviewed by Steve Falkenburg.
28205
28206         * WebKitTestRunner/win/TestControllerWin.cpp:
28207         (WTR::runRunLoopUntil): Moved code to run the run loop here here from platformRunUntil.
28208         Generalized the code slightly to handle the optional object to wait on.
28209         (WTR::TestController::platformRunUntil): Use the new runRunLoopUntil function, and also use
28210         it (with a timeout) when waiting for a crash log for the web process to be saved. This will
28211         prevent us from waiting forever.
28212
28213 2011-03-01  Adam Roben  <aroben@apple.com>
28214
28215         Give WebKitLauncherWin's PDB file a name different from WebKit's
28216
28217         WebKitLauncherWin (which builds WebKit.exe) and WebKit (which builds WebKit.dll) were
28218         fighting over the same PDB file, meaning that one of them was going to lose and not have
28219         symbols. WebKitLauncherWin's PDB file is now named WebKitLauncherWin.pdb.
28220
28221         Fixes <http://webkit.org/b/55495> Crashes in WebKit.dll have bad backtraces for some test
28222         runs (due to missing symbols)
28223
28224         Reviewed by David Kilzer.
28225
28226         * WebKitLauncherWin/WebKitLauncherWinCommon.vsprops: Use the project name,
28227         "WebKitLauncherWin", instead of the target name, "WebKit", when naming the PDB file.
28228
28229 2011-03-01  Adam Roben  <aroben@apple.com>
28230
28231         Count web process crashes toward the --exit-after-n-crashes-or-timeouts limit
28232
28233         Fixes <http://webkit.org/b/55492> Web process crashes aren't counted toward the
28234         --exit-after-n-crashes-or-timeouts limit
28235
28236         Reviewed by David Kilzer.
28237
28238         * Scripts/old-run-webkit-tests:
28239         (stopRunningTestsEarlyIfNeeded): Include web process crashes in the computation, and print
28240         the number of tests that caused the web process to crash when exiting early.
28241
28242 2011-03-01  Adam Roben  <aroben@apple.com>
28243
28244         Save a crash log when the web process crashes
28245
28246         On Windows, WebKitTestRunner now detects when the web process is crashing and waits to exit
28247         until it has finished crashing, which guarantees that the crash log will have had time to be
28248         saved, too. On Mac, we always wait until ReportCrash has exited before capturing the crash
28249         log, so all we have to do is choose the right crash log out of the CrashReporter directory.
28250
28251         Fixes <http://webkit.org/b/44121> <rdar://problem/8320759> When the web process crashes and
28252         a crash log is being saved, WebKitTestRunner thinks the web process has become unresponsive
28253
28254         Reviewed by Sam Weinig.
28255
28256         * Scripts/old-run-webkit-tests:
28257         (testCrashedOrTimedOut): Don't kill WebKitTestRunner when the web process crashes. It will
28258         kill itself. On Windows, this will cause us to wait until the crash log has been saved. On
28259         Mac, it should have no effect. Capture saved crash logs for web process crashes, too.
28260         (captureSavedCrashLog): Added $webProcessCrashed argument. On Mac, look for
28261         WebProces_*.crash files when the web process crashes.
28262
28263         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
28264         (WTR::InjectedBundle::initialize): Added an initializationUserData argument. Updated for
28265         initializePlatformDefaults -> platformInitialize rename. Pass the initializationUserData
28266         along to platformInitialize.
28267
28268         * WebKitTestRunner/InjectedBundle/InjectedBundle.h: See above.
28269
28270         * WebKitTestRunner/InjectedBundle/InjectedBundleMain.cpp:
28271         (WKBundleInitialize): Pass along the initializationUserData to the InjectedBundle.
28272
28273         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
28274         (WTR::InjectedBundle::platformInitialize):
28275         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
28276         (WTR::InjectedBundle::platformInitialize):
28277         Updated function signature.
28278
28279         * WebKitTestRunner/InjectedBundle/win/InjectedBundleWin.cpp:
28280         (WTR::exceptionFilter): Added. Tells the UI process we're crashing by signaling the
28281         webProcessCrashingEvent, then lets the crash continue as normal.
28282
28283         (WTR::InjectedBundle::platformInitialize): Hook up exceptionFilter. Retrieve the name of the
28284         event we should use to tell the UI process we're crashing from the initializationUserData,
28285         and get a handle to that event.
28286
28287         * WebKitTestRunner/TestController.cpp:
28288         (WTR::TestController::TestController): Initialize new members.
28289         (WTR::TestController::processDidCrash): Removed unnecessary WKPageRef argument. Changed to
28290         only print the "#CRASHED - WebProcess" message once, since this can be called more than once
28291         when a crash log is being saved on Windows. Exit right away if specified. (This is the
28292         default.)
28293
28294         * WebKitTestRunner/TestController.h: Added new members.
28295
28296         * WebKitTestRunner/win/TestControllerWin.cpp:
28297         (WTR::TestController::platformInitialize): Set up the event the web process will use to tell
28298         us it's crashing.
28299         (WTR::TestController::platformRunUntil): Pass MWMO_INPUTAVAILABLE to
28300         ::MsgWaitForMultipleObjectsEx so we'll process messages that have already been seen by
28301         ::PeekMessage. (This is unrelated to the bug fix.) Notice when the webProcessCrashingEvent
28302         has been signaled. When this happens, print the "#CRASHED - WebProcess" message right away
28303         so the test harness will know the web process has crashed and not try to kill us, then wait
28304         for the web process to finish crashing so a crash log will have time to be saved.
28305         (WTR::toWK): Simple hepler function.
28306         (WTR::TestController::platformInitializeContext): Pass along the name of the event the web
28307         process should use to tell us it is crashing in the context's initialization user data.
28308
28309 2011-03-01  Dimitri Glazkov  <dglazkov@chromium.org>
28310
28311         Reviewed by Tony Gentilcore.
28312
28313         Move the checks to exclude Python/Perl tests for Chromium Win build to the right place.
28314         https://bugs.webkit.org/show_bug.cgi?id=55476
28315
28316         Chromium Win uses BuildAndTestFactory, not TestFactory.
28317
28318         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Moved to the right place.
28319
28320 2011-03-01  Adam Roben  <aroben@apple.com>
28321
28322         Test that the WebKit2 responsiveness timer doesn't fire too early
28323
28324         Test for <http://webkit.org/b/55417> <rdar://problem/9065287> RunLoop::Timer fires 1000x too
28325         early on Windows
28326
28327         Reviewed by Anders Carlsson.
28328
28329         * TestWebKitAPI/PlatformUtilities.h: Added sleep().
28330
28331         * TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly.cpp: Added.
28332         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
28333         (TestWebKitAPI::didFinishLoadForFrame):
28334         (TestWebKitAPI::processDidBecomeUnresponsive):
28335         (TestWebKitAPI::setInjectedBundleClient):
28336         (TestWebKitAPI::setPageLoaderClient):
28337         Simple helper functions.
28338
28339         (TestWebKitAPI::TEST): Load an HTML file to make sure the web process is initialized. Then
28340         tell the web process to pause and press the spacebar key. The spacebar keypress should cause
28341         the responsiveness timer to start, but the web process should unpause before it has a chance
28342         to fire. Run until the web process has unpaused, and assert that the timer didn't fire.
28343
28344         * TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly_Bundle.cpp: Added.
28345         (TestWebKitAPI::ResponsivenessTimerDoesntFireEarlyTest::ResponsivenessTimerDoesntFireEarlyTest):
28346         Call up to the base class.
28347         (TestWebKitAPI::ResponsivenessTimerDoesntFireEarlyTest::didReceiveMessage): When asked to
28348         pause, sleep for 0.5 seconds, then send back a message saying we paused.
28349
28350         * TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
28351         (TestWebKitAPI::Util::sleep): Added. Calls through to usleep.
28352         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp:
28353         (TestWebKitAPI::Util::sleep): Added. Calls through to ::Sleep.
28354
28355         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
28356         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
28357         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
28358         Added new files.
28359
28360 2011-03-01  Andras Becsi  <abecsi@webkit.org>
28361
28362         Reviewed by Csaba Osztrogonác.
28363
28364         [Qt] Clean up the project files and move common options to WebKit.pri.
28365
28366         * WebKitTestRunner/qt/WebKitTestRunner.pro: Deduplicate.
28367
28368
28369 2011-03-01  Adam Roben  <aroben@apple.com>
28370
28371         Skip another multiprocessing test on Windows
28372
28373         It is sometimes failing (and may be leaving child Python processes around).
28374         <http://webkit.org/b/55087> tracks the failure.
28375
28376         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
28377         (FuncitonTests.test_get__processes): Skip this test on Windows.
28378
28379 2011-03-01  Ilya Tikhonovsky  <loislo@chromium.org>
28380
28381         Reviewed by Pavel Feldman.
28382
28383         Web Inspector: reduce inspector tests flakiness.
28384         https://bugs.webkit.org/show_bug.cgi?id=55401
28385
28386         * DumpRenderTree/chromium/TestShell.cpp:
28387         (TestShell::runFileTest):
28388         * DumpRenderTree/chromium/WebPreferences.cpp:
28389         (WebPreferences::reset):
28390
28391 2011-02-28  Mihai Parparita  <mihaip@chromium.org>
28392
28393         Reviewed by James Robinson.
28394
28395         NRWT should still upload results when it exits early due to too many failures
28396         https://bugs.webkit.org/show_bug.cgi?id=55432
28397
28398         Only keyboard interrupts should prevent results from being uploaded.
28399
28400         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
28401         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
28402
28403 2011-02-28  Alice Liu  <alice.liu@apple.com>
28404
28405         Reviewed by Adam Roben and Gavin Barraclough.
28406
28407         Test for https://bugs.webkit.org/show_bug.cgi?id=54898
28408         InjectedBundleNodeHandle dies too early in WKBundleHitTestResultGetNodeHandle.
28409
28410         * TestWebKitAPI/InjectedBundleController.h:
28411         (TestWebKitAPI::InjectedBundleController::bundle):
28412         * TestWebKitAPI/PlatformWebView.h:
28413         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle.cpp: Added.
