From 40736c5763bf61337c8c14e16d8587db021a87d4 Mon Sep 17 00:00:00 2001 From: Simon Hausmann Date: Fri, 6 Jan 2012 14:44:00 +0100 Subject: Imported WebKit commit 2ea9d364d0f6efa8fa64acf19f451504c59be0e4 (http://svn.webkit.org/repository/webkit/trunk@104285) --- Source/JavaScriptCore/parser/Lexer.cpp | 1543 ++++++++++++++++++++++++++++++++ 1 file changed, 1543 insertions(+) create mode 100644 Source/JavaScriptCore/parser/Lexer.cpp (limited to 'Source/JavaScriptCore/parser/Lexer.cpp') diff --git a/Source/JavaScriptCore/parser/Lexer.cpp b/Source/JavaScriptCore/parser/Lexer.cpp new file mode 100644 index 000000000..e38b52480 --- /dev/null +++ b/Source/JavaScriptCore/parser/Lexer.cpp @@ -0,0 +1,1543 @@ +/* + * Copyright (C) 1999-2000 Harri Porten (porten@kde.org) + * Copyright (C) 2006, 2007, 2008, 2009 Apple Inc. All Rights Reserved. + * Copyright (C) 2007 Cameron Zwarich (cwzwarich@uwaterloo.ca) + * Copyright (C) 2010 Zoltan Herczeg (zherczeg@inf.u-szeged.hu) + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Library General Public + * License as published by the Free Software Foundation; either + * version 2 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Library General Public License for more details. + * + * You should have received a copy of the GNU Library General Public License + * along with this library; see the file COPYING.LIB. If not, write to + * the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, + * Boston, MA 02110-1301, USA. + * + */ + +#include "config.h" +#include "Lexer.h" + +#include "JSFunction.h" + +#include "JSGlobalObjectFunctions.h" +#include "Identifier.h" +#include "NodeInfo.h" +#include "Nodes.h" +#include "dtoa.h" +#include +#include +#include +#include + +using namespace WTF; +using namespace Unicode; + +#include "KeywordLookup.h" +#include "Lexer.lut.h" +#include "Parser.h" + +namespace JSC { + +Keywords::Keywords(JSGlobalData* globalData) + : m_globalData(globalData) + , m_keywordTable(JSC::mainTable) +{ +} + +enum CharacterType { + // Types for the main switch + + // The first three types are fixed, and also used for identifying + // ASCII alpha and alphanumeric characters (see isIdentStart and isIdentPart). + CharacterIdentifierStart, + CharacterZero, + CharacterNumber, + + CharacterInvalid, + CharacterLineTerminator, + CharacterExclamationMark, + CharacterOpenParen, + CharacterCloseParen, + CharacterOpenBracket, + CharacterCloseBracket, + CharacterComma, + CharacterColon, + CharacterQuestion, + CharacterTilde, + CharacterQuote, + CharacterDot, + CharacterSlash, + CharacterBackSlash, + CharacterSemicolon, + CharacterOpenBrace, + CharacterCloseBrace, + + CharacterAdd, + CharacterSub, + CharacterMultiply, + CharacterModulo, + CharacterAnd, + CharacterXor, + CharacterOr, + CharacterLess, + CharacterGreater, + CharacterEqual, + + // Other types (only one so far) + CharacterWhiteSpace, +}; + +// 128 ASCII codes +static const unsigned short typesOfASCIICharacters[128] = { +/* 0 - Null */ CharacterInvalid, +/* 1 - Start of Heading */ CharacterInvalid, +/* 2 - Start of Text */ CharacterInvalid, +/* 3 - End of Text */ CharacterInvalid, +/* 4 - End of Transm. */ CharacterInvalid, +/* 5 - Enquiry */ CharacterInvalid, +/* 6 - Acknowledgment */ CharacterInvalid, +/* 7 - Bell */ CharacterInvalid, +/* 8 - Back Space */ CharacterInvalid, +/* 9 - Horizontal Tab */ CharacterWhiteSpace, +/* 10 - Line Feed */ CharacterLineTerminator, +/* 11 - Vertical Tab */ CharacterWhiteSpace, +/* 12 - Form Feed */ CharacterWhiteSpace, +/* 13 - Carriage Return */ CharacterLineTerminator, +/* 14 - Shift Out */ CharacterInvalid, +/* 15 - Shift In */ CharacterInvalid, +/* 16 - Data Line Escape */ CharacterInvalid, +/* 17 - Device Control 1 */ CharacterInvalid, +/* 18 - Device Control 2 */ CharacterInvalid, +/* 19 - Device Control 3 */ CharacterInvalid, +/* 20 - Device Control 4 */ CharacterInvalid, +/* 21 - Negative Ack. */ CharacterInvalid, +/* 22 - Synchronous Idle */ CharacterInvalid, +/* 23 - End of Transmit */ CharacterInvalid, +/* 24 - Cancel */ CharacterInvalid, +/* 25 - End of Medium */ CharacterInvalid, +/* 26 - Substitute */ CharacterInvalid, +/* 27 - Escape */ CharacterInvalid, +/* 28 - File Separator */ CharacterInvalid, +/* 29 - Group Separator */ CharacterInvalid, +/* 30 - Record Separator */ CharacterInvalid, +/* 31 - Unit Separator */ CharacterInvalid, +/* 32 - Space */ CharacterWhiteSpace, +/* 33 - ! */ CharacterExclamationMark, +/* 34 - " */ CharacterQuote, +/* 35 - # */ CharacterInvalid, +/* 36 - $ */ CharacterIdentifierStart, +/* 37 - % */ CharacterModulo, +/* 38 - & */ CharacterAnd, +/* 39 - ' */ CharacterQuote, +/* 40 - ( */ CharacterOpenParen, +/* 41 - ) */ CharacterCloseParen, +/* 42 - * */ CharacterMultiply, +/* 43 - + */ CharacterAdd, +/* 44 - , */ CharacterComma, +/* 45 - - */ CharacterSub, +/* 46 - . */ CharacterDot, +/* 47 - / */ CharacterSlash, +/* 48 - 0 */ CharacterZero, +/* 49 - 1 */ CharacterNumber, +/* 50 - 2 */ CharacterNumber, +/* 51 - 3 */ CharacterNumber, +/* 52 - 4 */ CharacterNumber, +/* 53 - 5 */ CharacterNumber, +/* 54 - 6 */ CharacterNumber, +/* 55 - 7 */ CharacterNumber, +/* 56 - 8 */ CharacterNumber, +/* 57 - 9 */ CharacterNumber, +/* 58 - : */ CharacterColon, +/* 59 - ; */ CharacterSemicolon, +/* 60 - < */ CharacterLess, +/* 61 - = */ CharacterEqual, +/* 62 - > */ CharacterGreater, +/* 63 - ? */ CharacterQuestion, +/* 64 - @ */ CharacterInvalid, +/* 65 - A */ CharacterIdentifierStart, +/* 66 - B */ CharacterIdentifierStart, +/* 67 - C */ CharacterIdentifierStart, +/* 68 - D */ CharacterIdentifierStart, +/* 69 - E */ CharacterIdentifierStart, +/* 70 - F */ CharacterIdentifierStart, +/* 71 - G */ CharacterIdentifierStart, +/* 72 - H */ CharacterIdentifierStart, +/* 73 - I */ CharacterIdentifierStart, +/* 74 - J */ CharacterIdentifierStart, +/* 75 - K */ CharacterIdentifierStart, +/* 76 - L */ CharacterIdentifierStart, +/* 77 - M */ CharacterIdentifierStart, +/* 78 - N */ CharacterIdentifierStart, +/* 79 - O */ CharacterIdentifierStart, +/* 80 - P */ CharacterIdentifierStart, +/* 81 - Q */ CharacterIdentifierStart, +/* 82 - R */ CharacterIdentifierStart, +/* 83 - S */ CharacterIdentifierStart, +/* 84 - T */ CharacterIdentifierStart, +/* 85 - U */ CharacterIdentifierStart, +/* 