]> git.saurik.com Git - apple/javascriptcore.git/blobdiff - yarr/RegexCompiler.cpp
JavaScriptCore-903.tar.gz
[apple/javascriptcore.git] / yarr / RegexCompiler.cpp
diff --git a/yarr/RegexCompiler.cpp b/yarr/RegexCompiler.cpp
deleted file mode 100644 (file)
index 9fbe213..0000000
+++ /dev/null
@@ -1,635 +0,0 @@
-/*
- * Copyright (C) 2009 Apple Inc. All rights reserved.
- *
- * Redistribution and use in source and binary forms, with or without
- * modification, are permitted provided that the following conditions
- * are met:
- * 1. Redistributions of source code must retain the above copyright
- *    notice, this list of conditions and the following disclaimer.
- * 2. Redistributions in binary form must reproduce the above copyright
- *    notice, this list of conditions and the following disclaimer in the
- *    documentation and/or other materials provided with the distribution.
- *
- * THIS SOFTWARE IS PROVIDED BY APPLE INC. ``AS IS'' AND ANY
- * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
- * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
- * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL APPLE INC. OR
- * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
- * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
- * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
- * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
- * OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 
- */
-
-#include "config.h"
-#include "RegexCompiler.h"
-
-#include "RegexInterpreter.h"
-#include "RegexPattern.h"
-#include <wtf/Vector.h>
-
-#if ENABLE(YARR)
-
-using namespace WTF;
-
-namespace JSC { namespace Yarr {
-
-#include "RegExpJitTables.h"
-
-class CharacterClassConstructor {
-public:
-    CharacterClassConstructor(bool isCaseInsensitive = false)
-        : m_isCaseInsensitive(isCaseInsensitive)
-    {
-    }
-    
-    void reset()
-    {
-        m_matches.clear();
-        m_ranges.clear();
-        m_matchesUnicode.clear();
-        m_rangesUnicode.clear();
-    }
-
-    void append(const CharacterClass* other)
-    {
-        for (size_t i = 0; i < other->m_matches.size(); ++i)
-            addSorted(m_matches, other->m_matches[i]);
-        for (size_t i = 0; i < other->m_ranges.size(); ++i)
-            addSortedRange(m_ranges, other->m_ranges[i].begin, other->m_ranges[i].end);
-        for (size_t i = 0; i < other->m_matchesUnicode.size(); ++i)
-            addSorted(m_matchesUnicode, other->m_matchesUnicode[i]);
-        for (size_t i = 0; i < other->m_rangesUnicode.size(); ++i)
-            addSortedRange(m_rangesUnicode, other->m_rangesUnicode[i].begin, other->m_rangesUnicode[i].end);
-    }
-
-    void putChar(UChar ch)
-    {
-        if (ch <= 0x7f) {
-            if (m_isCaseInsensitive && isASCIIAlpha(ch)) {
-                addSorted(m_matches, toASCIIUpper(ch));
-                addSorted(m_matches, toASCIILower(ch));
-            } else
-                addSorted(m_matches, ch);
-        } else {
-            UChar upper, lower;
-            if (m_isCaseInsensitive && ((upper = Unicode::toUpper(ch)) != (lower = Unicode::toLower(ch)))) {
-                addSorted(m_matchesUnicode, upper);
-                addSorted(m_matchesUnicode, lower);
-            } else
-                addSorted(m_matchesUnicode, ch);
-        }
-    }
-
-    // returns true if this character has another case, and 'ch' is the upper case form.
-    static inline bool isUnicodeUpper(UChar ch)
-    {
-        return ch != Unicode::toLower(ch);
-    }
-
-    // returns true if this character has another case, and 'ch' is the lower case form.