28414         (TestWebKitAPI::didFinishLoadForFrame):
28415         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
28416         (TestWebKitAPI::setPageLoaderClient):
28417         (TestWebKitAPI::TEST):
28418         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle_Bundle.cpp: Added.
28419         (TestWebKitAPI::HitTestResultNodeHandleTest::HitTestResultNodeHandleTest):
28420         (TestWebKitAPI::HitTestResultNodeHandleTest::getContextMenuFromDefaultMenu):
28421         (TestWebKitAPI::HitTestResultNodeHandleTest::didCreatePage):
28422         * TestWebKitAPI/mac/PlatformWebViewMac.mm:
28423         (TestWebKitAPI::PlatformWebView::simulateRightClick):
28424         * TestWebKitAPI/win/PlatformWebViewWin.cpp:
28425         (TestWebKitAPI::PlatformWebView::simulateRightClick):
28426
28427         Added files to:
28428         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
28429         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
28430         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
28431
28432 2011-02-28  Adam Roben  <aroben@apple.com>
28433
28434         Start running InjectedBundle API tests on Windows
28435
28436         A new project, TestWebKitAPIInjectedBundle, builds the injected bundle DLL.
28437
28438         Fixes <http://webkit.org/b/55420> run-api-tests doesn't run injected bundle tests on Windows
28439
28440         Reviewed by Darin Adler.
28441
28442         * TestWebKitAPI/Configurations/TestWebKitAPIDebug.vsprops: Added.
28443         * TestWebKitAPI/Configurations/TestWebKitAPIDebugAll.vsprops: Added.
28444         * TestWebKitAPI/Configurations/TestWebKitAPIDebugCairoCFLite.vsprops: Added.
28445         * TestWebKitAPI/Configurations/TestWebKitAPIRelease.vsprops: Added.
28446         * TestWebKitAPI/Configurations/TestWebKitAPIReleaseCairoCFLite.vsprops: Added.
28447         These are just extracted from TestWebKitAPI.vcproj.
28448
28449         * TestWebKitAPI/Configurations/TestWebKitAPIInjectedBundleCommon.vsprops: Added. Just sets
28450         up the DLL's filename.
28451
28452         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp: Fixed DLL name.
28453         (TestWebKitAPI::Util::createInjectedBundlePath): Changed to use kCFAllocatorNull instead of
28454         0 (== kCFAllocatorDefault) so that CF won't try to deallocate our C string.
28455
28456         * TestWebKitAPI/win/TestWebKitAPI.sln: Added TestWebKitAPIInjectedBundle. Removed the bogus
28457         "all" configuration.
28458
28459         * TestWebKitAPI/win/TestWebKitAPI.vcproj: Changed to use the new .vsprops files. Added
28460         injected bundle tests. Let VS resort things.
28461
28462         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj: Copied from Tools/TestWebKitAPI/win/TestWebKitAPI.vcproj.
28463
28464 2011-02-28  Adam Roben  <aroben@apple.com>
28465
28466         Add a test for the paint rect passed via WM_PAINT to windowless plugins
28467
28468         Test for <http://webkit.org/b/55365> <rdar://problem/9031089> REGRESSION (r79040): WebKit2:
28469         Large portions of pages with plugins paint black
28470
28471         Reviewed by Anders Carlsson.
28472
28473         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
28474         (PluginTest::NPP_HandleEvent):
28475         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
28476         Added NPP_HandleEvent. Just returns 0 at this level.
28477
28478         * DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowlessPaintRectCoordinates.cpp: Added.
28479         (WindowlessPaintRectCoordinates::WindowlessPaintRectCoordinates): Call up to the base class.
28480         (WindowlessPaintRectCoordinates::NPP_New): Mark ourselves as windowless.
28481         (WindowlessPaintRectCoordinates::NPP_HandleEvent): Check that the paint rect passed via
28482         WM_PAINT has the expected coordinates.
28483
28484         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
28485         (NPP_HandleEvent): Give the PluginTest a chance to handle the event before doing anything
28486         else.
28487
28488         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Added new file, let VS
28489         resort things.
28490
28491         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
28492         (WTR::InjectedBundlePage::didFinishLoadForFrame): Added a call to WKBundlePageForceRepaint.
28493         This is roughly equivalent to the call to -[WebView displayIfNeeded] in -[FrameLoadDelegate
28494         webView:didFinishLoadForFrame:] in DumpRenderTree.
28495
28496         * WebKitTestRunner/win/PlatformWebViewWin.cpp:
28497         (WTR::PlatformWebView::PlatformWebView): Tell the WKView it's in a window so that plugins
28498         can start running.
28499
28500 2011-02-28  Adam Roben  <aroben@apple.com>
28501
28502         Tell MiniBrowser's WKViews that they're in a window
28503
28504         Fixes <http://webkit.org/b/55364> Plugins don't work in MiniBrowser on Windows
28505
28506         Reviewed by Anders Carlsson.
28507
28508         * MiniBrowser/win/BrowserView.cpp:
28509         (BrowserView::create): Call WKViewSetIsInWindow after creating the view.
28510
28511 2011-02-28  Alexis Menard  <alexis.menard@openbossa.org>
28512
28513         Reviewed by Csaba Osztrogonác.
28514
28515         The new QML tests require QML_IMPORT_PATH variable to be set in order to
28516         load the proper plugin.
28517
28518         * Scripts/run-qtwebkit-tests:
28519
28520 2011-02-28  Joone Hur  <joone.hur@collabora.co.uk>
28521
28522         Unreviewed, fix spelling mistake.
28523
28524         hardware-accelated => hardware-accelerated.
28525
28526         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
28527
28528 2011-02-28  Kristian Amlie  <kristian.amlie@nokia.com>
28529
28530         Reviewed by Andreas Kling.
28531
28532         Made sure that the build-webkit --qmake argument is always respected.
28533
28534         Previously, it would be respected during the build, but not during
28535         the early feature detection. The build would also fail if qmake was
28536         not in your path, even if you specified --qmake with a valid qmake
28537         binary.
28538
28539         [Qt] WebKit patches required to work with a modularized version of Qt
28540         https://bugs.webkit.org/show_bug.cgi?id=53916
28541
28542         * Scripts/build-webkit:
28543         * Scripts/webkitdirs.pm:
28544
28545 2011-02-27  Ojan Vafai  <ojan@chromium.org>
28546
28547         Fix failing tests from http://trac.webkit.org/changeset/79837.
28548
28549         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
28550         This was just a mistake that would break if someone tried printing with this configuration.
28551         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
28552         Method signature changed. Didn't notice that this method was being calling in a unittest.
28553
28554 2011-02-24  Hayato Ito  <hayato@chromium.org>
28555
28556         Reviewed by Eric Seidel.
28557
28558         [NRWT] Remove test_types/* classes.
28559
28560         This is the last patch in the series of efforts which tried to remove test_types classes.
28561         A single_test_runner module will take over responsibilities of test_type classes.
28562
28563         From this patch, we won't display statistics per test_type even if a '--verbose' option is given.
28564
28565         https://bugs.webkit.org/show_bug.cgi?id=55123
28566
28567         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
28568         * Scripts/webkitpy/layout_tests/layout_package/test_results.py:
28569         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py:
28570         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
28571         * Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py:
28572         * Scripts/webkitpy/layout_tests/test_types/__init__.py: Removed.
28573         * Scripts/webkitpy/layout_tests/test_types/image_diff.py: Removed.
28574         * Scripts/webkitpy/layout_tests/test_types/test_type_base.py: Removed.
28575         * Scripts/webkitpy/layout_tests/test_types/test_type_base_unittest.py: Removed.
28576         * Scripts/webkitpy/layout_tests/test_types/text_diff.py: Removed.
28577
28578 2011-02-25  Ojan Vafai  <ojan@chromium.org>
28579
28580         Reviewed by Tony Chang.
28581
28582         Change results.json format to the one used by unexpected_results.json
28583         https://bugs.webkit.org/show_bug.cgi?id=52267
28584
28585         Also add runtimes in milliseconds to the JSON and make the output format more compact.
28586         Named the file full_results.json to avoid conflicting with the results.json
28587         file the test-results server currently serves up.
28588
28589         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
28590         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
28591         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
28592         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
28593         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
28594         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
28595
28596 2011-02-27  Adam Roben  <aroben@apple.com>
28597
28598         Followup fix to r79827
28599
28600         * DumpRenderTree/win/UIDelegate.cpp:
28601         (UIDelegate::webViewAddMessageToConsole): Replaced call to urlSuitableForTestResult with
28602         lastPathComponent. This matches what we did here before r79827, and matches Mac.
