2 * Copyright (C) 2005-2013 Team XBMC
5 * This Program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2, or (at your option)
10 * This Program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
15 * You should have received a copy of the GNU General Public License
16 * along with XBMC; see the file COPYING. If not, see
17 * <http://www.gnu.org/licenses/>.
20 //-----------------------------------------------------------------------
22 // File: StringUtils.cpp
24 // Purpose: ATL split string utility
25 // Author: Paul J. Weiss
27 // Modified to use J O'Leary's CStdString class by kraqh3d
29 //------------------------------------------------------------------------
32 #include "StringUtils.h"
33 #include "utils/RegExp.h"
34 #include "utils/fstrcmp.h"
41 #define FORMAT_BLOCK_SIZE 2048 // # of bytes to increment per try
45 const char* ADDON_GUID_RE = "^(\\{){0,1}[0-9a-fA-F]{8}\\-[0-9a-fA-F]{4}\\-[0-9a-fA-F]{4}\\-[0-9a-fA-F]{4}\\-[0-9a-fA-F]{12}(\\}){0,1}$";
47 /* empty string for use in returns by ref */
48 const CStdString StringUtils::EmptyString = "";
49 const std::string StringUtils::Empty = "";
50 CStdString StringUtils::m_lastUUID = "";
52 string StringUtils::Format(const char *fmt, ...)
56 string str = FormatV(fmt, args);
62 string StringUtils::FormatV(const char *fmt, va_list args)
67 int size = FORMAT_BLOCK_SIZE;
70 char *cstr = reinterpret_cast<char*>(malloc(sizeof(char) * size));
76 va_copy(argCopy, args);
78 int nActual = vsnprintf(cstr, size, fmt, argCopy);
81 if (nActual > -1 && nActual < size) // We got a valid result
83 string str(cstr, nActual);
87 if (nActual > -1) // Exactly what we will need (glibc 2.1)
89 else // Let's try to double the size (glibc 2.0)
92 char *new_cstr = reinterpret_cast<char*>(realloc(cstr, sizeof(char) * size));
106 wstring StringUtils::Format(const wchar_t *fmt, ...)
110 wstring str = FormatV(fmt, args);
116 wstring StringUtils::FormatV(const wchar_t *fmt, va_list args)
121 int size = FORMAT_BLOCK_SIZE;
124 wchar_t *cstr = reinterpret_cast<wchar_t*>(malloc(sizeof(wchar_t) * size));
130 va_copy(argCopy, args);
132 int nActual = vswprintf(cstr, size, fmt, argCopy);
135 if (nActual > -1 && nActual < size) // We got a valid result
137 wstring str(cstr, nActual);
141 if (nActual > -1) // Exactly what we will need (glibc 2.1)
143 else // Let's try to double the size (glibc 2.0)
146 wchar_t *new_cstr = reinterpret_cast<wchar_t*>(realloc(cstr, sizeof(wchar_t) * size));
147 if (new_cstr == NULL)
159 void StringUtils::ToUpper(string &str)
161 transform(str.begin(), str.end(), str.begin(), ::toupper);
164 void StringUtils::ToLower(string &str)
166 transform(str.begin(), str.end(), str.begin(), ::tolower);
169 bool StringUtils::EqualsNoCase(const std::string &str1, const std::string &str2)
171 return EqualsNoCase(str1.c_str(), str2.c_str());
174 bool StringUtils::EqualsNoCase(const std::string &str1, const char *s2)
176 return EqualsNoCase(str1.c_str(), s2);
179 bool StringUtils::EqualsNoCase(const char *s1, const char *s2)
181 char c2; // we need only one char outside the loop
184 const char c1 = *s1++; // const local variable should help compiler to optimize
186 if (c1 != c2 && ::tolower(c1) != ::tolower(c2)) // This includes the possibility that one of the characters is the null-terminator, which implies a string mismatch.
