Changeset 243950 in webkit

Timestamp:

Apr 5, 2019, 3:08:26 PM (7 years ago)

Author:

Alan Coon

Message:

Revert r243642. rdar://problem/49654398

Location:

tags/Safari-608.1.15/Source/JavaScriptCore

Files:

• ChangeLog (modified) (1 diff)
• runtime/RegExp.cpp (modified) (2 diffs)
• runtime/RegExpInlines.h (modified) (1 diff)
• yarr/YarrInterpreter.cpp (modified) (2 diffs)
• yarr/YarrJIT.cpp (modified) (23 diffs)
• yarr/YarrPattern.cpp (modified) (10 diffs)
• yarr/YarrPattern.h (modified) (6 diffs)
• yarr/create_regex_tables (modified) (2 diffs)
• yarr/generateYarrUnicodePropertyTables.py (modified) (5 diffs)

tags/Safari-608.1.15/Source/JavaScriptCore/ChangeLog

@@ +1 @@
+2019-04-05  Kocsen Chung  <kocsen_chung@apple.com>
+
+        Revert r243642. rdar://problem/49654398
+
 2019-04-05  Kocsen Chung  <kocsen_chung@apple.com>
 

tags/Safari-608.1.15/Source/JavaScriptCore/runtime/RegExp.cpp

@@ -386 +386 @@
         interpreterOffsetVector[j] = -1;
 
-    interpreterResult = Yarr::interpret(m_regExpBytecode.get(), s, startOffset, reinterpret_cast<unsigned*>(interpreterOffsetVector));
+    interpreterResult = Yarr::interpret(m_regExpBytecode.get(), s, startOffset, interpreterOffsetVector);
 
     if (jitResult != interpreterResult)
@@ -403 +403 @@
 
         if (jitResult != interpreterResult) {
-            dataLogF("    JIT result = %d, interpreted result = %d\n", jitResult, interpreterResult);
+            dataLogF("    JIT result = %d, blah interpreted result = %d\n", jitResult, interpreterResult);
             differences--;
         } else {

tags/Safari-608.1.15/Source/JavaScriptCore/runtime/RegExpInlines.h

@@ -182 +182 @@
 
 #if ENABLE(YARR_JIT_DEBUG)
-        if (m_state == JITCode) {
-            byteCodeCompileIfNecessary(&vm);
-            matchCompareWithInterpreter(s, startOffset, offsetVector, result);
-        }
+        matchCompareWithInterpreter(s, startOffset, offsetVector, result);
 #endif
     } else

tags/Safari-608.1.15/Source/JavaScriptCore/yarr/YarrInterpreter.cpp

@@ -429 +429 @@
         return invert ? !match : match;
     }
-    
-    bool checkCharacterClassDontAdvanceInputForNonBMP(CharacterClass* characterClass, unsigned negativeInputOffset)
-    {
-        int readCharacter = characterClass->hasOnlyNonBMPCharacters() ? input.readSurrogatePairChecked(negativeInputOffset) :  input.readChecked(negativeInputOffset);
-        return testCharacterClass(characterClass, readCharacter);
-    }
 
     bool tryConsumeBackReference(int matchBegin, int matchEnd, unsigned negativeInputOffset)
@@ -565 +559 @@
         case QuantifierFixedCount: {
             if (unicode) {
-                CharacterClass* charClass = term.atom.characterClass;
                 backTrack->begin = input.getPos();
                 unsigned matchAmount = 0;
                 for (matchAmount = 0; matchAmount < term.atom.quantityMaxCount; ++matchAmount) {
-                    if (term.invert()) {
-                        if (!checkCharacterClass(charClass, term.invert(), term.inputPosition - matchAmount)) {
-                            input.setPos(backTrack->begin);
-                            return false;
-                        }
-                    } else {
-                        unsigned matchOffset = matchAmount * (charClass->hasOnlyNonBMPCharacters() ? 2 : 1);
-                        if (!checkCharacterClassDontAdvanceInputForNonBMP(charClass, term.inputPosition - matchOffset)) {
-                            input.setPos(backTrack->begin);
-                            return false;
-                        }
+                    if (!checkCharacterClass(term.atom.characterClass, term.invert(), term.inputPosition - matchAmount)) {
+                        input.setPos(backTrack->begin);
+                        return false;
                     }
                 }