28603
28604 2011-02-27  Dan Bernstein  <mitz@apple.com>
28605
28606         Reviewed by Adam Roben.
28607
28608         Make bisect-builds print a trac link for the regression range
28609         https://bugs.webkit.org/show_bug.cgi?id=55331
28610
28611         * Scripts/bisect-builds:
28612
28613 2011-02-27  Adam Roben  <aroben@apple.com>
28614
28615         Use iswalpha instead of isalpha when dealing with wchar_ts in EditingDelegate
28616
28617         Fixes <http://webkit.org/b/55062> <rdar://problem/9059907> Crash beneath
28618         EditingDelegate::checkSpellingOfString when running fast/forms/input-text-maxlength.html or
28619         fast/forms/input-text-paste-maxlength.html on Windows with full page heap enabled
28620
28621         Reviewed by Anders Carlsson.
28622
28623         * DumpRenderTree/win/EditingDelegate.cpp:
28624         (indexOfFirstWordCharacter):
28625         (wordLength):
28626         Use iswalpha instead of isalpha, since TCHAR is really just wchar_t.
28627
28628 2011-02-27  Adam Roben  <aroben@apple.com>
28629
28630         Make ResourceLoadDelegate print URLs relative to the main resource on Windows
28631
28632         This matches Mac. (In fact, the code was ported from Mac!)
28633
28634         Fixes <http://webkit.org/b/55328> URLs printed by ResourceLoadDelegate on Windows don't
28635         match Mac
28636
28637         Reviewed by Anders Carlsson.
28638
28639         * DumpRenderTree/win/DumpRenderTree.cpp:
28640         (substringFromIndex): Added. Emulates -[NSString substringFromIndex:].
28641         (urlSuitableForTestResult): Ported code from -[NSURL(DRTExtras)
28642         _drt_descriptionSuitableForTestResult].
28643         (cfStringRefToWString): Moved here from LayoutTestControllerWin.cpp.
28644
28645         * DumpRenderTree/win/DumpRenderTreeWin.h: Added declaration of cfStringRefToWString.
28646
28647         * DumpRenderTree/win/LayoutTestControllerWin.cpp: Moved cfStringRefToWString to
28648         DumpRenderTree.cpp.
28649
28650 2011-02-27  Adam Roben  <aroben@apple.com>
28651
28652         Change the order of Chang Shu's email addresses so the first one is their Bugzilla username
28653
28654         This makes Bugzilla autocompletion work correctly.
28655
28656         * Scripts/webkitpy/common/config/committers.py:
28657
28658 2011-02-27  Patrick Gansterer  <paroga@webkit.org>
28659
28660         Reviewed by Andreas Kling.
28661
28662         [Qt] fast/loader/user-stylesheet-fast-path.html fails
28663         https://bugs.webkit.org/show_bug.cgi?id=50911
28664
28665         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
28666         (LayoutTestController::setUserStyleSheetLocation):
28667         Use QUrl::fromEncoded since the passed string is already encoded.
28668
28669 2011-02-27  Adam Roben  <aroben@apple.com>
28670
28671         Ensure $testResultsDirectory is an absolute path before setting up Windows crash log saving
28672
28673         Windows 7 (and maybe Vista?) launches the post-mortem debugger with a working directory that
28674         is different from old-run-webkit-test's. Because of this, the path to the crash log file
28675         (which is relative to $testResultsDirectory) must be an absolute path. See the bug for more
28676         details.
28677
28678         Fixes <http://webkit.org/b/55318> Crash logs aren't saved on Windows 7 when a relative
28679         --results-directory path is used (like on the test slaves)
28680
28681         Reviewed by Dan Bernstein.
28682
28683         * Scripts/old-run-webkit-tests: Moved the call to setUpWindowsCrashLogSaving after we've
28684         made $testResultsDirectory an absolute path.
28685
28686 2011-02-26  Dominic Mazzoni  <dmazzoni@google.com>
28687
28688         Reviewed by Adam Barth.
28689
28690         m_dumpAccessibilityNotifications should be initialized.
28691         https://bugs.webkit.org/show_bug.cgi?id=54874
28692
28693         * DumpRenderTree/chromium/AccessibilityController.cpp:
28694         (AccessibilityController::AccessibilityController):
28695
28696 2011-02-26  Vsevolod Vlasov  <vsevik@chromium.org>
28697
28698         Reviewed by Pavel Feldman.
28699
28700         DumpRenderTree should reset frame opener between tests.
28701         https://bugs.webkit.org/show_bug.cgi?id=54874
28702
28703         Added clearOpener() calls between tests.
28704         Some tests are using window.open() to change current window location.
28705         This changes window.opener property whose nullity is checked before
28706         showing xml viewer.
28707
28708         * DumpRenderTree/chromium/TestShell.cpp:
28709         (TestShell::resetTestController):
28710         * DumpRenderTree/gtk/DumpRenderTree.cpp:
28711         (resetDefaultsToConsistentValues):
28712         * DumpRenderTree/mac/DumpRenderTree.mm:
28713         (resetWebViewToConsistentStateBeforeTesting):
28714         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
28715         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
28716         * DumpRenderTree/win/DumpRenderTree.cpp:
28717         (resetWebViewToConsistentStateBeforeTesting):
28718         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
28719         (WTR::InjectedBundlePage::reset):
28720
28721 2011-02-26  Yael Aharon  <yael.aharon@nokia.com>
28722
28723         Reviewed by Andreas Kling.
28724
28725         [Qt] Fix the focus in MiniBrowser.
28726         https://bugs.webkit.org/show_bug.cgi?id=55288
28727
28728         Give focus to the web view when a load starts.
28729
28730         * MiniBrowser/qt/BrowserWindow.cpp:
28731         (BrowserWindow::urlChanged):
28732
28733 2011-02-26  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
28734
28735         Reviewed by Adam Barth.
28736
28737         [EFL] Add coding style exceptions for EFL port
28738         https://bugs.webkit.org/show_bug.cgi?id=54733
28739
28740         The EFL APIs use EFL naming style, which includes both lower-cased and camel-cased,
28741         underscore-sparated values. But, style checking rule have found parameter_name errors
28742         in efl directories so far.
28743
28744         * Scripts/webkitpy/style/checker.py: Add prameter_name rule's exception and efl directories.
28745
28746 2011-02-26  David Levin  <levin@chromium.org>
28747
28748         Reviewed by Darin Adler.
28749
28750         check-webkit-style should check for spaces after periods in comments
28751         https://bugs.webkit.org/show_bug.cgi?id=55253
28752
28753         * Scripts/webkitpy/style/checkers/cpp.py: Added the check.
28754         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added tests
28755           and fixed the name of one related test.
28756
28757 2011-02-26  Kenneth Rohde Christiansen  <kenneth@webkit.org>
28758
28759         Reviewed by Andreas Kling.
28760
28761         Make it possible to test the targetdensity-dpi support
28762         https://bugs.webkit.org/show_bug.cgi?id=55142
28763
28764         Test the viewport meta tag feature targetdensity-dpi by
28765         adding extra arguments to dumpConfigurationForViewport
28766
28767         Add the new arguments to the shared header and all
28768         implementation.
28769
28770         * DumpRenderTree/LayoutTestController.cpp:
28771         (dumpConfigurationForViewportCallback):
28772         * DumpRenderTree/LayoutTestController.h:
28773         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
28774         (LayoutTestController::dumpConfigurationForViewport):
28775         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
28776         (LayoutTestController::dumpConfigurationForViewport):
28777         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
28778         (LayoutTestController::dumpConfigurationForViewport):
28779         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
28780         (LayoutTestController::dumpConfigurationForViewport):
28781         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
28782         (LayoutTestController::dumpConfigurationForViewport):
28783         * DumpRenderTree/qt/LayoutTestControllerQt.h:
28784
28785 2011-02-26  Sheriff Bot  <webkit.review.bot@gmail.com>
28786
28787         Unreviewed, rolling out r79764.
28788         http://trac.webkit.org/changeset/79764
28789         https://bugs.webkit.org/show_bug.cgi?id=55295
28790
28791         "broke Chromium builds" (Requested by rniwa on #webkit).
28792
28793         * DumpRenderTree/chromium/TestShell.cpp:
28794         (TestShell::resetTestController):
28795         * DumpRenderTree/gtk/DumpRenderTree.cpp:
28796         (resetDefaultsToConsistentValues):
28797         * DumpRenderTree/mac/DumpRenderTree.mm:
28798         (resetWebViewToConsistentStateBeforeTesting):
28799         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
28800         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
28801         * DumpRenderTree/win/DumpRenderTree.cpp:
28802         (resetWebViewToConsistentStateBeforeTesting):
28803         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
28804         (WTR::InjectedBundlePage::reset):
28805
28806 2011-02-26  Yongjun Zhang  <yongjun_zhang@apple.com>
28807
28808         Reviewed by David Kilzer.
28809
28810         https://bugs.webkit.org/show_bug.cgi?id=48781
28811
28812         Add a resource load delegate method to query if WebCore should paint the default broken image for failed images.
28813
28814         Add a new resource load client method (shouldPaintBrokenImage).  WebKit client can decide if WebCore
28815         should paint the default broken image when an image fails to load or decode.
28816
28817         * DumpRenderTree/LayoutTestController.cpp:
28818         (LayoutTestController::LayoutTestController):
28819         (setShouldPaintBrokenImageCallback):
28820         (LayoutTestController::staticFunctions):
28821         (LayoutTestController::setShouldPaintBrokenImage):
28822         * DumpRenderTree/LayoutTestController.h:
28823         (LayoutTestController::shouldPaintBrokenImage):
28824         * DumpRenderTree/mac/ResourceLoadDelegate.mm:
28825         (-[ResourceLoadDelegate webView:shouldPaintBrokenImageForURL:]):
28826
28827 2011-02-26  Vsevolod Vlasov  <vsevik@chromium.org>
28828
28829         Reviewed by Pavel Feldman.
28830
28831         DumpRenderTree should reset frame opener between tests.
28832         https://bugs.webkit.org/show_bug.cgi?id=54874
28833
28834         Added clearOpener() calls between tests.
28835         Some tests are using window.open() to change current window location.
28836         This changes window.opener property whose nullity is checked before
28837         showing xml viewer.
