runtime/RegExp.cpp

   1 /*
   2  *  Copyright (C) 1999-2001, 2004 Harri Porten (porten@kde.org)
   3  *  Copyright (c) 2007, 2008 Apple Inc. All rights reserved.
   4  *  Copyright (C) 2009 Torch Mobile, Inc.
   5  *
   6  *  This library is free software; you can redistribute it and/or
   7  *  modify it under the terms of the GNU Lesser General Public
   8  *  License as published by the Free Software Foundation; either
   9  *  version 2 of the License, or (at your option) any later version.
  10  *
  11  *  This library is distributed in the hope that it will be useful,
  12  *  but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14  *  Lesser General Public License for more details.
  15  *
  16  *  You should have received a copy of the GNU Lesser General Public
  17  *  License along with this library; if not, write to the Free Software
  18  *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
  19  *
  20  */
  21
  22 #include "config.h"
  23 #include "RegExp.h"
  24 #include "Lexer.h"
  25 #include <stdio.h>
  26 #include <stdlib.h>
  27 #include <string.h>
  28 #include <wtf/Assertions.h>
  29 #include <wtf/OwnArrayPtr.h>
  30
  31
  32 #if ENABLE(YARR)
  33
  34 #include "yarr/RegexCompiler.h"
  35 #if ENABLE(YARR_JIT)
  36 #include "yarr/RegexJIT.h"
  37 #else
  38 #include "yarr/RegexInterpreter.h"
  39 #endif
  40
  41 #else
  42
  43 #if ENABLE(WREC)
  44 #include "JIT.h"
  45 #include "WRECGenerator.h"
  46 #endif
  47 #include <pcre/pcre.h>
  48
  49 #endif
  50
  51 namespace JSC {
  52
  53 #if ENABLE(WREC)
  54 using namespace WREC;
  55 #endif
  56
  57 inline RegExp::RegExp(JSGlobalData* globalData, const UString& pattern)
  58     : m_pattern(pattern)
  59     , m_flagBits(0)
  60     , m_constructionError(0)
  61     , m_numSubpatterns(0)
  62 {
  63     compile(globalData);
  64 }
  65
  66 inline RegExp::RegExp(JSGlobalData* globalData, const UString& pattern, const UString& flags)
  67     : m_pattern(pattern)
  68     , m_flagBits(0)
  69     , m_constructionError(0)
  70     , m_numSubpatterns(0)
  71 {
  72     // NOTE: The global flag is handled on a case-by-case basis by functions like
  73     // String::match and RegExpObject::match.
  74     if (flags.find('g') != -1)
  75         m_flagBits |= Global;
  76     if (flags.find('i') != -1)
  77         m_flagBits |= IgnoreCase;
  78     if (flags.find('m') != -1)
  79         m_flagBits |= Multiline;
  80
  81     compile(globalData);
  82 }
  83
  84 #if !ENABLE(YARR)
  85 RegExp::~RegExp()
  86 {
  87     jsRegExpFree(m_regExp);
  88 }
  89 #endif
  90
  91 PassRefPtr<RegExp> RegExp::create(JSGlobalData* globalData, const UString& pattern)
  92 {
  93     return adoptRef(new RegExp(globalData, pattern));
  94 }
  95
  96 PassRefPtr<RegExp> RegExp::create(JSGlobalData* globalData, const UString& pattern, const UString& flags)
  97 {
  98     return adoptRef(new RegExp(globalData, pattern, flags));
  99 }
 100
 101 #if ENABLE(YARR)
 102
 103 void RegExp::compile(JSGlobalData* globalData)
 104 {
 105 #if ENABLE(YARR_JIT)
 106     Yarr::jitCompileRegex(globalData, m_regExpJITCode, m_pattern, m_numSubpatterns, m_constructionError, ignoreCase(), multiline());
 107 #else
 108     UNUSED_PARAM(globalData);
 109     m_regExpBytecode.set(Yarr::byteCompileRegex(m_pattern, m_numSubpatterns, m_constructionError, ignoreCase(), multiline()));
 110 #endif
 111 }
 112
 113 int RegExp::match(const UString& s, int startOffset, Vector<int, 32>* ovector)
 114 {
 115     if (startOffset < 0)
 116         startOffset = 0;
 117     if (ovector)
 118         ovector->clear();
 119
 120     if (startOffset > s.size() || s.isNull())
 121         return -1;
 122
 123 #if ENABLE(YARR_JIT)
 124     if (!!m_regExpJITCode) {
 125 #else
 126     if (m_regExpBytecode) {
 127 #endif
 128         int offsetVectorSize = (m_numSubpatterns + 1) * 3; // FIXME: should be 2 - but adding temporary fallback to pcre.