188 } while (c2 != '\0'); // At this point, we know c1 == c2, so there's no need to test them both.
192 string StringUtils::Left(const string &str, size_t count)
194 count = max((size_t)0, min(count, str.size()));
195 return str.substr(0, count);
198 string StringUtils::Mid(const string &str, size_t first, size_t count /* = string::npos */)
200 if (first + count > str.size())
201 count = str.size() - first;
203 if (first > str.size())
206 ASSERT(first + count <= str.size());
208 return str.substr(first, count);
211 string StringUtils::Right(const string &str, size_t count)
213 count = max((size_t)0, min(count, str.size()));
214 return str.substr(str.size() - count);
217 std::string& StringUtils::Trim(std::string &str)
220 return TrimRight(str);
223 // hack to ensure that std::string::iterator will be dereferenced as _unsigned_ char
224 // without this hack "TrimX" functions failed on Win32 with UTF-8 strings
225 static int isspace_c(char c)
227 return ::isspace((unsigned char)c);
230 std::string& StringUtils::TrimLeft(std::string &str)
232 str.erase(str.begin(), ::find_if(str.begin(), str.end(), ::not1(::ptr_fun(isspace_c))));
236 std::string& StringUtils::TrimRight(std::string &str)
238 str.erase(::find_if(str.rbegin(), str.rend(), ::not1(::ptr_fun(isspace_c))).base(), str.end());
242 std::string& StringUtils::RemoveDuplicatedSpacesAndTabs(std::string& str)
244 std::string::iterator it = str.begin();
245 bool onSpace = false;
246 while(it != str.end())
269 int StringUtils::Replace(string &str, char oldChar, char newChar)
271 int replacedChars = 0;
272 for (string::iterator it = str.begin(); it != str.end(); it++)
281 return replacedChars;
284 int StringUtils::Replace(std::string &str, const std::string &oldStr, const std::string &newStr)
289 int replacedChars = 0;
292 while (index < str.size() && (index = str.find(oldStr, index)) != string::npos)
294 str.replace(index, oldStr.size(), newStr);
295 index += newStr.size();
299 return replacedChars;
302 bool StringUtils::StartsWith(const std::string &str1, const std::string &str2)
304 return str1.compare(0, str2.size(), str2) == 0;
307 bool StringUtils::StartsWith(const std::string &str1, const char *s2)
309 return StartsWith(str1.c_str(), s2);
312 bool StringUtils::StartsWith(const char *s1, const char *s2)
324 bool StringUtils::StartsWithNoCase(const std::string &str1, const std::string &str2)
326 return StartsWithNoCase(str1.c_str(), str2.c_str());
329 bool StringUtils::StartsWithNoCase(const std::string &str1, const char *s2)
331 return StartsWithNoCase(str1.c_str(), s2);
334 bool StringUtils::StartsWithNoCase(const char *s1, const char *s2)
338 if (::tolower(*s1) != ::tolower(*s2))
346 bool StringUtils::EndsWith(const std::string &str1, const std::string &str2)
348 if (str1.size() < str2.size())
350 return str1.compare(str1.size() - str2.size(), str2.size(), str2) == 0;
353 bool StringUtils::EndsWith(const std::string &str1, const char *s2)
355 size_t len2 = strlen(s2);
356 if (str1.size() < len2)
358 return str1.compare(str1.size() - len2, len2, s2) == 0;
361 bool StringUtils::EndsWithNoCase(const std::string &str1, const std::string &str2)
363 if (str1.size() < str2.size())
365 const char *s1 = str1.c_str() + str1.size() - str2.size();
366 const char *s2 = str2.c_str();
369 if (::tolower(*s1) != ::tolower(*s2))
377 bool StringUtils::EndsWithNoCase(const std::string &str1, const char *s2)
379 size_t len2 = strlen(s2);
380 if (str1.size() < len2)
382 const char *s1 = str1.c_str() + str1.size() - len2;
385 if (::tolower(*s1) != ::tolower(*s2))
393 void StringUtils::JoinString(const CStdStringArray &strings, const CStdString& delimiter, CStdString& result)
396 for(CStdStringArray::const_iterator it = strings.begin(); it != strings.end(); it++ )
397 result += (*it) + delimiter;
400 result.Delete(result.size()-delimiter.size(), delimiter.size());
403 CStdString StringUtils::JoinString(const CStdStringArray &strings, const CStdString& delimiter)
406 JoinString(strings, delimiter, result);
410 CStdString StringUtils::Join(const vector<string> &strings, const CStdString& delimiter)
412 CStdStringArray strArray;
413 for (unsigned int index = 0; index < strings.size(); index++)
414 strArray.push_back(strings.at(index));
416 return JoinString(strArray, delimiter);
419 // Splits the string input into pieces delimited by delimiter.