28838
28839         * DumpRenderTree/chromium/TestShell.cpp:
28840         (TestShell::resetTestController):
28841         * DumpRenderTree/gtk/DumpRenderTree.cpp:
28842         (resetDefaultsToConsistentValues):
28843         * DumpRenderTree/mac/DumpRenderTree.mm:
28844         (resetWebViewToConsistentStateBeforeTesting):
28845         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
28846         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
28847         * DumpRenderTree/win/DumpRenderTree.cpp:
28848         (resetWebViewToConsistentStateBeforeTesting):
28849         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
28850         (WTR::InjectedBundlePage::reset):
28851
28852 2011-02-25  Lukasz Slachciak  <l.slachciak@samsung.com>
28853
28854         Reviewed by Kenneth Rohde Christiansen.
28855
28856         [EFL] Added EWebLauncher command line option to turn on/off frame flattening
28857         https://bugs.webkit.org/show_bug.cgi?id=54809
28858
28859         * EWebLauncher/main.c: Handling -f option added.
28860         (on_key_down):
28861         (browserCreate):
28862         (main):
28863
28864 2011-02-25  Sheriff Bot  <webkit.review.bot@gmail.com>
28865
28866         Unreviewed, rolling out r79324.
28867         http://trac.webkit.org/changeset/79324
28868         https://bugs.webkit.org/show_bug.cgi?id=55260
28869
28870         Leopard Debug is once again running out of address space while
28871         linking (Requested by aroben on #webkit).
28872
28873         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
28874         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
28875
28876 2011-02-25  Tony Chang  <tony@chromium.org>
28877
28878         Reviewed by James Robinson.
28879
28880         Clean up temp files left by crashing DRTs on chromium-win
28881         https://bugs.webkit.org/show_bug.cgi?id=55182
28882
28883         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
28884         * BuildSlaveSupport/chromium/remove-crash-logs: Added.
28885
28886 2011-02-25  Adam Roben  <aroben@apple.com>
28887
28888         Double-check that modifying the registry worked
28889
28890         On Windows Vista/7 with UAC enabled, regtool will fail to modify the registry, but it will
28891         still return a successful exit code. So we double-check here that the value we tried to
28892         write to the registry was really written.
28893
28894         Fixes <http://webkit.org/b/55227> old-run-webkit-tests falsely claims to be saving crash
28895         logs on Windows Vista/7
28896
28897         Reviewed by Darin Adler.
28898
28899         * Scripts/old-run-webkit-tests:
28900         (writeRegistryString): After trying to write the string value, check that it was really
28901         written.
28902
28903 2011-02-25  Adam Roben  <aroben@apple.com>
28904
28905         Set the 32-bit post-mortem debugger on 64-bit OSes
28906
28907         The post-mortem debugger is controlled by some registry values in
28908         HKLM\SOFTWARE\Microsoft\Windows NT\CurrentVersion\AeDebug. On 64-bit OSes, you can either
28909         set the values in that key, which will be used for 64-bit apps, or in the equivalent key
28910         under Wow6432Node, which will be used for 32-bit apps. Since DumpRenderTree and
28911         WebKitTestRunner are 32-bit, we want to use the latter location.
28912
28913         Fixes <http://webkit.org/b/55225> old-run-webkit-tests sets the post-mortem debugger for
28914         64-bit apps, but should set it for 32-bit apps instead
28915
28916         Reviewed by Darin Adler.
28917
28918         * Scripts/old-run-webkit-tests:
28919         (readRegistryString):
28920         (writeRegistryString):
28921         Pass --wow32 to regtool so it will get/set values beneath Wow6432Node, which is where
28922         Windows looks for the post-mortem debugger values when a 32-bit app crashes on a 64-bit OS.
28923         Passing --wow32 on a 32-bit OS will not have any effect.
28924
28925 2011-02-25  Adam Roben  <aroben@apple.com>
28926
28927         Clean up code to set up crash log saving on Windows
28928
28929         Fixes <http://webkit.org/b/55224> old-run-webkit-test's Windows registry code is ugly
28930
28931         Reviewed by David Kilzer.
28932
28933         * Scripts/old-run-webkit-tests:
28934         (readRegistryString):
28935         (writeRegistryString):
28936         Moved code to read strings from and write strings to the registry here...
28937
28938         (setUpWindowsCrashLogSaving): ...from here.
28939
28940 2011-02-25  Adam Roben  <aroben@apple.com>
28941
28942         Work around Cygwin's crash-suppression behavior
28943
28944         Cygwin calls ::SetErrorMode(SEM_FAILCRITICALERRORS), which any processes it launches will
28945         inherit. This is bad for testing/debugging, as it causes the post-mortem debugger not to be
28946         invoked. (Cygwin does this because it makes crashes more UNIX-y.) We reset the error mode
28947         when our test apps launch to work around Cygwin's behavior.
28948
28949         Fixes <http://webkit.org/b/55222> Test apps crash silently (without invoking post-mortem
28950         debugger) when launched from Cygwin 1.7
28951
28952         Reviewed by Darin Adler.
28953
28954         * DumpRenderTree/win/DumpRenderTree.cpp:
28955         (main):
28956         * TestWebKitAPI/win/main.cpp:
28957         (main):
28958         * WebKitAPITest/main.cpp:
28959         (main):
28960         * WebKitTestRunner/win/TestControllerWin.cpp:
28961         (WTR::TestController::platformInitialize):
28962         Call ::SetErrorMode(0) to undo Cygwin's folly.
28963
28964 2011-02-24  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
28965
28966         Reviewed by Andreas Kling.
28967
28968         [Qt] Revert the support for QNAM affined to a different thread.
28969         https://bugs.webkit.org/show_bug.cgi?id=55149
28970
28971         Qt 4.8 will have QNAM use its own thread internally by default,
28972         no need to keep this complexity in WebKit.
28973
28974         This mainly reverts:
28975         http://trac.webkit.org/changeset/73710
28976         http://trac.webkit.org/changeset/73712
28977
28978         * QtTestBrowser/launcherwindow.cpp:
28979         (LauncherWindow::~LauncherWindow):
28980         (LauncherWindow::initializeView):
28981         (LauncherWindow::createChrome):
28982         * QtTestBrowser/launcherwindow.h:
28983         (WindowOptions::WindowOptions):
28984         * QtTestBrowser/webpage.cpp:
28985         (WebPage::WebPage):
28986         * QtTestBrowser/webpage.h:
28987
28988 2011-02-24  Adam Barth  <abarth@webkit.org>
28989
28990         Reviewed by Eric Seidel.
28991
28992         Add a script for creating gypi files
28993         https://bugs.webkit.org/show_bug.cgi?id=55113
28994
28995         This is a trivial script to create gypi files from the source code in
28996         the current directory.
28997
28998         * Scripts/make-gypi: Added.
28999
29000 2011-02-24  Sheriff Bot  <webkit.review.bot@gmail.com>
29001
29002         Unreviewed, rolling out r79628.
29003         http://trac.webkit.org/changeset/79628
29004         https://bugs.webkit.org/show_bug.cgi?id=55195
29005
29006         "broke chromium win tests" (Requested by dpranke on #webkit).
29007
29008         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
29009         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
29010
29011 2011-02-24  Dirk Pranke  <dpranke@chromium.org>
29012
29013         Reviewed by Mihai Parparita.
29014
29015         rebaseline-chromium-webkit-tests can't handle GPU variants
29016         https://bugs.webkit.org/show_bug.cgi?id=55002
29017
29018         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
29019         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
29020
29021 2011-02-24  Dirk Pranke  <dpranke@chromium.org>
29022
29023         Reviewed by Mihai Parparita.
29024
29025         nrwt: log command used to start web server to --verbose
29026
29027         https://bugs.webkit.org/show_bug.cgi?id=55188
29028
29029         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
29030         * Scripts/webkitpy/layout_tests/port/http_server.py:
29031
29032 2011-02-24  Victoria Kirst  <vrk@google.com>
29033
29034         Reviewed by Mihai Parparita.
29035
29036         [chromium] Move media-file.js and video-test.js back to media/ for UILayoutTests
29037         https://bugs.webkit.org/show_bug.cgi?id=55089
29038
29039         This adds an alias into the media/ directory so that http/tests/media
29040         tests can access the media resources when running in an httpd process.
29041
29042         * Scripts/webkitperl/httpd.pm:
29043         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
29044         * Scripts/webkitpy/layout_tests/port/http_server.py:
29045
29046 2011-02-24  Dirk Pranke  <dpranke@chromium.org>
29047
29048         Reviewed by Mihai Parparita.
29049
29050         nrwt: enable multiple processes by default on Chromium Win.
29051
29052         https://bugs.webkit.org/show_bug.cgi?id=55163
29053
29054         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
29055         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
29056
29057 2011-02-24  Sam Weinig  <sam@webkit.org>
29058
29059         Attempt to fix Qt.
29060
29061         * WebKitTestRunner/TestInvocation.cpp:
29062
29063 2011-02-24  James Robinson  <jamesr@chromium.org>
29064
29065         Unreviewed, rolling out r79598.
29066         http://trac.webkit.org/changeset/79598
29067         https://bugs.webkit.org/show_bug.cgi?id=55089
29068
29069         Appears to be causing trouble with httpd on windows
29070
29071         * Scripts/webkitperl/httpd.pm:
29072         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
29073         * Scripts/webkitpy/layout_tests/port/http_server.py:
29074
29075 2011-02-24  Sam Weinig  <sam@webkit.org>
29076
29077         Reviewed by Darin Adler.
29078
29079         Add basic pixel testing support to WebKitTestRunner
29080         https://bugs.webkit.org/show_bug.cgi?id=55161
29081
29082         * Scripts/old-run-webkit-tests:
29083         Don't print color space information when using WebKitTestRunner since
29084         we don't need to change the screen.