tags/Safari-608.1.15/Source/JavaScriptCore/yarr/YarrJIT.cpp

@@ -73 +73 @@
     static const RegisterID initialStart = ARM64Registers::x11;
     static const RegisterID supplementaryPlanesBase = ARM64Registers::x12;
-    static const RegisterID leadingSurrogateTag = ARM64Registers::x13;
-    static const RegisterID trailingSurrogateTag = ARM64Registers::x14;
-    static const RegisterID endOfStringAddress = ARM64Registers::x15;
+    static const RegisterID surrogateTagMask = ARM64Registers::x13;
+    static const RegisterID leadingSurrogateTag = ARM64Registers::x14;
+    static const RegisterID trailingSurrogateTag = ARM64Registers::x15;
 
     static const RegisterID returnRegister = ARM64Registers::x0;
     static const RegisterID returnRegister2 = ARM64Registers::x1;
 
-    const TrustedImm32 surrogateTagMask = TrustedImm32(0xfffffc00);
 #define HAVE_INITIAL_START_REG
 #define JIT_UNICODE_EXPRESSIONS
@@ -145 +144 @@
     static const RegisterID regUnicodeInputAndTrail = X86Registers::r13;
     static const RegisterID leadingSurrogateTag = X86Registers::r14;
-    static const RegisterID endOfStringAddress = X86Registers::r15;
+    static const RegisterID trailingSurrogateTag = X86Registers::r15;
 
     static const RegisterID returnRegister = X86Registers::eax;
@@ -151 +150 @@
 
     const TrustedImm32 supplementaryPlanesBase = TrustedImm32(0x10000);
-    const TrustedImm32 trailingSurrogateTag = TrustedImm32(0xdc00);
     const TrustedImm32 surrogateTagMask = TrustedImm32(0xfffffc00);
 #define HAVE_INITIAL_START_REG
@@ -322 +320 @@
             if ((term.type == PatternTerm::TypeCharacterClass)
                 && (term.quantityType == QuantifierFixedCount)
-                && (!m_decodeSurrogatePairs || (term.characterClass->hasOneCharacterSize() && !term.m_invert))
+                && (!m_decodeSurrogatePairs || (!term.characterClass->m_hasNonBMPCharacters && !term.m_invert))
                 && (nextTerm.type == PatternTerm::TypePatternCharacter)
                 && (nextTerm.quantityType == QuantifierFixedCount)) {
@@ -386 +384 @@
             return;
         }
-
         JumpList unicodeFail;
         if (charClass->m_matchesUnicode.size() || charClass->m_rangesUnicode.size()) {
@@ -451 +448 @@
             unicodeFail.link(this);
     }
-
-#ifdef JIT_UNICODE_EXPRESSIONS
-    void advanceIndexAfterCharacterClassTermMatch(const PatternTerm* term, JumpList& failures, const RegisterID character)
-    {
-        ASSERT(term->type == PatternTerm::TypeCharacterClass);
-
-        if (term->characterClass->hasOneCharacterSize() && !term->invert())
-            add32(TrustedImm32(term->characterClass->hasNonBMPCharacters() ? 2 : 1), index);
-        else {
-            add32(TrustedImm32(1), index);
-            failures.append(atEndOfInput());
-            Jump isBMPChar = branch32(LessThan, character, supplementaryPlanesBase);
-            add32(TrustedImm32(1), index);
-            isBMPChar.link(this);
-        }
-    }
-#endif
 
     // Jumps if input not available; will have (incorrectly) incremented already!
@@ -541 +521 @@
 