-    static inline bool isUnicodeLower(UChar ch)
-    {
-        return ch != Unicode::toUpper(ch);
-    }
-
-    void putRange(UChar lo, UChar hi)
-    {
-        if (lo <= 0x7f) {
-            char asciiLo = lo;
-            char asciiHi = std::min(hi, (UChar)0x7f);
-            addSortedRange(m_ranges, lo, asciiHi);
-            
-            if (m_isCaseInsensitive) {
-                if ((asciiLo <= 'Z') && (asciiHi >= 'A'))
-                    addSortedRange(m_ranges, std::max(asciiLo, 'A')+('a'-'A'), std::min(asciiHi, 'Z')+('a'-'A'));
-                if ((asciiLo <= 'z') && (asciiHi >= 'a'))
-                    addSortedRange(m_ranges, std::max(asciiLo, 'a')+('A'-'a'), std::min(asciiHi, 'z')+('A'-'a'));
-            }
-        }
-        if (hi >= 0x80) {
-            uint32_t unicodeCurr = std::max(lo, (UChar)0x80);
-            addSortedRange(m_rangesUnicode, unicodeCurr, hi);
-            
-            if (m_isCaseInsensitive) {
-                while (unicodeCurr <= hi) {
-                    // If the upper bound of the range (hi) is 0xffff, the increments to
-                    // unicodeCurr in this loop may take it to 0x10000.  This is fine
-                    // (if so we won't re-enter the loop, since the loop condition above
-                    // will definitely fail) - but this does mean we cannot use a UChar
-                    // to represent unicodeCurr, we must use a 32-bit value instead.
-                    ASSERT(unicodeCurr <= 0xffff);
-
-                    if (isUnicodeUpper(unicodeCurr)) {
-                        UChar lowerCaseRangeBegin = Unicode::toLower(unicodeCurr);
-                        UChar lowerCaseRangeEnd = lowerCaseRangeBegin;
-                        while ((++unicodeCurr <= hi) && isUnicodeUpper(unicodeCurr) && (Unicode::toLower(unicodeCurr) == (lowerCaseRangeEnd + 1)))
-                            lowerCaseRangeEnd++;
-                        addSortedRange(m_rangesUnicode, lowerCaseRangeBegin, lowerCaseRangeEnd);
-                    } else if (isUnicodeLower(unicodeCurr)) {
-                        UChar upperCaseRangeBegin = Unicode::toUpper(unicodeCurr);
-                        UChar upperCaseRangeEnd = upperCaseRangeBegin;
-                        while ((++unicodeCurr <= hi) && isUnicodeLower(unicodeCurr) && (Unicode::toUpper(unicodeCurr) == (upperCaseRangeEnd + 1)))
-                            upperCaseRangeEnd++;
-                        addSortedRange(m_rangesUnicode, upperCaseRangeBegin, upperCaseRangeEnd);
-                    } else
-                        ++unicodeCurr;
-                }
-            }
-        }
-    }
-
-    CharacterClass* charClass()
-    {
-        CharacterClass* characterClass = new CharacterClass(0);
-
-        characterClass->m_matches.append(m_matches);
-        characterClass->m_ranges.append(m_ranges);
-        characterClass->m_matchesUnicode.append(m_matchesUnicode);
-        characterClass->m_rangesUnicode.append(m_rangesUnicode);
-
-        reset();
-
-        return characterClass;
-    }
-
-private:
-    void addSorted(Vector<UChar>& matches, UChar ch)
-    {
-        unsigned pos = 0;
-        unsigned range = matches.size();
-
-        // binary chop, find position to insert char.
-        while (range) {
-            unsigned index = range >> 1;
-
-            int val = matches[pos+index] - ch;
-            if (!val)
-                return;
-            else if (val > 0)
-                range = index;
-            else {
-                pos += (index+1);
-                range -= (index+1);
-            }
-        }
-        
-        if (pos == matches.size())
-            matches.append(ch);
-        else
-            matches.insert(pos, ch);
-    }
-
-    void addSortedRange(Vector<CharacterRange>& ranges, UChar lo, UChar hi)
-    {
-        unsigned end = ranges.size();
-        
-        // Simple linear scan - I doubt there are that many ranges anyway...
-        // feel free to fix this with something faster (eg binary chop).
-        for (unsigned i = 0; i < end; ++i) {
-            // does the new range fall before the current position in the array
-            if (hi < ranges[i].begin) {
-                // optional optimization: concatenate appending ranges? - may not be worthwhile.
-                if (hi == (ranges[i].begin - 1)) {
-                    ranges[i].begin = lo;
-                    return;
-                }
-                ranges.insert(i, CharacterRange(lo, hi));
-                return;
-            }
-            // Okay, since we didn't hit the last case, the end of the new range is definitely at or after the begining
-            // If the new range start at or before the end of the last range, then the overlap (if it starts one after the
-            // end of the last range they concatenate, which is just as good.
-            if (lo <= (ranges[i].end + 1)) {
-                // found an intersect! we'll replace this entry in the array.
-                ranges[i].begin = std::min(ranges[i].begin, lo);
-                ranges[i].end = std::max(ranges[i].end, hi);
-
-                // now check if the new range can subsume any subsequent ranges.