29085
29086         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
29087         (WTR::InjectedBundle::initialize):
29088         Add call to initializePlatformDefaults.
29089
29090         (WTR::InjectedBundle::done):
29091         Return a dictionary on completion with both the text output and pixel dump override.
29092
29093         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
29094         Add declaration for initializePlatformDefaults.
29095
29096         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
29097         (WTR::LayoutTestController::LayoutTestController):
29098         (WTR::LayoutTestController::dumpAsText):
29099         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
29100         (WTR::LayoutTestController::shouldDumpPixels):
29101         Make setting dumpAsText override the dump pixels preference.
29102
29103         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm: Added.
29104         (WTR::InjectedBundle::initializePlatformDefaults):
29105         Set platform defaults matching DumpRenderTree.
29106
29107         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp: Added.
29108         (WTR::InjectedBundle::initializePlatformDefaults):
29109         * WebKitTestRunner/InjectedBundle/win/InjectedBundleWin.cpp: Added.
29110         (WTR::InjectedBundle::initializePlatformDefaults):
29111         Add stubs for initializePlatformDefaults.
29112
29113         * WebKitTestRunner/TestController.cpp:
29114         (WTR::TestController::runTest):
29115         (WTR::TestController::runTestingServerLoop):
29116         (WTR::TestController::run):
29117         Parse expected pixel results out of stdin and setup the invocation with it.
29118
29119         * WebKitTestRunner/TestInvocation.cpp:
29120         (WTR::TestInvocation::TestInvocation):
29121         Initialize new members.
29122
29123         (WTR::TestInvocation::~TestInvocation):
29124         (WTR::TestInvocation::setIsPixelTest):
29125         (WTR::sizeWebViewForCurrentTest):
29126         (WTR::TestInvocation::invoke):
29127         (WTR::TestInvocation::dump):
29128         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
29129         * WebKitTestRunner/TestInvocation.h:
29130         Dump pixels in addition to text by calling dumpPixelsAndCompareWithExpected.
29131
29132         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
29133         (WTR::PlatformWebView::PlatformWebView):
29134         Use -[NSColorSpace genericRGBColorSpace] instead of the main screen color space.
29135
29136         * WebKitTestRunner/mac/TestInvocationMac.mm: Added.
29137         (WTR::createCGContextFromPlatformView):
29138         (WTR::computeMD5HashStringForContext):
29139         (WTR::dumpBitmap):
29140         (WTR::forceRepaintFunction):
29141         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
29142         Add mac specific pixel dumping support. Right now, we always pull pixels
29143         from the window server.
29144
29145         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
29146         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
29147         * WebKitTestRunner/qt/WebKitTestRunner.pro:
29148         * WebKitTestRunner/win/InjectedBundle.vcproj:
29149         * WebKitTestRunner/win/WebKitTestRunner.vcproj:
29150         Add new files.
29151
29152 2011-02-24  Chang Shu  <cshu@webkit.org>
29153
29154         Reviewed by Kenneth Rohde Christiansen.
29155
29156         [Qt] Launch MiniBrowser if the test run is WebKit2.
29157         https://bugs.webkit.org/show_bug.cgi?id=55145
29158
29159         * Scripts/old-run-webkit-tests:
29160         * Scripts/run-launcher:
29161         * Scripts/webkitdirs.pm:
29162
29163 2011-02-24  Victoria Kirst  <vrk@google.com>
29164
29165         Reviewed by Mihai Parparita.
29166
29167         [chromium] Move media-file.js and video-test.js back to media/ for UILayoutTests
29168         https://bugs.webkit.org/show_bug.cgi?id=55089
29169
29170         This adds an alias into the media/ directory so that http/tests/media
29171         tests can access the media resources when running in an httpd process.
29172
29173         * Scripts/webkitperl/httpd.pm:
29174         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
29175         * Scripts/webkitpy/layout_tests/port/http_server.py:
29176
29177 2011-02-24  Andrew Wilson  <atwilson@chromium.org>
29178
29179         Unreviewed, rolling out r79570.
29180         http://trac.webkit.org/changeset/79570
29181         https://bugs.webkit.org/show_bug.cgi?id=54874
29182
29183         Breaks chromium build because glue/mocks/mock_web_frame.h/cc
29184         was not updated
29185
29186         * DumpRenderTree/chromium/TestShell.cpp:
29187         (TestShell::resetTestController):
29188         * DumpRenderTree/gtk/DumpRenderTree.cpp:
29189         (resetDefaultsToConsistentValues):
29190         * DumpRenderTree/mac/DumpRenderTree.mm:
29191         (resetWebViewToConsistentStateBeforeTesting):
29192         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
29193         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
29194         * DumpRenderTree/win/DumpRenderTree.cpp:
29195         (resetWebViewToConsistentStateBeforeTesting):
29196         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
29197         (WTR::InjectedBundlePage::reset):
29198
29199 2011-02-24  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
29200
29201         Reviewed by Laszlo Gombos.
29202
29203         [Qt] add android 2.2 user agent to QtTestBrowser
29204         https://bugs.webkit.org/show_bug.cgi?id=55085
29205
29206         Credit to Forrest Hodgkins to find right user-agent which works for youtube.
29207         * QtTestBrowser/useragentlist.txt:
29208
29209 2011-02-24  Adam Roben  <aroben@apple.com>
29210
29211         Change FrameLoadDelegate to support any number of delegates with delayed work to process
29212
29213         This makes our behavior match Mac more closely, and allows us to remove an incorrect
29214         assertion that was firing during some tests. (The assertion was claiming that there was
29215         never more than one delegate with delayed work to process, but that was not the case.)
29216
29217         Fixes <http://webkit.org/b/55146> Assertion failure in FrameLoadDelegate::locationChangeDone
29218         when running http/tests/navigation/back-twice-without-commit.html
29219
29220         Reviewed by Eric Carlson.
29221
29222         * DumpRenderTree/win/FrameLoadDelegate.cpp:
29223         (delegatesWithDelayedWork): Added. Returns all FrameLoadDelegates that have delayed work to
29224         process. A single delegate may appear in this Vector more than once (just as, on Mac, a
29225         single delegate may have multiple performSelector requests).
29226         (processWorkTimer): Pass the HWND to ::KillTimer, for pedantic brownie points. Added an
29227         assertion that the timer firing is the shared process work timer. Instead of using the
29228         single, global "delegate waiting for timer" delegate, give all delegates that have delayed
29229         work to process a chance to process their work.
29230         (FrameLoadDelegate::locationChangeDone): If we don't already have an active timer for
29231         processing delayed work, create one. Then add ourselves to the delegatesWithDelayedWork
29232         Vector so our processWork function will be called when the timer fires.
29233
29234 2011-02-24  Vsevolod Vlasov  <vsevik@chromium.org>
29235
29236         Reviewed by Alexey Proskuryakov.
29237
29238         DumpRenderTree should reset frame opener between tests.
29239         https://bugs.webkit.org/show_bug.cgi?id=54874
29240
29241         Added clearOpener() calls between tests.
29242         Some tests are using window.open() to change current window location.
29243         This changes window.opener property whose nullity is checked before
29244         showing xml viewer.
29245
29246         * DumpRenderTree/chromium/TestShell.cpp:
29247         (TestShell::resetTestController):
29248         * DumpRenderTree/gtk/DumpRenderTree.cpp:
29249         (resetDefaultsToConsistentValues):
29250         * DumpRenderTree/mac/DumpRenderTree.mm:
29251         (resetWebViewToConsistentStateBeforeTesting):
29252         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
29253         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
29254         * DumpRenderTree/win/DumpRenderTree.cpp:
29255         (resetWebViewToConsistentStateBeforeTesting):
29256         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
29257         (WTR::InjectedBundlePage::reset):
29258
29259 2011-02-24  Adam Roben  <aroben@apple.com>
29260
29261         Don't trigger Windows builds for changes to WebCore.exp.in or the top-level GNUmakefile.am
29262
29263         Fixes <http://webkit.org/b/55144> Windows builders should not have built r79343 or r79440,
29264         but did
29265
29266         Reviewed by Anders Carlsson.
29267
29268         * Scripts/webkitpy/common/config/build.py:
29269         (_should_file_trigger_build): Loosened the pattern for GNUmakefile.am to include the
29270         top-level GNUmakefile.am, and loosened the pattern for .exp files to include WebCore.exp.in.
29271
29272         * Scripts/webkitpy/common/config/build_unittest.py:
29273         (ShouldBuildTest): Added some more tests.
29274
29275 2011-02-24  Andras Becsi  <abecsi@webkit.org>
29276
29277         Reviewed by Laszlo Gombos.
29278
29279         [Qt] MinGW build fails to link
29280         https://bugs.webkit.org/show_bug.cgi?id=55050
29281
29282         Prepend the libraries of subcomponents instead of appending them
29283         to fix the library order according to the dependency of the libraries
29284
29285         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
29286         * WebKitTestRunner/qt/WebKitTestRunner.pro:
29287
29288 2011-02-24  Gabor Loki  <loki@webkit.org>
29289
29290         Reviewed by Csaba Osztrogonác.
29291
29292         [Qt] Remove ARMv5 Release buildbot.
29293
29294         Remove ARMv5 Release buildbot, because there is very little difference
29295         between ARMv5 and ARMv7 Release bots, and there is also very little
29296         interest in ARMv5 build slave (compared to ARMv7).
29297
29298         * BuildSlaveSupport/build.webkit.org-config/config.json:
29299         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
29300
29301 2011-02-22  Ojan Vafai  <ojan@chromium.org>
29302
29303         Reviewed by Tony Chang.
29304
29305         fix revision number finding when creating git patches
29306         https://bugs.webkit.org/show_bug.cgi?id=55024
29307
29308         HEAD~n does not walk merge commits correctly. git log however does.
29309
29310         * Scripts/webkitpy/common/checkout/scm.py:
29311         * Scripts/webkitpy/common/checkout/scm_unittest.py:
29312
29313 2011-02-23  Kenneth Russell  <kbr@google.com>
29314
29315         Unreviewed, rolling out r79387.
29316         https://bugs.webkit.org/show_bug.cgi?id=54885
29317
29318         Reapply the code changes from r79038 since they weren't the root
29319         cause of the test timeouts.