         JumpList notUnicode;
-
         load16Unaligned(regUnicodeInputAndTrail, resultReg);
         and32(surrogateTagMask, resultReg, regT2);
         notUnicode.append(branch32(NotEqual, regT2, leadingSurrogateTag));
         addPtr(TrustedImm32(2), regUnicodeInputAndTrail);
-        notUnicode.append(branchPtr(AboveOrEqual, regUnicodeInputAndTrail, endOfStringAddress));
+        getEffectiveAddress(BaseIndex(input, length, TimesTwo), regT2);
+        notUnicode.append(branch32(AboveOrEqual, regUnicodeInputAndTrail, regT2));
         load16Unaligned(Address(regUnicodeInputAndTrail), regUnicodeInputAndTrail);
         and32(surrogateTagMask, regUnicodeInputAndTrail, regT2);
@@ -1755 +1735 @@
         }
 #ifdef JIT_UNICODE_EXPRESSIONS
-        if (m_decodeSurrogatePairs && (!term->characterClass->hasOneCharacterSize() || term->invert())) {
+        if (m_decodeSurrogatePairs) {
             Jump isBMPChar = branch32(LessThan, character, supplementaryPlanesBase);
             add32(TrustedImm32(1), index);
@@ -1789 +1769 @@
 
         move(index, countRegister);
-
-        Checked<unsigned> scaledMaxCount = term->quantityMaxCount;
-
-#ifdef JIT_UNICODE_EXPRESSIONS
-        if (m_decodeSurrogatePairs && term->characterClass->hasOnlyNonBMPCharacters() && !term->invert())
-            scaledMaxCount *= 2;
-#endif
-        sub32(Imm32(scaledMaxCount.unsafeGet()), countRegister);
+        sub32(Imm32(term->quantityMaxCount.unsafeGet()), countRegister);
 
         Label loop(this);
         JumpList matchDest;
-        readCharacter(m_checkedOffset - term->inputPosition - scaledMaxCount, character, countRegister);
+        readCharacter(m_checkedOffset - term->inputPosition - term->quantityMaxCount, character, countRegister);
         // If we are matching the "any character" builtin class we only need to read the
         // character and don't need to match as it will always succeed.
@@ -1814 +1787 @@
         }
 
+        add32(TrustedImm32(1), countRegister);
 #ifdef JIT_UNICODE_EXPRESSIONS
         if (m_decodeSurrogatePairs) {
-            if (term->characterClass->hasOneCharacterSize() && !term->invert())
-                add32(TrustedImm32(term->characterClass->hasNonBMPCharacters() ? 2 : 1), countRegister);
-            else {
-                add32(TrustedImm32(1), countRegister);
-                Jump isBMPChar = branch32(LessThan, character, supplementaryPlanesBase);
-                op.m_jumps.append(atEndOfInput());
-                add32(TrustedImm32(1), countRegister);
-                add32(TrustedImm32(1), index);
-                isBMPChar.link(this);
-            }
-        } else
-#endif
+            Jump isBMPChar = branch32(LessThan, character, supplementaryPlanesBase);
+            op.m_jumps.append(atEndOfInput());
             add32(TrustedImm32(1), countRegister);
+            add32(TrustedImm32(1), index);
+            isBMPChar.link(this);
+        }
+#endif
         branch32(NotEqual, countRegister, index).linkTo(loop, this);
     }
@@ -1844 +1812 @@
         const RegisterID countRegister = regT1;
 
-        if (m_decodeSurrogatePairs && (!term->characterClass->hasOneCharacterSize() || term->invert()))
+        if (m_decodeSurrogatePairs)
             storeToFrame(index, term->frameLocation + BackTrackInfoCharacterClass::beginIndex());
         move(TrustedImm32(0), countRegister);
@@ -1858 +1826 @@
             JumpList matchDest;
             readCharacter(m_checkedOffset - term->inputPosition, character);
-            // If we are matching the "any character" builtin class for non-unicode patterns,
-            // we only need to read the character and don't need to match as it will always succeed.
+            // If we are matching the "any character" builtin class we only need to read the
+            // character and don't need to match as it will always succeed.
             if (!term->characterClass->m_anyCharacter) {
                 matchCharacterClass(character, matchDest, term->characterClass);
@@ -1867 +1835 @@
         }
 
+        add32(TrustedImm32(1), index);
 #ifdef JIT_UNICODE_EXPRESSIONS
-        if (m_decodeSurrogatePairs)
-            advanceIndexAfterCharacterClassTermMatch(term, failures, character);
-        else
-#endif
+        if (m_decodeSurrogatePairs) {
+            failures.append(atEndOfInput());
+            Jump isBMPChar = branch32(LessThan, character, supplementaryPlanesBase);
             add32(TrustedImm32(1), index);
+            isBMPChar.link(this);
+        }
+#endif
         add32(TrustedImm32(1), countRegister);
 