 129         int* offsetVector;
 130         Vector<int, 32> nonReturnedOvector;
 131         if (ovector) {
 132             ovector->resize(offsetVectorSize);
 133             offsetVector = ovector->data();
 134         } else {
 135             nonReturnedOvector.resize(offsetVectorSize);
 136             offsetVector = nonReturnedOvector.data();
 137         }
 138
 139         ASSERT(offsetVector);
 140         for (int j = 0; j < offsetVectorSize; ++j)
 141             offsetVector[j] = -1;
 142
 143
 144 #if ENABLE(YARR_JIT)
 145         int result = Yarr::executeRegex(m_regExpJITCode, s.data(), startOffset, s.size(), offsetVector, offsetVectorSize);
 146 #else
 147         int result = Yarr::interpretRegex(m_regExpBytecode.get(), s.data(), startOffset, s.size(), offsetVector);
 148 #endif
 149
 150         if (result < 0) {
 151 #ifndef NDEBUG
 152             // TODO: define up a symbol, rather than magic -1
 153             if (result != -1)
 154                 fprintf(stderr, "jsRegExpExecute failed with result %d\n", result);
 155 #endif
 156             if (ovector)
 157                 ovector->clear();
 158         }
 159         return result;
 160     }
 161
 162     return -1;
 163 }
 164
 165 #else
 166
 167 void RegExp::compile(JSGlobalData* globalData)
 168 {
 169     m_regExp = 0;
 170 #if ENABLE(WREC)
 171     m_wrecFunction = Generator::compileRegExp(globalData, m_pattern, &m_numSubpatterns, &m_constructionError, m_executablePool, ignoreCase(), multiline());
 172     if (m_wrecFunction || m_constructionError)
 173         return;
 174     // Fall through to non-WREC case.
 175 #else
 176     UNUSED_PARAM(globalData);
 177 #endif
 178
 179     JSRegExpIgnoreCaseOption ignoreCaseOption = ignoreCase() ? JSRegExpIgnoreCase : JSRegExpDoNotIgnoreCase;
 180     JSRegExpMultilineOption multilineOption = multiline() ? JSRegExpMultiline : JSRegExpSingleLine;
 181     m_regExp = jsRegExpCompile(reinterpret_cast<const UChar*>(m_pattern.data()), m_pattern.size(), ignoreCaseOption, multilineOption, &m_numSubpatterns, &m_constructionError);
 182 }
 183
 184 int RegExp::match(const UString& s, int startOffset, Vector<int, 32>* ovector)
 185 {
 186     if (startOffset < 0)
 187         startOffset = 0;
 188     if (ovector)
 189         ovector->clear();
 190
 191     if (startOffset > s.size() || s.isNull())
 192         return -1;
 193
 194 #if ENABLE(WREC)
 195     if (m_wrecFunction) {
 196         int offsetVectorSize = (m_numSubpatterns + 1) * 2;
 197         int* offsetVector;
 198         Vector<int, 32> nonReturnedOvector;
 199         if (ovector) {
 200             ovector->resize(offsetVectorSize);
 201             offsetVector = ovector->data();
 202         } else {
 203             nonReturnedOvector.resize(offsetVectorSize);
 204             offsetVector = nonReturnedOvector.data();
 205         }
 206         ASSERT(offsetVector);
 207         for (int j = 0; j < offsetVectorSize; ++j)
 208             offsetVector[j] = -1;
 209
 210         int result = m_wrecFunction(s.data(), startOffset, s.size(), offsetVector);
 211
 212         if (result < 0) {
 213 #ifndef NDEBUG
 214             // TODO: define up a symbol, rather than magic -1
 215             if (result != -1)
 216                 fprintf(stderr, "jsRegExpExecute failed with result %d\n", result);
 217 #endif
 218             if (ovector)
 219                 ovector->clear();
 220         }
 221         return result;
 222     } else
 223 #endif
 224     if (m_regExp) {
 225         // Set up the offset vector for the result.
 226         // First 2/3 used for result, the last third used by PCRE.
 227         int* offsetVector;
 228         int offsetVectorSize;
 229         int fixedSizeOffsetVector[3];
 230         if (!ovector) {
 231             offsetVectorSize = 3;
 232             offsetVector = fixedSizeOffsetVector;
 233         } else {
 234             offsetVectorSize = (m_numSubpatterns + 1) * 3;
 235             ovector->resize(offsetVectorSize);
 236             offsetVector = ovector->data();
 237         }
 238
 239         int numMatches = jsRegExpExecute(m_regExp, reinterpret_cast<const UChar*>(s.data()), s.size(), startOffset, offsetVector, offsetVectorSize);
 240
 241         if (numMatches < 0) {
 242 #ifndef NDEBUG
 243             if (numMatches != JSRegExpErrorNoMatch)
 244                 fprintf(stderr, "jsRegExpExecute failed with result %d\n", numMatches);
 245 #endif
 246             if (ovector)
 247                 ovector->clear();
 248             return -1;
 249         }
 250
 251         return offsetVector[0];
 252     }
 253
 254     return -1;
 255 }
 256
 257 #endif
 258
 259 } // namespace JSC