29320
29321         * DumpRenderTree/chromium/WebViewHost.cpp:
29322         (WebViewHost::reset):
29323
29324 2011-02-23  Siddharth Mathur  <siddharth.mathur@nokia.com>
29325
29326         Reviewed by Laszlo Gombos.
29327
29328         [Qt] Make sure Symbian binary UIDs are unique
29329
29330         * MiniBrowser/qt/MiniBrowser.pro: Resolve the UID collision with
29331         QtTestBrowser.pro by changing the UID.
29332
29333 2011-02-23  Patrick Gansterer  <paroga@webkit.org>
29334
29335         Reviewed by Darin Adler.
29336
29337         Rename PLATFORM(CF) to USE(CF)
29338         https://bugs.webkit.org/show_bug.cgi?id=53540
29339
29340         * DumpRenderTree/config.h:
29341
29342 2011-02-23  Dirk Pranke  <dpranke@chromium.org>
29343
29344         Reviewed by Ojan Vafai.
29345
29346         nrwt: turn on multiprocessing on chromium_linux, clean up unit tests for worker model defaults
29347         https://bugs.webkit.org/show_bug.cgi?id=55016
29348
29349         * Scripts/webkitpy/layout_tests/port/base.py:
29350         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
29351         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py: Added.
29352         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
29353         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
29354         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
29355         * Scripts/webkitpy/layout_tests/port/mac.py:
29356         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
29357         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
29358         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
29359
29360 2011-02-23  Tony Chang  <tony@chromium.org>
29361
29362         Reviewed by Kent Tamura.
29363
29364         [chromium] DumpRenderTree --check-layout-test-sys-deps leaks 4 temp dirs
29365         https://bugs.webkit.org/show_bug.cgi?id=55004
29366
29367         * DumpRenderTree/chromium/DumpRenderTree.cpp:
29368         (WebKitSupportTestEnvironment::WebKitSupportTestEnvironment):
29369         (WebKitSupportTestEnvironment::~WebKitSupportTestEnvironment):
29370         (main): Use a C++ object to ensure that TearDownTestEnvironment
29371             gets called when --check-layout-test-sys-deps is called
29372
29373 2011-02-23  Gopal Raghavan  <gopal.1.raghavan@nokia.com>
29374
29375         Reviewed by Anders Carlsson.
29376
29377         [Qt] [Qt] check webkit style fails in qt/tests
29378         https://bugs.webkit.org/show_bug.cgi?id=54998
29379
29380         check-webkit-style was complaining about missing config.h in WebKit/qt/test/qdeclarativewebview. This patch was originally part of 50222. Reviewer suggested to split and create a separate patch for checker.
29381
29382         * Scripts/webkitpy/style/checker.py:
29383
29384 2011-02-22  Vsevolod Vlasov  <vsevik@chromium.org>
29385
29386         Reviewed by Pavel Feldman.
29387
29388         DumpRenderTree should dump as text if test path contains "dumpAsText/"
29389         https://bugs.webkit.org/show_bug.cgi?id=54873
29390
29391         * DumpRenderTree/chromium/LayoutTestController.h:
29392         (LayoutTestController::setShouldDumpAsText):
29393         (LayoutTestController::setShouldGeneratePixelResults):
29394         * DumpRenderTree/chromium/TestShell.cpp:
29395         (TestShell::runFileTest):
29396         * DumpRenderTree/gtk/DumpRenderTree.cpp:
29397         (shouldDumpAsText):
29398         (dump):
29399         (runTest):
29400         * DumpRenderTree/mac/DumpRenderTree.mm:
29401         (shouldDumpAsText):
29402         (runTest):
29403         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
29404         (WebCore::isDumpAsTextTest):
29405         (WebCore::DumpRenderTree::open):
29406         * DumpRenderTree/win/DumpRenderTree.cpp:
29407         (shouldDumpAsText):
29408         (runTest):
29409         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
29410         (WTR::InjectedBundlePage::dump):
29411
29412 2011-02-22  Kenneth Russell  <kbr@google.com>
29413
29414         Reviewed by James Robinson.
29415
29416         Many tests started timing out on the chrome mac canaries around r79035
29417         https://bugs.webkit.org/show_bug.cgi?id=54885
29418
29419         Revert the code changes, but not the Chromium DEPS roll, from
29420         r79038 to see whether this clears up the timing-out tests. Was not
29421         able to reproduce the timeouts locally with a Release mode build
29422         of DRT (on Snow Leopard).
29423
29424         * DumpRenderTree/chromium/WebViewHost.cpp:
29425         (WebViewHost::reset):
29426
29427 2011-02-22  Dirk Pranke  <dpranke@chromium.org>
29428
29429         Reviewed by James Robinson.
29430
29431         nrwt: --platform chromium-gpu doesn't work with --worker-model=processes
29432         https://bugs.webkit.org/show_bug.cgi?id=55009
29433
29434         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
29435
29436 2011-02-18  Mihai Parparita  <mihaip@chromium.org>
29437
29438         Reviewed by Ojan Vafai.
29439
29440         Default to 500 for --exit-after-n-failures
29441         https://bugs.webkit.org/show_bug.cgi?id=54773
29442
29443         Make 500 the default value for --exit-after-n-failures (ORWT bots
29444         already default to this as of r75726). Similarly, change the default for
29445         --exit-after-n-crashes-or-timeouts to 20.
29446
29447         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
29448
29449 2011-02-02  Martin Robinson  <mrobinson@igalia.com>
29450
29451         Reviewed by Xan Lopez.
29452
29453         [GTK] DRT needs an implementation of LayoutTestController.setSelectTrailingWhitespace
29454         https://bugs.webkit.org/show_bug.cgi?id=53603
29455
29456         Add support for LayoutTestController.setSelectTrailingWhitespace by calling through
29457         to DumpRenderTreeSupportGtk for this functionality.
29458
29459         * DumpRenderTree/gtk/DumpRenderTree.cpp:
29460         (resetDefaultsToConsistentValues): Always call LayoutTestController.setSelectTrailingWhitespace to
29461         false when reseting values between tests.
29462         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
29463         (LayoutTestController::setSelectTrailingWhitespaceEnabled): Call LayoutTestController.setSelectTrailingWhitespace
29464         with the appropriate value.
29465
29466 2011-02-22  Adam Roben  <aroben@apple.com>
29467
29468         Don't trigger a Windows build when the top-level ChangeLog changes
29469
29470         r79320 is an example of a change that shouldn't have built on Windows, but did, because of
29471         this bug.
29472
29473         Fixes <http://webkit.org/b/54957> Windows bots build when top-level ChangeLog changes, but
29474         shouldn't
29475
29476         Reviewed by Anders Carlsson.
29477
29478         * Scripts/webkitpy/common/config/build.py:
29479         (_should_file_trigger_build): Modified the ChangeLog pattern to match the top-level
29480         ChangeLog, and re-sorted it in the list.
29481
29482         * Scripts/webkitpy/common/config/build_unittest.py:
29483         (ShouldBuildTest): Added a test that ChangeLogs don't trigger builds.
29484
29485 2011-02-22  Sheriff Bot  <webkit.review.bot@gmail.com>
29486
29487         Unreviewed, rolling out r78691.
29488         http://trac.webkit.org/changeset/78691
29489         https://bugs.webkit.org/show_bug.cgi?id=54956
29490
29491         Leopard Debug is successfully building again (Requested by
29492         aroben on #webkit).
29493
29494         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
29495         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
29496
29497 2011-02-22  Ilya Tikhonovsky  <loislo@chromium.org>
29498
29499         Reviewed by Pavel Feldman.
29500
29501         Web Inspector: flakyness of inspector tests.
29502         https://bugs.webkit.org/show_bug.cgi?id=54729
29503
29504         As far as we have the protocol with sequence numbers we can simplify test harness support
29505         and drop out chomium specific methods.
29506
29507         * DumpRenderTree/DumpRenderTree.gypi:
29508         * DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
29509         (DRTDevToolsAgent::sendMessageToInspectorFrontend):
29510         (DRTDevToolsAgent::runtimePropertyChanged):
29511         (DRTDevToolsAgent::asyncCall):
29512         (DRTDevToolsAgent::call):
29513         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
29514         * DumpRenderTree/chromium/DRTDevToolsCallArgs.cpp: Removed.
29515         * DumpRenderTree/chromium/DRTDevToolsCallArgs.h: Removed.
29516         * DumpRenderTree/chromium/DRTDevToolsClient.cpp:
29517         (DRTDevToolsClient::sendMessageToBackend):
29518         (DRTDevToolsClient::asyncCall):
29519         (DRTDevToolsClient::call):
29520         * DumpRenderTree/chromium/DRTDevToolsClient.h:
29521
29522 2011-02-22  Andras Becsi  <abecsi@webkit.org>
29523
29524         Reviewed by Laszlo Gombos.
29525
29526         [Qt] Redesign the build system
29527         https://bugs.webkit.org/show_bug.cgi?id=51339
29528
29529         Part 2.
29530
29531         Build WebCore as a static library, compile the WebKit API and WebKit2 API
29532         in a final step and link to WebKit2, WebCore and JSC libraries to fix
29533         linking issues resulting from stripped away symbols.
29534
29535         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro: Remove superfluous includepaths.
29536
29537 2011-02-21  Lukasz Slachciak  <l.slachciak@samsung.com>
29538
29539         Reviewed by Antonio Gomes.
29540
29541         [EFL] Restored spatial navigation option in EWebLauncher since #18662 bug is closed
29542         https://bugs.webkit.org/show_bug.cgi?id=54806
29543
29544         * EWebLauncher/main.c:
29545         (on_key_down): calling ewk_view_setting_spatial_navigation_set when "F12" pressed added.