420 // if 2 delimiters are in a row, it will include the empty string between them.
421 // added MaxStrings parameter to restrict the number of returned substrings (like perl and python)
422 int StringUtils::SplitString(const CStdString& input, const CStdString& delimiter, CStdStringArray &results, unsigned int iMaxStrings /* = 0 */)
426 int sizeS2 = delimiter.GetLength();
427 int isize = input.GetLength();
431 vector<unsigned int> positions;
433 newPos = input.Find (delimiter, 0);
437 results.push_back(input);
441 while ( newPos > iPos )
443 positions.push_back(newPos);
445 newPos = input.Find (delimiter, iPos + sizeS2);
448 // numFound is the number of delimiters which is one less
449 // than the number of substrings
450 unsigned int numFound = positions.size();
451 if (iMaxStrings > 0 && numFound >= iMaxStrings)
452 numFound = iMaxStrings - 1;
454 for ( unsigned int i = 0; i <= numFound; i++ )
462 s = input.Mid( i, positions[i] );
466 int offset = positions[i - 1] + sizeS2;
467 if ( offset < isize )
470 s = input.Mid(offset);
472 s = input.Mid( positions[i - 1] + sizeS2,
473 positions[i] - positions[i - 1] - sizeS2 );
476 results.push_back(s);
478 // return the number of substrings
479 return results.size();
482 CStdStringArray StringUtils::SplitString(const CStdString& input, const CStdString& delimiter, unsigned int iMaxStrings /* = 0 */)
484 CStdStringArray result;
485 SplitString(input, delimiter, result, iMaxStrings);
489 vector<string> StringUtils::Split(const std::string& input, const std::string& delimiter, unsigned int iMaxStrings /* = 0 */)
491 CStdStringArray result;
492 SplitString(input, delimiter, result, iMaxStrings);
494 vector<string> strArray;
495 for (unsigned int index = 0; index < result.size(); index++)
496 strArray.push_back(result.at(index));
501 // returns the number of occurrences of strFind in strInput.
502 int StringUtils::FindNumber(const CStdString& strInput, const CStdString &strFind)
504 int pos = strInput.Find(strFind, 0);
509 pos = strInput.Find(strFind, pos + 1);
514 // Compares separately the numeric and alphabetic parts of a string.
515 // returns negative if left < right, positive if left > right
516 // and 0 if they are identical (essentially calculates left - right)
517 int64_t StringUtils::AlphaNumericCompare(const wchar_t *left, const wchar_t *right)
519 wchar_t *l = (wchar_t *)left;
520 wchar_t *r = (wchar_t *)right;
524 const collate<wchar_t>& coll = use_facet< collate<wchar_t> >( locale() );
526 while (*l != 0 && *r != 0)
528 // check if we have a numerical value
529 if (*l >= L'0' && *l <= L'9' && *r >= L'0' && *r <= L'9')
533 while (*ld >= L'0' && *ld <= L'9' && ld < l + 15)
534 { // compare only up to 15 digits
540 while (*rd >= L'0' && *rd <= L'9' && rd < r + 15)
541 { // compare only up to 15 digits
543 rnum += *rd++ - L'0';
545 // do we have numbers?