@@ -1898 +1869 @@
         m_backtrackingState.append(branchTest32(Zero, countRegister));
         sub32(TrustedImm32(1), countRegister);
-        storeToFrame(countRegister, term->frameLocation + BackTrackInfoCharacterClass::matchAmountIndex());
-
         if (!m_decodeSurrogatePairs)
             sub32(TrustedImm32(1), index);
-        else if (term->characterClass->hasOneCharacterSize() && !term->invert())
-            sub32(TrustedImm32(term->characterClass->hasNonBMPCharacters() ? 2 : 1), index);
         else {
+            const RegisterID character = regT0;
+
+            loadFromFrame(term->frameLocation + BackTrackInfoCharacterClass::beginIndex(), index);
             // Rematch one less
-            const RegisterID character = regT0;
-
-            loadFromFrame(term->frameLocation + BackTrackInfoCharacterClass::beginIndex(), index);
+            storeToFrame(countRegister, term->frameLocation + BackTrackInfoCharacterClass::matchAmountIndex());
 
             Label rematchLoop(this);
@@ -1938 +1906 @@
         move(TrustedImm32(0), countRegister);
         op.m_reentry = label();
-        if (m_decodeSurrogatePairs) {
-            if (!term->characterClass->hasOneCharacterSize() || term->invert())
-                storeToFrame(index, term->frameLocation + BackTrackInfoCharacterClass::beginIndex());
-            storeToFrame(countRegister, term->frameLocation + BackTrackInfoCharacterClass::matchAmountIndex());
-        }
+        if (m_decodeSurrogatePairs)
+            storeToFrame(index, term->frameLocation + BackTrackInfoCharacterClass::beginIndex());
+        storeToFrame(countRegister, term->frameLocation + BackTrackInfoCharacterClass::matchAmountIndex());
     }
 
@@ -1957 +1923 @@
         m_backtrackingState.link(this);
 
-        if (m_decodeSurrogatePairs) {
-            if (!term->characterClass->hasOneCharacterSize() || term->invert())
-                loadFromFrame(term->frameLocation + BackTrackInfoCharacterClass::beginIndex(), index);
-            loadFromFrame(term->frameLocation + BackTrackInfoCharacterClass::matchAmountIndex(), countRegister);
-        }
+        if (m_decodeSurrogatePairs)
+            loadFromFrame(term->frameLocation + BackTrackInfoCharacterClass::beginIndex(), index);
+        loadFromFrame(term->frameLocation + BackTrackInfoCharacterClass::matchAmountIndex(), countRegister);
 
         nonGreedyFailures.append(atEndOfInput());
@@ -1968 +1932 @@
         JumpList matchDest;
         readCharacter(m_checkedOffset - term->inputPosition, character);
-        // If we are matching the "any character" builtin class for non-unicode patterns,
-        // we only need to read the character and don't need to match as it will always succeed.
+        // If we are matching the "any character" builtin class we only need to read the
+        // character and don't need to match as it will always succeed.
         if (term->invert() || !term->characterClass->m_anyCharacter) {
             matchCharacterClass(character, matchDest, term->characterClass);
@@ -1981 +1945 @@
         }
 
+        add32(TrustedImm32(1), index);
 #ifdef JIT_UNICODE_EXPRESSIONS
-        if (m_decodeSurrogatePairs)
-            advanceIndexAfterCharacterClassTermMatch(term, nonGreedyFailures, character);
-        else
-#endif
+        if (m_decodeSurrogatePairs) {
+            nonGreedyFailures.append(atEndOfInput());
+            Jump isBMPChar = branch32(LessThan, character, supplementaryPlanesBase);
             add32(TrustedImm32(1), index);
+            isBMPChar.link(this);
+        }
+#endif
         add32(TrustedImm32(1), countRegister);
 