29546
29547 2011-02-21  Dirk Pranke  <dpranke@chromium.org>
29548
29549         Reviewed by Kent Tamura.
29550
29551         nrwt: revert behavior on 10.5 to "old-threads". It's possible
29552         that r79062 has introduced a bunch of flakiness, although I
29553         can't explain why right now. This change should restore the
29554         old behavior, and then we can look at the build logs to see
29555         if the tests stop timing out or being flaky.
29556
29557         https://bugs.webkit.org/show_bug.cgi?id=54925
29558
29559         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
29560         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
29561         * Scripts/webkitpy/layout_tests/port/mac.py:
29562         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
29563
29564 2011-02-21  Dirk Pranke  <dpranke@chromium.org>
29565
29566         Reviewed by Dimitri Glazkov.
29567
29568         nrwt: fix typo causing us to try and repeatedly shut down the servers
29569         https://bugs.webkit.org/show_bug.cgi?id=54904
29570
29571         * Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py:
29572
29573 2011-02-21  Alejandro G. Castro  <alex@igalia.com>
29574
29575         Reviewed by Csaba Osztrogonác.
29576
29577         REGRESSION(r79157):
29578         plugins/get-url-with-javascript-destroying-plugin.html make the
29579         next test crash
29580         https://bugs.webkit.org/show_bug.cgi?id=54863
29581
29582         * GNUmakefile.am:
29583
29584 2011-02-20  Dirk Pranke  <dpranke@chromium.org>
29585
29586         Reviewed by Ojan Vafai.
29587
29588         test-webkitpy: stop skipping a bunch of tests on win32
29589         https://bugs.webkit.org/show_bug.cgi?id=54788
29590
29591         * Scripts/webkitpy/common/net/testoutputset_unittest.py:
29592         * Scripts/webkitpy/test/main.py:
29593
29594 2011-02-20  Dirk Pranke  <dpranke@chromium.org>
29595
29596         Reviewed by Ojan Vafai.
29597
29598         Fix various bugs related to handling of dir separators and real
29599         vs. fake filesystems that were causing test-webkitpy to fail
29600         tests on win32. These bugs have a variety of causes but are
29601         lumped together because they should be easy to review.
29602
29603         https://bugs.webkit.org/show_bug.cgi?id=54700
29604
29605         * Scripts/webkitpy/common/net/testoutput.py:
29606           Here we were using os.path.sep but not escaping it prior to
29607           using in a regexp, which  caused bad things on win32.
29608         * Scripts/webkitpy/common/net/testoutput_unittest.py:
29609           Same here.
29610         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
29611           Here we were using the native filesystem instead of the mock
29612           filesystem, making some tests unnecessarily platform dependent.
29613           Also we weren't setting the default configuration for the
29614           TestChromiumWinPort.
29615         * Scripts/webkitpy/layout_tests/update_webgl_conformance_tests.py:
29616           Here we *shouldn't* use os.path.join(), just so the code becomes
29617           less platform dependent.
29618         * Scripts/webkitpy/style/checker.py:
29619           Here we should be using os.path.sep instead of "/"; the tests
29620           require the native filesystem.
29621         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
29622           Here the underlying code is using os.path.sep, so we probably
29623           should as well.
29624         * Scripts/webkitpy/style/test_expectations_unittest.py:
29625           Here we should be using the filesystem hanging of the port
29626           object (which is the mocked filesystem) rather than the native
29627           one.
29628
29629 2011-02-20  Anders Carlsson  <andersca@apple.com>
29630
29631         Reviewed by Maciej Stachowiak.
29632
29633         Crash when a plug-in requests a javascript: url that destroys the plug-in
29634         https://bugs.webkit.org/show_bug.cgi?id=54837
29635         <rdar://problem/9005475>
29636
29637         Add new plug-in test.
29638
29639         * DumpRenderTree/DumpRenderTree.gypi:
29640         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
29641         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
29642         (PluginTest::NPN_GetURL):
29643         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
29644         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLWithJavaScriptURLDestroyingPlugin.cpp: Added.
29645         (GetURLWithJavaScriptURLDestroyingPlugin::GetURLWithJavaScriptURLDestroyingPlugin):
29646         (GetURLWithJavaScriptURLDestroyingPlugin::NPP_New):
29647         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
29648         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
29649
29650 2011-02-19  Patrick Gansterer  <paroga@webkit.org>
29651
29652         Reviewed by Andreas Kling.
29653
29654         Make WinCE a core builder
29655         https://bugs.webkit.org/show_bug.cgi?id=54801
29656
29657         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
29658         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
29659
29660 2011-02-19  Lukasz Slachciak  <l.slachciak@samsung.com>
29661
29662         Reviewed by Andreas Kling.
29663
29664         Fixed problem with launching EWebLauncher with "run-launcher --efl". Incorrect path referring to non-existing .libs directory ws set.
29665         https://webkit.org/b/54778
29666
29667         * Scripts/webkitdirs.pm:
29668
29669 2011-02-18  David Levin  <levin@chromium.org>
29670
29671         Reviewed by Eric Seidel.
29672
29673         check-webkit-style: Misses brace style error when the line with the { has a } in it.
29674         https://bugs.webkit.org/show_bug.cgi?id=54769
29675
29676         * Scripts/webkitpy/style/checkers/cpp.py: Fixed the check for the close brace
29677           to only look after the last open brace, so that the open brace in this line
29678           "} else {" will still be able to trigger the error.
29679         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added related unit tests.
29680
29681 2011-02-18  Zan Dobersek  <zandobersek@gmail.com>
29682
29683         Reviewed by Martin Robinson.
29684
29685         [Gtk] Re-enable meter tag support
29686         https://bugs.webkit.org/show_bug.cgi?id=54762
29687
29688         Meter tag is supported on the Gtk port, enable it by default.
29689
29690         * Scripts/build-webkit:
29691
29692 2011-02-18  David Levin  <levin@chromium.org>
29693
29694         Reviewed by Eric Seidel.
29695
29696         check-webkit-style falsely complains about WebKitGTK+ public headers
29697         https://bugs.webkit.org/show_bug.cgi?id=54650
29698
29699         * Scripts/webkitpy/style/checker.py: Add some exceptions for the gtk files.
29700         * Scripts/webkitpy/style/checkers/cpp.py: Restrict the checks for WEBKIT_API
29701           to the chromium directory (and improved the checks slightly).
29702         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added corresponding tests.
29703
29704 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
29705
29706         Reviewed by James Robinson.
29707
29708         new-run-webkit-tests: enable multiple processes by default on
29709         mac. This change removes the artificial restrictions we placed
29710         on the # of child processes we used with the old threading
29711         model, and switches to the new message-based model and multiple
29712         processes, where available. If multiple processes are not
29713         available (Leopard / Python 2.5), then we use the 'inline' model
29714         instead of the 'old-threads' model on the 'mac' port or one
29715         process on the 'chromium-mac' port. We need additional testing
29716         to see if the new 'threads' model works reliably and is worth
29717         supporting, or if we should just live with things being slightly
29718         slow.
29719
29720         https://bugs.webkit.org/show_bug.cgi?id=54596
29721
29722         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
29723         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
29724         * Scripts/webkitpy/layout_tests/port/mac.py:
29725         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
29726         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
29727
29728 2011-02-18  Andrew Wilson  <atwilson@chromium.org>
29729
29730         Unreviewed, rolling out r79047.
29731         http://trac.webkit.org/changeset/79047
29732         https://bugs.webkit.org/show_bug.cgi?id=54596
29733
29734         Broke canary bots - please watch the canaries the next time
29735         you land this
29736
29737         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
29738         * Scripts/webkitpy/layout_tests/port/mac.py:
29739
29740 2011-02-18  Emil A Eklund  <eae@chromium.org>
29741
29742         Reviewed by Darin Adler.
29743
29744         Crash in EventHandler::sendContextMenuEventForKey
29745         https://bugs.webkit.org/show_bug.cgi?id=54495
29746
29747         Add support for the context menu key (VK_APPS) to EventSender for
29748         platforms that support the key.
29749
29750         * DumpRenderTree/chromium/EventSender.cpp:
29751         (EventSender::keyDown):
29752         * DumpRenderTree/gtk/EventSender.cpp:
29753         (keyDownCallback):
29754         * DumpRenderTree/qt/EventSenderQt.cpp:
29755         (EventSender::keyDown):
29756         * DumpRenderTree/win/EventSender.cpp:
29757         (keyDownCallback):
29758
29759 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
29760
29761         Reviewed by Ojan Vafai.
29762
29763         new-run-webkit-tests: enable multiple processes by default on mac
29764         https://bugs.webkit.org/show_bug.cgi?id=54596
29765
29766         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
29767         * Scripts/webkitpy/layout_tests/port/mac.py:
29768
29769 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
29770
29771         Reviewed by Adam Roben.
29772
29773         Skip pretty patch unit tests if ruby isn't installed.
29774
29775         https://bugs.webkit.org/show_bug.cgi?id=54699
29776
29777         * Scripts/webkitpy/common/prettypatch_unittest.py:
29778
29779 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
29780
29781         Reviewed by Adam Roben.
29782
29783         Skip mac.test_skipped_file_paths() when running on win32; the
29784         test has hardcoded '/' directory separators and since it's a
29785         port-specific test and we have coverage on other ports it's not
29786         worth it to make the test more generic.