547 { // yes - and they're different!
554 // do case less comparison
556 if (lc >= L'A' && lc <= L'Z')
559 if (rc >= L'A' && rc <= L'Z')
562 // ok, do a normal comparison, taking current locale into account. Add special case stuff (eg '(' characters)) in here later
563 if ((cmp_res = coll.compare(&lc, &lc + 1, &rc, &rc + 1)) != 0)
577 return 0; // files are the same
580 int StringUtils::DateStringToYYYYMMDD(const CStdString &dateString)
582 CStdStringArray days;
583 int splitCount = StringUtils::SplitString(dateString, "-", days);
585 return atoi(days[0].c_str());
586 else if (splitCount == 2)
587 return atoi(days[0].c_str())*100+atoi(days[1].c_str());
588 else if (splitCount == 3)
589 return atoi(days[0].c_str())*10000+atoi(days[1].c_str())*100+atoi(days[2].c_str());
594 long StringUtils::TimeStringToSeconds(const CStdString &timeString)
596 CStdString strCopy(timeString);
597 strCopy.TrimLeft(" \n\r\t");
598 strCopy.TrimRight(" \n\r\t");
599 if(StringUtils::EndsWithNoCase(strCopy, " min"))
601 // this is imdb format of "XXX min"
602 return 60 * atoi(strCopy.c_str());
606 CStdStringArray secs;
607 StringUtils::SplitString(strCopy, ":", secs);
609 for (unsigned int i = 0; i < 3 && i < secs.size(); i++)
612 timeInSecs += atoi(secs[i]);
618 CStdString StringUtils::SecondsToTimeString(long lSeconds, TIME_FORMAT format)
620 int hh = lSeconds / 3600;
621 lSeconds = lSeconds % 3600;
622 int mm = lSeconds / 60;
623 int ss = lSeconds % 60;
625 if (format == TIME_FORMAT_GUESS)
626 format = (hh >= 1) ? TIME_FORMAT_HH_MM_SS : TIME_FORMAT_MM_SS;
628 if (format & TIME_FORMAT_HH)
629 strHMS.AppendFormat("%02.2i", hh);
630 else if (format & TIME_FORMAT_H)
631 strHMS.AppendFormat("%i", hh);
632 if (format & TIME_FORMAT_MM)
633 strHMS.AppendFormat(strHMS.IsEmpty() ? "%02.2i" : ":%02.2i", mm);
634 if (format & TIME_FORMAT_SS)
635 strHMS.AppendFormat(strHMS.IsEmpty() ? "%02.2i" : ":%02.2i", ss);
639 bool StringUtils::IsNaturalNumber(const CStdString& str)
642 // allow whitespace,digits,whitespace
643 while (i < str.size() && isspace((unsigned char) str[i]))
645 while (i < str.size() && isdigit((unsigned char) str[i]))
649 while (i < str.size() && isspace((unsigned char) str[i]))
651 return i == str.size() && n > 0;
654 bool StringUtils::IsInteger(const CStdString& str)
657 // allow whitespace,-,digits,whitespace
658 while (i < str.size() && isspace((unsigned char) str[i]))
660 if (i < str.size() && str[i] == '-')
662 while (i < str.size() && isdigit((unsigned char) str[i]))
666 while (i < str.size() && isspace((unsigned char) str[i]))
668 return i == str.size() && n > 0;
671 void StringUtils::RemoveCRLF(CStdString& strLine)
673 while ( strLine.size() && (strLine.Right(1) == "\n" || strLine.Right(1) == "\r") )
675 strLine = strLine.Left(std::max(0, (int)strLine.size() - 1));
679 CStdString StringUtils::SizeToString(int64_t size)
682 const char prefixes[] = {' ','k', 'M', 'G', 'T', 'P', 'E', 'Z', 'Y'};
684 double s = (double)size;
685 while (i < sizeof(prefixes)/sizeof(prefixes[0]) && s >= 1000.0)
692 strLabel.Format("%.0lf %cB ", s, prefixes[i]);
694 strLabel.Format("%.1lf %cB", s, prefixes[i]);
696 strLabel.Format("%.2lf %cB", s, prefixes[i]);
701 // return -1 if not, else return the utf8 char length.