-                unsigned next = i+1;
-                // each iteration of the loop we will either remove something from the list, or break the loop.
-                while (next < ranges.size()) {
-                    if (ranges[next].begin <= (ranges[i].end + 1)) {
-                        // the next entry now overlaps / concatenates this one.
-                        ranges[i].end = std::max(ranges[i].end, ranges[next].end);
-                        ranges.remove(next);
-                    } else
-                        break;
-                }
-                
-                return;
-            }
-        }
-
-        // CharacterRange comes after all existing ranges.
-        ranges.append(CharacterRange(lo, hi));
-    }
-
-    bool m_isCaseInsensitive;
-
-    Vector<UChar> m_matches;
-    Vector<CharacterRange> m_ranges;
-    Vector<UChar> m_matchesUnicode;
-    Vector<CharacterRange> m_rangesUnicode;
-};
-
-class RegexPatternConstructor {
-public:
-    RegexPatternConstructor(RegexPattern& pattern)
-        : m_pattern(pattern)
-        , m_characterClassConstructor(pattern.m_ignoreCase)
-    {
-    }
-
-    ~RegexPatternConstructor()
-    {
-    }
-
-    void reset()
-    {
-        m_pattern.reset();
-        m_characterClassConstructor.reset();
-    }
-    
-    void assertionBOL()
-    {
-        m_alternative->m_terms.append(PatternTerm::BOL());
-    }
-    void assertionEOL()
-    {
-        m_alternative->m_terms.append(PatternTerm::EOL());
-    }
-    void assertionWordBoundary(bool invert)
-    {
-        m_alternative->m_terms.append(PatternTerm::WordBoundary(invert));
-    }
-
-    void atomPatternCharacter(UChar ch)
-    {
-        // We handle case-insensitive checking of unicode characters which do have both
-        // cases by handling them as if they were defined using a CharacterClass.
-        if (m_pattern.m_ignoreCase && !isASCII(ch) && (Unicode::toUpper(ch) != Unicode::toLower(ch))) {
-            atomCharacterClassBegin();
-            atomCharacterClassAtom(ch);
-            atomCharacterClassEnd();
-        } else
-            m_alternative->m_terms.append(PatternTerm(ch));
-    }
-
-    void atomBuiltInCharacterClass(BuiltInCharacterClassID classID, bool invert)
-    {
-        switch (classID) {
-        case DigitClassID:
-            m_alternative->m_terms.append(PatternTerm(m_pattern.digitsCharacterClass(), invert));
-            break;
-        case SpaceClassID:
-            m_alternative->m_terms.append(PatternTerm(m_pattern.spacesCharacterClass(), invert));
-            break;
-        case WordClassID:
-            m_alternative->m_terms.append(PatternTerm(m_pattern.wordcharCharacterClass(), invert));
-            break;
-        case NewlineClassID:
-            m_alternative->m_terms.append(PatternTerm(m_pattern.newlineCharacterClass(), invert));
-            break;
-        }
-    }
-
-    void atomCharacterClassBegin(bool invert = false)
-    {
-        m_invertCharacterClass = invert;
-    }
-
-    void atomCharacterClassAtom(UChar ch)
-    {
-        m_characterClassConstructor.putChar(ch);
-    }
-
-    void atomCharacterClassRange(UChar begin, UChar end)
-    {
-        m_characterClassConstructor.putRange(begin, end);
-    }
-
-    void atomCharacterClassBuiltIn(BuiltInCharacterClassID classID, bool invert)
-    {
-        ASSERT(classID != NewlineClassID);
-
-        switch (classID) {
-        case DigitClassID:
-            m_characterClassConstructor.append(invert ? m_pattern.nondigitsCharacterClass() : m_pattern.digitsCharacterClass());
-            break;
-        
-        case SpaceClassID:
-            m_characterClassConstructor.append(invert ? m_pattern.nonspacesCharacterClass() : m_pattern.spacesCharacterClass());
-            break;
-        
-        case WordClassID:
-            m_characterClassConstructor.append(invert ? m_pattern.nonwordcharCharacterClass() : m_pattern.wordcharCharacterClass());
-            break;
-        
-        default:
-            ASSERT_NOT_REACHED();
-        }
-    }
-
-    void atomCharacterClassEnd()
-    {
-        CharacterClass* newCharacterClass = m_characterClassConstructor.charClass();
-        m_pattern.m_userCharacterClasses.append(newCharacterClass);