29787
29788         https://bugs.webkit.org/show_bug.cgi?id=54698
29789
29790         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
29791
29792 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
29793
29794         Reviewed by Adam Roben.
29795
29796         Fix webkitpy.layout_tests.port.config_unittest to work on Win32.
29797
29798         https://bugs.webkit.org/show_bug.cgi?id=54703
29799
29800         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
29801
29802 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
29803
29804         Reviewed by Adam Roben.
29805
29806         Change the default port we're mocking out on win32 when we
29807         run mock_drt_unittest. Normally we'd default to the
29808         port-specific default, but that would be "win", which doesn't
29809         currently work.
29810
29811         https://bugs.webkit.org/show_bug.cgi?id=54708
29812
29813         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
29814
29815 2011-02-18  Kenneth Russell  <kbr@google.com>
29816
29817         Reviewed by James Robinson.
29818
29819         [chromium] Clean up initialization of minimum timer interval
29820         https://bugs.webkit.org/show_bug.cgi?id=54772
29821
29822         Use new GetForegroundTabTimerInterval function in webkit_support
29823         to reset the minimum timer interval before each test.
29824
29825         * DumpRenderTree/chromium/WebViewHost.cpp:
29826         (WebViewHost::reset):
29827
29828 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
29829
29830         Reviewed by Adam Roben.
29831
29832         r78494 fixed the chromium_win unittest failing under win32,
29833         but not cygwin. This fixes cygwin as well.
29834
29835         https://bugs.webkit.org/show_bug.cgi?id=54652
29836
29837         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
29838
29839 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
29840
29841         Reviewed by Ojan Vafai, Mihai Parparita.
29842
29843         new-run-webkit-tests: r78522 made it impossible to create a
29844         'chromium-win' or 'chromium-mac' port; they would always get the
29845         version tacked on to the end. It turns out that we actually need
29846         to be able to create these ports in order for
29847         rebaseline-chromium-webkit-tests to work correctly given the
29848         broken coverage of our layout bots. As soon as we get SL and
29849         Win 7 bots running reliably on the canaries, we can get rid of
29850         this change. Note that we only need to change the chromium_mac
29851         and win ports because linux doesn't yet use versions and that
29852         script doesn't work with non-chromium ports.
29853
29854         Also, this change fixes a bug where we were using
29855         "platform/chromium-win-win7" and
29856         "platform/chromium-mac-snowleopard" as the baseline dirs for
29857         Win 7 and SL; we should've been using the generic directories
29858         instead since they are the newest versions of the os's.
29859
29860         https://bugs.webkit.org/show_bug.cgi?id=54691
29861
29862         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
29863         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
29864         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
29865         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
29866         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
29867
29868 2011-02-16  MORITA Hajime  <morrita@google.com>
29869
29870         Reviewed by Kent Tamura.
29871
29872         [Chromium][DRT] some PostDelayedTask() accesses possibly dangling this pointer.
29873         http://webkit.org/b/53899
29874
29875         Replaced static invocation wrappers with MethodTask objects,
29876         which are removed on WebViewHost destructor.
29877
29878         * DumpRenderTree/chromium/WebViewHost.cpp:
29879         (WebViewHost::requestCheckingOfText):
29880         (WebViewHost::scheduleAnimation):
29881         (WebViewHost::closeWidgetSoon):
29882         * DumpRenderTree/chromium/WebViewHost.h:
29883         (WebViewHost::HostMethodTask::HostMethodTask):
29884         (WebViewHost::HostMethodTask::runIfValid):
29885         (WebViewHost::taskList):
29886
29887 2011-02-07  Hayato Ito  <hayato@chromium.org>
29888
29889         Reviewed by Ojan Vafai.
29890
29891         Get rid of code which writes test results from test_type's
29892         compare_output() method.
29893         Writing test results now happens in single_test_runner, using
29894         newly introduced test_result_writer module.
29895
29896         Subsequent patches will eliminate test_type/* classes.
29897
29898         https://bugs.webkit.org/show_bug.cgi?id=53908
29899
29900         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
29901         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py: Added.
29902         * Scripts/webkitpy/layout_tests/test_types/image_diff.py:
29903         * Scripts/webkitpy/layout_tests/test_types/test_type_base.py:
29904         * Scripts/webkitpy/layout_tests/test_types/text_diff.py:
29905
29906 2011-02-17  Dirk Pranke  <dpranke@chromium.org>
29907
29908         Reviewed by Adam Roben.
29909
29910         Add blacklist of remaining webkitpy/* modules that fail on
29911         Win32, so that test-webkitpy will at least run cleanly.
29912
29913         https://bugs.webkit.org/show_bug.cgi?id=54709
29914
29915         * Scripts/webkitpy/layout_tests/test/main.py:
29916
29917 2011-02-17  Simon Fraser  <simon.fraser@apple.com>
29918
29919         Reviewed by Sam Weinig.
29920
29921         Set a preference to ensure that scrollbars are in a predictable
29922         state for DRT.
29923
29924         * DumpRenderTree/mac/DumpRenderTree.mm:
29925         (resetDefaultsToConsistentValues):
29926
29927 2011-02-17  Dirk Pranke  <dpranke@chromium.org>
29928
29929         Reviewed by Adam Roben.
29930
29931         test-webkitpy doesn't work with the multiprocessing module on win32
29932         or cygwin, so we skip it for now.
29933
29934         https://bugs.webkit.org/show_bug.cgi?id=54520
29935
29936         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
29937         * Scripts/webkitpy/layout_tests/layout_package/run_webkit_tests_unittest.py:
29938
29939 2011-02-17  Dirk Pranke  <dpranke@chromium.org>
29940
29941         Reviewed by Adam Roben.
29942
29943         Fix version, baseline_search_path for chromium-win-win7. Also
29944         fix chromium_win_unittests, and fix port_testcase.make_port()
29945         to correctly use the options passed into it and only set
29946         results_directory() if necessary.
29947
29948         https://bugs.webkit.org/show_bug.cgi?id=54652
29949
29950         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
29951         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
29952         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
29953
29954 2011-02-17  Dirk Pranke  <dpranke@chromium.org>
29955
29956         Reviewed by Ojan Vafai.
29957
29958         Fix a hang in new-run-webkit-tests when run with
29959         --dry-run --worker-model=processes
29960
29961         https://bugs.webkit.org/show_bug.cgi?id=54595
29962
29963         * Scripts/webkitpy/layout_tests/layout_package/test_runner2.py:
29964         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
29965
29966 2011-02-17  Kevin Ollivier  <kevino@theolliviers.com>
29967
29968         [wx] Build fixes after recent changes.
29969
29970         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
29971         (LayoutTestController::setMinimumTimerInterval):
29972         * wx/build/settings.py:
29973
29974 2011-02-17  Csaba Osztrogonác  <ossy@webkit.org>
29975
29976         Reviewed by Andreas Kling.
29977
29978         [Qt] ImageDiff sometimes hangs
29979         https://bugs.webkit.org/show_bug.cgi?id=54641
29980
29981         * DumpRenderTree/qt/ImageDiff.cpp: Clear actualImage and baselineImage after all comparison.
29982         (main):
29983
29984 2011-02-17  Peter Varga  <pvarga@webkit.org>
29985
29986         Reviewed by Csaba Osztrogonác.
29987
29988         The run-sunspider --v8-suite doesn't work
29989         https://bugs.webkit.org/show_bug.cgi?id=54660
29990
29991         * Scripts/run-sunspider:
29992
29993 2011-02-17  Gabor Rapcsanyi  <rgabor@webkit.org>
29994
29995         Reviewed by Csaba Osztrogonác.
29996
29997         [NRWT] Fix environment setup on Qt port.
29998         https://bugs.webkit.org/show_bug.cgi?id=54549
29999
30000         * Scripts/webkitpy/layout_tests/port/qt.py:
30001
30002 2011-02-16  Yael Aharon  <yael.aharon@nokia.com>
30003
30004         Reviewed by Andreas Kling.
30005
30006         [Qt] Enable tiled backing store by default in MiniBrowser.
30007         https://bugs.webkit.org/show_bug.cgi?id=54581
30008
30009         Instead of a flag to turn it on, a new flag is defined to turn
30010         tiled backing store off.
30011
30012         * MiniBrowser/qt/MiniBrowserApplication.cpp:
30013         (MiniBrowserApplication::handleUserOptions):
30014         * MiniBrowser/qt/MiniBrowserApplication.h:
30015         (WindowOptions::WindowOptions):
30016
30017 2011-02-16  Eric Seidel  <eric@webkit.org>
30018
30019         Reviewed by Mihai Parparita.
30020
30021         flaky test archive uploads are always 22 bytes long
30022         https://bugs.webkit.org/show_bug.cgi?id=54593
30023
30024         Speculative fix for the 22-bytes bug.
30025
30026         * Scripts/webkitpy/tool/bot/flakytestreporter.py:
30027
30028 2011-02-16  Anna Cavender  <annacc@chromium.org>
30029
30030         Reviewed by Mihai Parparita.
30031
30032         http/tests/media should be served over HTTP (not from local file)
30033         https://bugs.webkit.org/show_bug.cgi?id=54028
30034
30035         Removed special file:// treatment of http/tests/media
30036
30037         * Scripts/old-run-webkit-tests:
30038         * Scripts/webkitpy/layout_tests/port/base.py:
30039
30040 2011-02-16  Mihai Parparita  <mihaip@chromium.org>
30041
30042         Reviewed by Ojan Vafai.
30043
30044         ChangeLog.parse_latest_entry_from_file does not handle rolled over ChangeLogs
30045         https://bugs.webkit.org/show_bug.cgi?id=54609
30046
30047         r78737 rolled over ChangeLogs, and webkit-patch complains when landing
30048         patches such as this one that are the first to add an entry to the
30049         ChangeLog (since they don't see a second date line to know where the
30050         entry ends).
30051
30052         * Scripts/webkitpy/common/checkout/changelog.py:
30053         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
30054
30055 == Rolled over to ChangeLog-2011-02-16 ==