@@ -3734 +3701 @@
 
             move(TrustedImm32(0xd800), leadingSurrogateTag);
+            move(TrustedImm32(0xdc00), trailingSurrogateTag);
         }
         // The ABI doesn't guarantee the upper bits are zero on unsigned arguments, so clear them ourselves.
@@ -3767 +3735 @@
             pushPair(framePointerRegister, linkRegister);
             move(TrustedImm32(0x10000), supplementaryPlanesBase);
+            move(TrustedImm32(0xfffffc00), surrogateTagMask);
             move(TrustedImm32(0xd800), leadingSurrogateTag);
             move(TrustedImm32(0xdc00), trailingSurrogateTag);
@@ -3847 +3816 @@
         , m_decodeSurrogatePairs(m_charSize == Char16 && m_pattern.unicode())
         , m_unicodeIgnoreCase(m_pattern.unicode() && m_pattern.ignoreCase())
-        , m_fixedSizedAlternative(false)
         , m_canonicalMode(m_pattern.unicode() ? CanonicalMode::Unicode : CanonicalMode::UCS2)
 #if ENABLE(YARR_JIT_ALL_PARENS_EXPRESSIONS)
@@ -3901 +3869 @@
         generateFailReturn();
         hasInput.link(this);
-
-#ifdef JIT_UNICODE_EXPRESSIONS
-        if (m_decodeSurrogatePairs)
-            getEffectiveAddress(BaseIndex(input, length, TimesTwo), endOfStringAddress);
-#endif
 
 #if ENABLE(YARR_JIT_ALL_PARENS_EXPRESSIONS)
@@ -4201 +4164 @@
     bool m_decodeSurrogatePairs;
     bool m_unicodeIgnoreCase;
-    bool m_fixedSizedAlternative;
     CanonicalMode m_canonicalMode;
 #if ENABLE(YARR_JIT_ALL_PARENS_EXPRESSIONS)

tags/Safari-608.1.15/Source/JavaScriptCore/yarr/YarrPattern.cpp

@@ -46 +46 @@
     CharacterClassConstructor(bool isCaseInsensitive, CanonicalMode canonicalMode)
         : m_isCaseInsensitive(isCaseInsensitive)
+        , m_hasNonBMPCharacters(false)
         , m_anyCharacter(false)
-        , m_characterWidths(CharacterClassWidths::Unknown)
         , m_canonicalMode(canonicalMode)
     {
@@ -58 +58 @@
         m_matchesUnicode.clear();
         m_rangesUnicode.clear();
+        m_hasNonBMPCharacters = false;
         m_anyCharacter = false;
-        m_characterWidths = CharacterClassWidths::Unknown;
     }
 
@@ -247 +247 @@
         characterClass->m_matchesUnicode.swap(m_matchesUnicode);
         characterClass->m_rangesUnicode.swap(m_rangesUnicode);
+        characterClass->m_hasNonBMPCharacters = hasNonBMPCharacters();
         characterClass->m_anyCharacter = anyCharacter();
-        characterClass->m_characterWidths = characterWidths();
-
+
+        m_hasNonBMPCharacters = false;
         m_anyCharacter = false;
-        m_characterWidths = CharacterClassWidths::Unknown;
 
         return characterClass;
@@ -267 +267 @@
         unsigned range = matches.size();
 
-        m_characterWidths |= (U_IS_BMP(ch) ? CharacterClassWidths::HasBMPChars : CharacterClassWidths::HasNonBMPChars);
+        if (!U_IS_BMP(ch))
+            m_hasNonBMPCharacters = true;
 
         // binary chop, find position to insert char.
@@ -316 +317 @@
         size_t end = ranges.size();
 
-        if (U_IS_BMP(lo))
-            m_characterWidths |= CharacterClassWidths::HasBMPChars;
         if (!U_IS_BMP(hi))
-            m_characterWidths |= CharacterClassWidths::HasNonBMPChars;
+            m_hasNonBMPCharacters = true;
 
         // Simple linear scan - I doubt there are that many ranges anyway...
@@ -410 +409 @@
     bool hasNonBMPCharacters()
     {
-        return m_characterWidths & CharacterClassWidths::HasNonBMPChars;
-    }
-
-    CharacterClassWidths characterWidths()
-    {
-        return m_characterWidths;
+        return m_hasNonBMPCharacters;
     }
 