-        m_alternative->m_terms.append(PatternTerm(newCharacterClass, m_invertCharacterClass));
-    }
-
-    void atomParenthesesSubpatternBegin(bool capture = true)
-    {
-        unsigned subpatternId = m_pattern.m_numSubpatterns + 1;
-        if (capture)
-            m_pattern.m_numSubpatterns++;
-
-        PatternDisjunction* parenthesesDisjunction = new PatternDisjunction(m_alternative);
-        m_pattern.m_disjunctions.append(parenthesesDisjunction);
-        m_alternative->m_terms.append(PatternTerm(PatternTerm::TypeParenthesesSubpattern, subpatternId, parenthesesDisjunction, capture));
-        m_alternative = parenthesesDisjunction->addNewAlternative();
-    }
-
-    void atomParentheticalAssertionBegin(bool invert = false)
-    {
-        PatternDisjunction* parenthesesDisjunction = new PatternDisjunction(m_alternative);
-        m_pattern.m_disjunctions.append(parenthesesDisjunction);
-        m_alternative->m_terms.append(PatternTerm(PatternTerm::TypeParentheticalAssertion, m_pattern.m_numSubpatterns + 1, parenthesesDisjunction, invert));
-        m_alternative = parenthesesDisjunction->addNewAlternative();
-    }
-
-    void atomParenthesesEnd()
-    {
-        ASSERT(m_alternative->m_parent);
-        ASSERT(m_alternative->m_parent->m_parent);
-        m_alternative = m_alternative->m_parent->m_parent;
-        
-        m_alternative->lastTerm().parentheses.lastSubpatternId = m_pattern.m_numSubpatterns;
-    }
-
-    void atomBackReference(unsigned subpatternId)
-    {
-        ASSERT(subpatternId);
-        m_pattern.m_shouldFallBack = true;
-        m_pattern.m_maxBackReference = std::max(m_pattern.m_maxBackReference, subpatternId);
-
-        if (subpatternId > m_pattern.m_numSubpatterns) {
-            m_alternative->m_terms.append(PatternTerm::ForwardReference());
-            return;
-        }
-
-        PatternAlternative* currentAlternative = m_alternative;
-        ASSERT(currentAlternative);
-
-        // Note to self: if we waited until the AST was baked, we could also remove forwards refs 
-        while ((currentAlternative = currentAlternative->m_parent->m_parent)) {
-            PatternTerm& term = currentAlternative->lastTerm();
-            ASSERT((term.type == PatternTerm::TypeParenthesesSubpattern) || (term.type == PatternTerm::TypeParentheticalAssertion));
-
-            if ((term.type == PatternTerm::TypeParenthesesSubpattern) && term.invertOrCapture && (subpatternId == term.subpatternId)) {
-                m_alternative->m_terms.append(PatternTerm::ForwardReference());
-                return;
-            }
-        }
-
-        m_alternative->m_terms.append(PatternTerm(subpatternId));
-    }
-
-    PatternDisjunction* copyDisjunction(PatternDisjunction* disjunction)
-    {
-        PatternDisjunction* newDisjunction = new PatternDisjunction();
-
-        newDisjunction->m_parent = disjunction->m_parent;
-        for (unsigned alt = 0; alt < disjunction->m_alternatives.size(); ++alt) {
-            PatternAlternative* alternative = disjunction->m_alternatives[alt];
-            PatternAlternative* newAlternative = newDisjunction->addNewAlternative();
-            for (unsigned i = 0; i < alternative->m_terms.size(); ++i)
-                newAlternative->m_terms.append(copyTerm(alternative->m_terms[i]));
-        }
-
-        m_pattern.m_disjunctions.append(newDisjunction);
-        return newDisjunction;
-    }
-
-    PatternTerm copyTerm(PatternTerm& term)
-    {
-        if ((term.type != PatternTerm::TypeParenthesesSubpattern) && (term.type != PatternTerm::TypeParentheticalAssertion))
-            return PatternTerm(term);
-
-        PatternTerm termCopy = term;
-        termCopy.parentheses.disjunction = copyDisjunction(termCopy.parentheses.disjunction);
-        return termCopy;
-    }
-
-    void quantifyAtom(unsigned min, unsigned max, bool greedy)
-    {
-        ASSERT(min <= max);
-        ASSERT(m_alternative->m_terms.size());
-
-        if (!max) {
-            m_alternative->removeLastTerm();
-            return;
-        }
-
-        PatternTerm& term = m_alternative->lastTerm();
-        ASSERT(term.type > PatternTerm::TypeAssertionWordBoundary);