702 int IsUTF8Letter(const unsigned char *str)
705 // unicode -> utf8 table: http://www.utf8-chartable.de/
706 // latin characters in unicode: http://en.wikipedia.org/wiki/Latin_characters_in_Unicode
707 unsigned char ch = str[0];
710 if ((ch >= 'a' && ch <= 'z') || (ch >= 'A' && ch <= 'Z'))
714 unsigned char ch2 = str[1];
717 // check latin 1 letter table: http://en.wikipedia.org/wiki/C1_Controls_and_Latin-1_Supplement
718 if (ch == 0xC3 && ch2 >= 0x80 && ch2 <= 0xBF && ch2 != 0x97 && ch2 != 0xB7)
720 // check latin extended A table: http://en.wikipedia.org/wiki/Latin_Extended-A
721 if (ch >= 0xC4 && ch <= 0xC7 && ch2 >= 0x80 && ch2 <= 0xBF)
723 // check latin extended B table: http://en.wikipedia.org/wiki/Latin_Extended-B
724 // and International Phonetic Alphabet: http://en.wikipedia.org/wiki/IPA_Extensions_(Unicode_block)
725 if (((ch == 0xC8 || ch == 0xC9) && ch2 >= 0x80 && ch2 <= 0xBF)
726 || (ch == 0xCA && ch2 >= 0x80 && ch2 <= 0xAF))
731 size_t StringUtils::FindWords(const char *str, const char *wordLowerCase)
733 // NOTE: This assumes word is lowercase!
734 unsigned char *s = (unsigned char *)str;
737 // start with a compare
738 unsigned char *c = s;
739 unsigned char *w = (unsigned char *)wordLowerCase;
741 while (same && *c && *w)
743 unsigned char lc = *c++;
744 if (lc >= 'A' && lc <= 'Z')
747 if (lc != *w++) // different
750 if (same && *w == 0) // only the same if word has been exhausted
751 return (const char *)s - str;
753 // otherwise, skip current word (composed by latin letters) or number
755 if (*s >= '0' && *s <= '9')
758 while (*s >= '0' && *s <= '9') ++s;
760 else if ((l = IsUTF8Letter(s)) > 0)
763 while ((l = IsUTF8Letter(s)) > 0) s += l;
767 while (*s && *s == ' ') s++;
769 // and repeat until we're done
772 return CStdString::npos;
775 // assumes it is called from after the first open bracket is found
776 int StringUtils::FindEndBracket(const CStdString &str, char opener, char closer, int startPos)
779 for (unsigned int i = startPos; i < str.size(); i++)
781 if (str[i] == opener)
783 else if (str[i] == closer)
791 return (int)CStdString::npos;
794 void StringUtils::WordToDigits(CStdString &word)
796 static const char word_to_letter[] = "22233344455566677778889999";
798 for (unsigned int i = 0; i < word.size(); ++i)
799 { // NB: This assumes ascii, which probably needs extending at some point.
800 char letter = word[i];
801 if ((letter >= 'a' && letter <= 'z')) // assume contiguous letter range
803 word[i] = word_to_letter[letter-'a'];
805 else if (letter < '0' || letter > '9') // We want to keep 0-9!
807 word[i] = ' '; // replace everything else with a space
812 CStdString StringUtils::CreateUUID()
814 /* This function generate a DCE 1.1, ISO/IEC 11578:1996 and IETF RFC-4122
815 * Version 4 conform local unique UUID based upon random number generation.