@@ -424 +418 @@
 
     bool m_isCaseInsensitive : 1;
+    bool m_hasNonBMPCharacters : 1;
     bool m_anyCharacter : 1;
-    CharacterClassWidths m_characterWidths;
-    
     CanonicalMode m_canonicalMode;
 
@@ -844 +837 @@
                     term.frameLocation = currentCallFrameSize;
                     currentCallFrameSize += YarrStackSpaceForBackTrackInfoCharacterClass;
-                    if (term.characterClass->hasOneCharacterSize() && !term.invert()) {
-                        Checked<unsigned, RecordOverflow> tempCount = term.quantityMaxCount;
-                        tempCount *= term.characterClass->hasNonBMPCharacters() ? 2 : 1;
-                        if (tempCount.hasOverflowed())
-                            return ErrorCode::OffsetTooLarge;
-                        currentInputPosition += tempCount;
-                    } else {
-                        currentInputPosition += term.quantityMaxCount;
-                        alternative->m_hasFixedSize = false;
-                    }
+                    currentInputPosition += term.quantityMaxCount;
+                    alternative->m_hasFixedSize = false;
                 } else
                     currentInputPosition += term.quantityMaxCount;
@@ -1334 +1319 @@
     case TypeCharacterClass:
         out.print("character class ");
-        out.printf("inputPosition %u ", inputPosition);
         dumpCharacterClass(out, thisPattern, characterClass);
         dumpQuantifier(out);
@@ -1478 +1462 @@
     characterClass->m_ranges.append(CharacterRange(0x00, 0x7f));
     characterClass->m_rangesUnicode.append(CharacterRange(0x0080, 0x10ffff));
-    characterClass->m_characterWidths = CharacterClassWidths::HasBothBMPAndNonBMP;
+    characterClass->m_hasNonBMPCharacters = true;
     characterClass->m_anyCharacter = true;
     return characterClass;

tags/Safari-608.1.15/Source/JavaScriptCore/yarr/YarrPattern.h

@@ -53 +53 @@
 };
 
-enum struct CharacterClassWidths : unsigned char {
-    Unknown = 0x0,
-    HasBMPChars = 0x1,
-    HasNonBMPChars = 0x2,
-    HasBothBMPAndNonBMP = HasBMPChars | HasNonBMPChars
-};
-
-inline CharacterClassWidths operator|(CharacterClassWidths lhs, CharacterClassWidths rhs)
-{
-    return static_cast<CharacterClassWidths>(static_cast<unsigned>(lhs) | static_cast<unsigned>(rhs));
-}
-
-inline bool operator&(CharacterClassWidths lhs, CharacterClassWidths rhs)
-{
-    return static_cast<unsigned>(lhs) & static_cast<unsigned>(rhs);
-}
-
-inline CharacterClassWidths& operator|=(CharacterClassWidths& lhs, CharacterClassWidths rhs)
-{
-    lhs = lhs | rhs;
-    return lhs;
-}
-
 struct CharacterClass {
     WTF_MAKE_FAST_ALLOCATED;
@@ -84 +61 @@
     CharacterClass()
         : m_table(0)
-        , m_characterWidths(CharacterClassWidths::Unknown)
+        , m_hasNonBMPCharacters(false)
         , m_anyCharacter(false)
     {
@@ -90 +67 @@
     CharacterClass(const char* table, bool inverted)
         : m_table(table)
-        , m_characterWidths(CharacterClassWidths::Unknown)
         , m_tableInverted(inverted)
+        , m_hasNonBMPCharacters(false)
         , m_anyCharacter(false)
     {
     }
-    CharacterClass(std::initializer_list<UChar32> matches, std::initializer_list<CharacterRange> ranges, std::initializer_list<UChar32> matchesUnicode, std::initializer_list<CharacterRange> rangesUnicode, CharacterClassWidths widths)
+    CharacterClass(std::initializer_list<UChar32> matches, std::initializer_list<CharacterRange> ranges, std::initializer_list<UChar32> matchesUnicode, std::initializer_list<CharacterRange> rangesUnicode)
         : m_matches(matches)
         , m_ranges(ranges)
@@ -101 +78 @@
         , m_rangesUnicode(rangesUnicode)
         , m_table(0)
-        , m_characterWidths(widths)
         , m_tableInverted(false)
+        , m_hasNonBMPCharacters(false)
         , m_anyCharacter(false)
     {
     }
 