-        ASSERT((term.quantityCount == 1) && (term.quantityType == QuantifierFixedCount));
-
-        // For any assertion with a zero minimum, not matching is valid and has no effect,
-        // remove it.  Otherwise, we need to match as least once, but there is no point
-        // matching more than once, so remove the quantifier.  It is not entirely clear
-        // from the spec whether or not this behavior is correct, but I believe this
-        // matches Firefox. :-/
-        if (term.type == PatternTerm::TypeParentheticalAssertion) {
-            if (!min)
-                m_alternative->removeLastTerm();
-            return;
-        }
-
-        if (max > 1 && term.type == PatternTerm::TypeParenthesesSubpattern)
-            m_pattern.m_shouldFallBack = true;
-
-        if (min == 0)
-            term.quantify(max, greedy   ? QuantifierGreedy : QuantifierNonGreedy);
-        else if (min == max)
-            term.quantify(min, QuantifierFixedCount);
-        else {
-            term.quantify(min, QuantifierFixedCount);
-            m_alternative->m_terms.append(copyTerm(term));
-            // NOTE: this term is interesting from an analysis perspective, in that it can be ignored.....
-            m_alternative->lastTerm().quantify((max == UINT_MAX) ? max : max - min, greedy ? QuantifierGreedy : QuantifierNonGreedy);
-            if (m_alternative->lastTerm().type == PatternTerm::TypeParenthesesSubpattern)
-                m_alternative->lastTerm().parentheses.isCopy = true;
-        }
-    }
-
-    void disjunction()
-    {
-        m_alternative = m_alternative->m_parent->addNewAlternative();
-    }
-
-    void regexBegin()
-    {
-        m_pattern.m_body = new PatternDisjunction();
-        m_alternative = m_pattern.m_body->addNewAlternative();
-        m_pattern.m_disjunctions.append(m_pattern.m_body);
-    }
-    void regexEnd()
-    {
-    }
-    void regexError()
-    {
-    }
-
-    unsigned setupAlternativeOffsets(PatternAlternative* alternative, unsigned currentCallFrameSize, unsigned initialInputPosition)
-    {
-        alternative->m_hasFixedSize = true;
-        unsigned currentInputPosition = initialInputPosition;
-
-        for (unsigned i = 0; i < alternative->m_terms.size(); ++i) {
-            PatternTerm& term = alternative->m_terms[i];
-
-            switch (term.type) {
-            case PatternTerm::TypeAssertionBOL:
-            case PatternTerm::TypeAssertionEOL:
-            case PatternTerm::TypeAssertionWordBoundary:
-                term.inputPosition = currentInputPosition;
-                break;
-
-            case PatternTerm::TypeBackReference:
-                term.inputPosition = currentInputPosition;
-                term.frameLocation = currentCallFrameSize;
-                currentCallFrameSize += RegexStackSpaceForBackTrackInfoBackReference;
-                alternative->m_hasFixedSize = false;
-                break;
-
-            case PatternTerm::TypeForwardReference:
-                break;
-
-            case PatternTerm::TypePatternCharacter:
-                term.inputPosition = currentInputPosition;
-                if (term.quantityType != QuantifierFixedCount) {
-                    term.frameLocation = currentCallFrameSize;
-                    currentCallFrameSize += RegexStackSpaceForBackTrackInfoPatternCharacter;
-                    alternative->m_hasFixedSize = false;
-                } else
-                    currentInputPosition += term.quantityCount;
-                break;
-
-            case PatternTerm::TypeCharacterClass:
-                term.inputPosition = currentInputPosition;
-                if (term.quantityType != QuantifierFixedCount) {
-                    term.frameLocation = currentCallFrameSize;
-                    currentCallFrameSize += RegexStackSpaceForBackTrackInfoCharacterClass;
-                    alternative->m_hasFixedSize = false;
-                } else
-                    currentInputPosition += term.quantityCount;
-                break;
-
-            case PatternTerm::TypeParenthesesSubpattern:
-                // Note: for fixed once parentheses we will ensure at least the minimum is available; others are on their own.
