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 #include <pcre/pcre.h>
  44
  45 #endif
  46
  47 namespace JSC {
  48
  49 inline RegExp::RegExp(JSGlobalData* globalData, const UString& pattern)
  50     : m_pattern(pattern)
  51     , m_flagBits(0)
  52     , m_constructionError(0)
  53     , m_numSubpatterns(0)
  54     , m_lastMatchStart(-1)
  55 {
  56     compile(globalData);
  57 }
  58
  59 inline RegExp::RegExp(JSGlobalData* globalData, const UString& pattern, const UString& flags)
  60     : m_pattern(pattern)
  61     , m_flagBits(0)
  62     , m_constructionError(0)
  63     , m_numSubpatterns(0)
  64     , m_lastMatchStart(-1)
  65 {
  66     // NOTE: The global flag is handled on a case-by-case basis by functions like
  67     // String::match and RegExpObject::match.
  68     if (flags.find('g') != UString::NotFound)
  69         m_flagBits |= Global;
  70     if (flags.find('i') != UString::NotFound)
  71         m_flagBits |= IgnoreCase;
  72     if (flags.find('m') != UString::NotFound)
  73         m_flagBits |= Multiline;
  74
  75     compile(globalData);
  76 }
  77
  78 #if !ENABLE(YARR)
  79 RegExp::~RegExp()
  80 {
  81     jsRegExpFree(m_regExp);
  82 }
  83 #endif
  84
  85 PassRefPtr<RegExp> RegExp::create(JSGlobalData* globalData, const UString& pattern)
  86 {
  87     return adoptRef(new RegExp(globalData, pattern));
  88 }
  89
  90 PassRefPtr<RegExp> RegExp::create(JSGlobalData* globalData, const UString& pattern, const UString& flags)
  91 {
  92     return adoptRef(new RegExp(globalData, pattern, flags));
  93 }
  94
  95 #if ENABLE(YARR)
  96
  97 void RegExp::compile(JSGlobalData* globalData)
  98 {
  99 #if ENABLE(YARR_JIT)
 100     Yarr::jitCompileRegex(globalData, m_regExpJITCode, m_pattern, m_numSubpatterns, m_constructionError, ignoreCase(), multiline());
 101 #else
 102     UNUSED_PARAM(globalData);
 103     m_regExpBytecode.set(Yarr::byteCompileRegex(m_pattern, m_numSubpatterns, m_constructionError, ignoreCase(), multiline()));
 104 #endif
 105 }
 106
 107 int RegExp::match(const UString& s, int startOffset, Vector<int, 32>* ovector)
 108 {
 109     if (startOffset < 0)
 110         startOffset = 0;
 111     if (ovector)
 112         ovector->clear();
 113
 114     if (static_cast<unsigned>(startOffset) > s.size() || s.isNull()) {
 115         m_lastMatchString = UString();
 116         m_lastMatchStart = -1;
 117         m_lastOVector.shrink(0);
 118         return -1;
 119     }
 120
 121     // Perform check to see if this match call is the same as the last match invocation
 122     // and if it is return the prior result.
 123     if ((startOffset == m_lastMatchStart) && (s.rep() == m_lastMatchString.rep())) {
 124         if (ovector)
 125             *ovector = m_lastOVector;
 126
 127         if (m_lastOVector.isEmpty())
 128             return -1;
 129
 130         return m_lastOVector.at(0);
 131     }
 132
 133 #if ENABLE(YARR_JIT)
 134     if (!!m_regExpJITCode) {
 135 #else
 136     if (m_regExpBytecode) {
 137 #endif
 138         int offsetVectorSize = (m_numSubpatterns + 1) * 3; // FIXME: should be 2 - but adding temporary fallback to pcre.
 139         int* offsetVector;
 140         Vector<int, 32> nonReturnedOvector;
 141         if (ovector) {
 142             ovector->resize(offsetVectorSize);
 143             offsetVector = ovector->data();
 144         } else {
 145             nonReturnedOvector.resize(offsetVectorSize);
 146             offsetVector = nonReturnedOvector.data();
 147         }
 148
 149         ASSERT(offsetVector);
 150         for (int j = 0; j < offsetVectorSize; ++j)
 151             offsetVector[j] = -1;
 152
 153
 154 #if ENABLE(YARR_JIT)
 155         int result = Yarr::executeRegex(m_regExpJITCode, s.data(), startOffset, s.size(), offsetVector, offsetVectorSize);
 156 #else
 157         int result = Yarr::interpretRegex(m_regExpBytecode.get(), s.data(), startOffset, s.size(), offsetVector);
 158 #endif
 159
 160         if (result < 0) {
 161 #ifndef NDEBUG
 162             // TODO: define up a symbol, rather than magic -1
 163             if (result != -1)
 164                 fprintf(stderr, "jsRegExpExecute failed with result %d\n", result);
 165 #endif
 166             if (ovector)
 167                 ovector->clear();
 168         }
 169
 170         m_lastMatchString = s;
 171         m_lastMatchStart = startOffset;
 172
 173         if (ovector)
 174             m_lastOVector = *ovector;
 175         else
 176             m_lastOVector = nonReturnedOvector;
 177
 178         return result;
 179     }
 180
 181     m_lastMatchString = UString();
 182     m_lastMatchStart = -1;
 183     m_lastOVector.shrink(0);
 184
 185     return -1;
 186 }
 187
 188 #else
 189
 190 void RegExp::compile(JSGlobalData*)
 191 {
 192     m_regExp = 0;
 193     JSRegExpIgnoreCaseOption ignoreCaseOption = ignoreCase() ? JSRegExpIgnoreCase : JSRegExpDoNotIgnoreCase;
 194     JSRegExpMultilineOption multilineOption = multiline() ? JSRegExpMultiline : JSRegExpSingleLine;
 195     m_regExp = jsRegExpCompile(reinterpret_cast<const UChar*>(m_pattern.data()), m_pattern.size(), ignoreCaseOption, multilineOption, &m_numSubpatterns, &m_constructionError);
 196 }
 197
 198 int RegExp::match(const UString& s, int startOffset, Vector<int, 32>* ovector)
 199 {
 200     if (startOffset < 0)
 201         startOffset = 0;
 202     if (ovector)
 203         ovector->clear();
 204
 205     if (static_cast<unsigned>(startOffset) > s.size() || s.isNull())
 206         return -1;
 207
 208     if (m_regExp) {
 209         // Set up the offset vector for the result.
 210         // First 2/3 used for result, the last third used by PCRE.
 211         int* offsetVector;
 212         int offsetVectorSize;
 213         int fixedSizeOffsetVector[3];
 214         if (!ovector) {
 215             offsetVectorSize = 3;
 216             offsetVector = fixedSizeOffsetVector;
 217         } else {
 218             offsetVectorSize = (m_numSubpatterns + 1) * 3;
 219             ovector->resize(offsetVectorSize);
 220             offsetVector = ovector->data();
 221         }
 222
 223         int numMatches = jsRegExpExecute(m_regExp, reinterpret_cast<const UChar*>(s.data()), s.size(), startOffset, offsetVector, offsetVectorSize);
 224
 225         if (numMatches < 0) {
 226 #ifndef NDEBUG
 227             if (numMatches != JSRegExpErrorNoMatch)
 228                 fprintf(stderr, "jsRegExpExecute failed with result %d\n", numMatches);
 229 #endif
 230             if (ovector)
 231                 ovector->clear();
 232             return -1;
 233         }
 234
 235         return offsetVector[0];
 236     }
 237
 238     return -1;
 239 }
 240
 241 #endif
 242
 243 } // namespace JSC