-    bool hasNonBMPCharacters() { return m_characterWidths & CharacterClassWidths::HasNonBMPChars; }
-
-    bool hasOneCharacterSize() { return m_characterWidths == CharacterClassWidths::HasBMPChars || m_characterWidths == CharacterClassWidths::HasNonBMPChars; }
-    bool hasOnlyNonBMPCharacters() { return m_characterWidths == CharacterClassWidths::HasNonBMPChars; }
-    
     Vector<UChar32> m_matches;
     Vector<CharacterRange> m_ranges;
@@ -118 +90 @@
 
     const char* m_table;
-    CharacterClassWidths m_characterWidths;
     bool m_tableInverted : 1;
+    bool m_hasNonBMPCharacters : 1;
     bool m_anyCharacter : 1;
 };
@@ -249 +221 @@
     }
     
-    bool invert() const
+    bool invert()
     {
         return m_invert;

tags/Safari-608.1.15/Source/JavaScriptCore/yarr/create_regex_tables

@@ -101 +101 @@
     else:
         function += ("    auto characterClass = std::make_unique<CharacterClass>();\n")
-    hasBMPCharacters = False
     hasNonBMPCharacters = False
     for (min, max) in ranges:
-        if min < 0x10000:
-            hasBMPCharacters = True
-        if max >= 0x10000:
-            hasNonBMPCharacters = True
         if (min == max):
             if (min > 127):
@@ -118 +113 @@
         else:
             function += ("    characterClass->m_ranges.append(CharacterRange(0x%02x, 0x%02x));\n" % (min, max))
-    function += ("    characterClass->m_characterWidths = CharacterClassWidths::%s;\n" % (("Unknown", "HasBMPChars", "HasNonBMPChars", "HasBothBMPAndNonBMP")[int(hasNonBMPCharacters) * 2 + int(hasBMPCharacters)]))
+        if max >= 0x10000:
+            hasNonBMPCharacters = True
+    function += ("    characterClass->m_hasNonBMPCharacters = %s;\n" % ("true" if hasNonBMPCharacters else "false"))
     function += ("    return characterClass;\n")
     function += ("}\n\n")

tags/Safari-608.1.15/Source/JavaScriptCore/yarr/generateYarrUnicodePropertyTables.py

@@ -36 +36 @@
 
 header = """/*
-* Copyright (C) 2017-2019 Apple Inc. All rights reserved.
+* Copyright (C) 2017-2018 Apple Inc. All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
@@ -226 +226 @@
         self.aliases = []
         self.index = len(PropertyData.allPropertyData)
-        self.hasBMPCharacters = False
         self.hasNonBMPCharacters = False
         self.matches = []
@@ -251 +250 @@
 
     def addMatch(self, codePoint):
-        if codePoint <= MaxBMP:
-            self.hasBMPCharacters = True
-        else:
+        if codePoint > MaxBMP:
             self.hasNonBMPCharacters = True
         if codePoint <= lastASCIICodePoint:
@@ -285 +282 @@
 
     def addRange(self, lowCodePoint, highCodePoint):
-        if lowCodePoint <= MaxBMP:
-            self.hasBMPCharacters = True
         if highCodePoint > MaxBMP:
             self.hasNonBMPCharacters = True
@@ -542 +537 @@
         file.write("        std::initializer_list<CharacterRange>(")
         self.dumpMatchData(file, 4, self.unicodeRanges, lambda file, range: (file.write("{{{0:0=#6x}, {1:0=#6x}}}".format(range[0], range[1]))))
-        file.write("),\n")
-
-        file.write("        CharacterClassWidths::{});\n".format(("Unknown", "HasBMPChars", "HasNonBMPChars", "HasBothBMPAndNonBMP")[int(self.hasNonBMPCharacters) * 2 + int(self.hasBMPCharacters)]))
+        file.write("));\n")
+
+        file.write("    characterClass->m_hasNonBMPCharacters = {};\n".format(("false", "true")[self.hasNonBMPCharacters]))
         file.write("    return characterClass;\n}\n\n")