818 char *pUuidStr = UuidStrTmp;
821 static bool m_uuidInitialized = false;
822 if (!m_uuidInitialized)
824 /* use current time as the seed for rand()*/
826 m_uuidInitialized = true;
829 /*Data1 - 8 characters.*/
830 for(i = 0; i < 8; i++, pUuidStr++)
831 ((*pUuidStr = (rand() % 16)) < 10) ? *pUuidStr += 48 : *pUuidStr += 55;
833 /*Data2 - 4 characters.*/
835 for(i = 0; i < 4; i++, pUuidStr++)
836 ((*pUuidStr = (rand() % 16)) < 10) ? *pUuidStr += 48 : *pUuidStr += 55;
838 /*Data3 - 4 characters.*/
840 for(i = 0; i < 4; i++, pUuidStr++)
841 ((*pUuidStr = (rand() % 16)) < 10) ? *pUuidStr += 48 : *pUuidStr += 55;
843 /*Data4 - 4 characters.*/
845 for(i = 0; i < 4; i++, pUuidStr++)
846 ((*pUuidStr = (rand() % 16)) < 10) ? *pUuidStr += 48 : *pUuidStr += 55;
848 /*Data5 - 12 characters.*/
850 for(i = 0; i < 12; i++, pUuidStr++)
851 ((*pUuidStr = (rand() % 16)) < 10) ? *pUuidStr += 48 : *pUuidStr += 55;
855 m_lastUUID = UuidStrTmp;
859 bool StringUtils::ValidateUUID(const CStdString &uuid)
862 guidRE.RegComp(ADDON_GUID_RE);
863 return (guidRE.RegFind(uuid.c_str()) == 0);
866 double StringUtils::CompareFuzzy(const CStdString &left, const CStdString &right)
868 return (0.5 + fstrcmp(left.c_str(), right.c_str(), 0.0) * (left.length() + right.length())) / 2.0;
871 int StringUtils::FindBestMatch(const CStdString &str, const CStdStringArray &strings, double &matchscore)
877 for (CStdStringArray::const_iterator it = strings.begin(); it != strings.end(); it++, i++)
879 int maxlength = max(str.length(), it->length());
880 double score = StringUtils::CompareFuzzy(str, *it) / maxlength;
881 if (score > matchscore)
890 bool StringUtils::ContainsKeyword(const CStdString &str, const CStdStringArray &keywords)
892 for (CStdStringArray::const_iterator it = keywords.begin(); it != keywords.end(); it++)
894 if (str.find(*it) != str.npos)
900 size_t StringUtils::utf8_strlen(const char *s)
905 if ((*s++ & 0xC0) != 0x80)
911 std::string StringUtils::Paramify(const std::string ¶m)
913 std::string result = param;
915 StringUtils::Replace(result, "\\", "\\\\");
916 // escape double quotes
917 StringUtils::Replace(result, "\"", "\\\"");
919 // add double quotes around the whole string
920 return "\"" + result + "\"";
923 void StringUtils::Tokenize(const std::string& input, std::vector<std::string>& tokens, const std::string& delimiters)
925 // Tokenize ripped from http://www.linuxselfhelp.com/HOWTO/C++Programming-HOWTO-7.html
926 // Skip delimiters at beginning.
927 string::size_type lastPos = input.find_first_not_of(delimiters, 0);
928 // Find first "non-delimiter".
929 string::size_type pos = input.find_first_of(delimiters, lastPos);
931 while (string::npos != pos || string::npos != lastPos)
933 // Found a token, add it to the vector.
934 tokens.push_back(input.substr(lastPos, pos - lastPos));
935 // Skip delimiters. Note the "not_of"
936 lastPos = input.find_first_not_of(delimiters, pos);
937 // Find next "non-delimiter"
938 pos = input.find_first_of(delimiters, lastPos);