-                term.frameLocation = currentCallFrameSize;
-                if ((term.quantityCount == 1) && !term.parentheses.isCopy) {
-                    if (term.quantityType == QuantifierFixedCount) {
-                        currentCallFrameSize = setupDisjunctionOffsets(term.parentheses.disjunction, currentCallFrameSize, currentInputPosition);
-                        currentInputPosition += term.parentheses.disjunction->m_minimumSize;
-                    } else {
-                        currentCallFrameSize += RegexStackSpaceForBackTrackInfoParenthesesOnce;
-                        currentCallFrameSize = setupDisjunctionOffsets(term.parentheses.disjunction, currentCallFrameSize, currentInputPosition);
-                    }
-                    term.inputPosition = currentInputPosition;
-                } else {
-                    term.inputPosition = currentInputPosition;
-                    setupDisjunctionOffsets(term.parentheses.disjunction, 0, currentInputPosition);
-                    currentCallFrameSize += RegexStackSpaceForBackTrackInfoParentheses;
-                }
-                // Fixed count of 1 could be accepted, if they have a fixed size *AND* if all alternatives are of the same length.
-                alternative->m_hasFixedSize = false;
-                break;
-
-            case PatternTerm::TypeParentheticalAssertion:
-                term.inputPosition = currentInputPosition;
-                term.frameLocation = currentCallFrameSize;
-                currentCallFrameSize = setupDisjunctionOffsets(term.parentheses.disjunction, currentCallFrameSize + RegexStackSpaceForBackTrackInfoParentheticalAssertion, currentInputPosition);
-                break;
-            }
-        }
-
-        alternative->m_minimumSize = currentInputPosition - initialInputPosition;
-        return currentCallFrameSize;
-    }
-
-    unsigned setupDisjunctionOffsets(PatternDisjunction* disjunction, unsigned initialCallFrameSize, unsigned initialInputPosition)
-    {
-        if ((disjunction != m_pattern.m_body) && (disjunction->m_alternatives.size() > 1))
-            initialCallFrameSize += RegexStackSpaceForBackTrackInfoAlternative;
-
-        unsigned minimumInputSize = UINT_MAX;
-        unsigned maximumCallFrameSize = 0;
-        bool hasFixedSize = true;
-
-        for (unsigned alt = 0; alt < disjunction->m_alternatives.size(); ++alt) {
-            PatternAlternative* alternative = disjunction->m_alternatives[alt];
-            unsigned currentAlternativeCallFrameSize = setupAlternativeOffsets(alternative, initialCallFrameSize, initialInputPosition);
-            minimumInputSize = min(minimumInputSize, alternative->m_minimumSize);
-            maximumCallFrameSize = max(maximumCallFrameSize, currentAlternativeCallFrameSize);
-            hasFixedSize &= alternative->m_hasFixedSize;
-        }
-        
-        ASSERT(minimumInputSize != UINT_MAX);
-        ASSERT(maximumCallFrameSize >= initialCallFrameSize);
-
-        disjunction->m_hasFixedSize = hasFixedSize;
-        disjunction->m_minimumSize = minimumInputSize;
-        disjunction->m_callFrameSize = maximumCallFrameSize;
-        return maximumCallFrameSize;
-    }
-
-    void setupOffsets()
-    {
-        setupDisjunctionOffsets(m_pattern.m_body, 0, 0);
-    }
-
-private:
-    RegexPattern& m_pattern;
-    PatternAlternative* m_alternative;
-    CharacterClassConstructor m_characterClassConstructor;
-    bool m_invertCharacterClass;
-};
-
-
-const char* compileRegex(const UString& patternString, RegexPattern& pattern)
-{
-    RegexPatternConstructor constructor(pattern);
-
-    if (const char* error = parse(constructor, patternString))
-        return error;
-    
-    // If the pattern contains illegal backreferences reset & reparse.
-    // Quoting Netscape's "What's new in JavaScript 1.2",
-    //      "Note: if the number of left parentheses is less than the number specified
-    //       in \#, the \# is taken as an octal escape as described in the next row."
-    if (pattern.containsIllegalBackReference()) {
-        unsigned numSubpatterns = pattern.m_numSubpatterns;
-
-        constructor.reset();
-#if !ASSERT_DISABLED
-        const char* error =
-#endif
-            parse(constructor, patternString, numSubpatterns);
-
-        ASSERT(!error);
-        ASSERT(numSubpatterns == pattern.m_numSubpatterns);
-    }
-
-    constructor.setupOffsets();
-
-    return false;
-};
-
-
-} }
-
-#endif