86 - V */ CharacterIdentifierStart, +/* 87 - W */ CharacterIdentifierStart, +/* 88 - X */ CharacterIdentifierStart, +/* 89 - Y */ CharacterIdentifierStart, +/* 90 - Z */ CharacterIdentifierStart, +/* 91 - [ */ CharacterOpenBracket, +/* 92 - \ */ CharacterBackSlash, +/* 93 - ] */ CharacterCloseBracket, +/* 94 - ^ */ CharacterXor, +/* 95 - _ */ CharacterIdentifierStart, +/* 96 - ` */ CharacterInvalid, +/* 97 - a */ CharacterIdentifierStart, +/* 98 - b */ CharacterIdentifierStart, +/* 99 - c */ CharacterIdentifierStart, +/* 100 - d */ CharacterIdentifierStart, +/* 101 - e */ CharacterIdentifierStart, +/* 102 - f */ CharacterIdentifierStart, +/* 103 - g */ CharacterIdentifierStart, +/* 104 - h */ CharacterIdentifierStart, +/* 105 - i */ CharacterIdentifierStart, +/* 106 - j */ CharacterIdentifierStart, +/* 107 - k */ CharacterIdentifierStart, +/* 108 - l */ CharacterIdentifierStart, +/* 109 - m */ CharacterIdentifierStart, +/* 110 - n */ CharacterIdentifierStart, +/* 111 - o */ CharacterIdentifierStart, +/* 112 - p */ CharacterIdentifierStart, +/* 113 - q */ CharacterIdentifierStart, +/* 114 - r */ CharacterIdentifierStart, +/* 115 - s */ CharacterIdentifierStart, +/* 116 - t */ CharacterIdentifierStart, +/* 117 - u */ CharacterIdentifierStart, +/* 118 - v */ CharacterIdentifierStart, +/* 119 - w */ CharacterIdentifierStart, +/* 120 - x */ CharacterIdentifierStart, +/* 121 - y */ CharacterIdentifierStart, +/* 122 - z */ CharacterIdentifierStart, +/* 123 - { */ CharacterOpenBrace, +/* 124 - | */ CharacterOr, +/* 125 - } */ CharacterCloseBrace, +/* 126 - ~ */ CharacterTilde, +/* 127 - Delete */ CharacterInvalid, +}; + +template +Lexer::Lexer(JSGlobalData* globalData) + : m_isReparsing(false) + , m_globalData(globalData) +{ +} + +template +Lexer::~Lexer() +{ +} + +template +UString Lexer::getInvalidCharMessage() +{ + switch (m_current) { + case 0: + return "Invalid character: '\\0'"; + case 10: + return "Invalid character: '\\n'"; + case 11: + return "Invalid character: '\\v'"; + case 13: + return "Invalid character: '\\r'"; + case 35: + return "Invalid character: '#'"; + case 64: + return "Invalid character: '@'"; + case 96: + return "Invalid character: '`'"; + default: + return String::format("Invalid character '\\u%04u'", m_current).impl(); + } +} + +template +ALWAYS_INLINE const T* Lexer::currentCharacter() const +{ + ASSERT(m_code <= m_codeEnd); + return m_code; +} + +template +void Lexer::setCode(const SourceCode& source, ParserArena* arena) +{ + m_arena = &arena->identifierArena(); + + m_lineNumber = source.firstLine(); + m_delimited = false; + m_lastToken = -1; + + const StringImpl* sourceString = source.provider()->data(); + + if (sourceString) + setCodeStart(sourceString); + else + m_codeStart = 0; + + m_source = &source; + m_code = m_codeStart + source.startOffset(); + m_codeEnd = m_codeStart + source.endOffset(); + m_error = false; + m_atLineStart = true; + m_lexErrorMessage = UString(); + + m_buffer8.reserveInitialCapacity(initialReadBufferCapacity); + m_buffer16.reserveInitialCapacity((m_codeEnd - m_code) / 2); + + if (LIKELY(m_code < m_codeEnd)) + m_current = *m_code; + else + m_current = -1; + ASSERT(currentOffset() == source.startOffset()); +} + +template +template ALWAYS_INLINE void Lexer::internalShift() +{ + m_code += shiftAmount; + m_current = *m_code; +} + +template +ALWAYS_INLINE void Lexer::shift() +{ + // Faster than an if-else sequence + ASSERT(m_current != -1); + m_current = -1; + m_code++; + if (LIKELY(m_code < m_codeEnd)) + m_current = *m_code; +} + +template +ALWAYS_INLINE int Lexer::peek(int offset) +{ + // Only use if necessary + ASSERT(offset > 0 && offset < 5); + const T* code = m_code + offset; + return (code < m_codeEnd) ? *code : -1; +} + +template +int Lexer::getUnicodeCharacter() +{ + int char1 = peek(1); + int char2 = peek(2); + int char3 = peek(3); + + if (UNLIKELY(!isASCIIHexDigit(m_current) || !isASCIIHexDigit(char1) || !isASCIIHexDigit(char2) || !isASCIIHexDigit(char3))) + return -1; + + int result = convertUnicode(m_current, char1, char2, char3); + shift(); + shift(); + shift(); + shift(); + return result; +} + +template +void Lexer::shiftLineTerminator() +{ + ASSERT(isLineTerminator(m_current)); + + int m_prev = m_current; + shift(); + + // Allow both CRLF and LFCR. + if (m_prev + m_current == '\n' + '\r') + shift(); + + ++m_lineNumber; +} + +template +ALWAYS_INLINE bool Lexer::lastTokenWasRestrKeyword() const +{ + return m_lastToken == CONTINUE || m_lastToken == BREAK || m_lastToken == RETURN || m_lastToken == THROW; +} + +static NEVER_INLINE bool isNonASCIIIdentStart(int c) +{ + return category(c) & (Letter_Uppercase | Letter_Lowercase | Letter_Titlecase | Letter_Modifier | Letter_Other); +} + +static inline bool isIdentStart(int c) +{ + return isASCII(c) ? typesOfASCIICharacters[c] == CharacterIdentifierStart : isNonASCIIIdentStart(c); +} + +static NEVER_INLINE bool isNonASCIIIdentPart(int c) +{ + return category(c) & (Letter_Uppercase | Letter_Lowercase | Letter_Titlecase | Letter_Modifier | Letter_Other + | Mark_NonSpacing | Mark_SpacingCombining | Number_DecimalDigit | Punctuation_Connector); +} + +static ALWAYS_INLINE bool isIdentPart(int c) +{ + // Character types are divided into two groups depending on whether they can be part of an + // identifier or not. Those whose type value is less or equal than CharacterNumber can be + // part of an identifier. (See the CharacterType definition for more details.) + return isASCII(c) ? typesOfASCIICharacters[c] <= CharacterNumber : isNonASCIIIdentPart(c); +} + +static inline int singleEscape(int c) +{ + switch (c) { + case 'b': + return 0x08; + case 't': + return 0x09; + case 'n': + return 0x0A; + case 'v': + return 0x0B; + case 'f': + return 0x0C; + case 'r': + return 0x0D; + case '\\': + return '\\'; + case '\'': + return '\''; + case '"': + return '"'; + default: + return 0; + } +} + +template +inline void Lexer::record8(int c) +{ + ASSERT(c >= 0); + ASSERT(c <= 0xFF); + m_buffer8.append(static_cast(c)); +} + +template +inline void assertCharIsIn8BitRange(T c) +{ + UNUSED_PARAM(c); + ASSERT(c >= 0); + ASSERT(c <= 0xFF); +} + +template <> +inline void assertCharIsIn8BitRange(UChar c) +{ + UNUSED_PARAM(c); + ASSERT(c <= 0xFF); +} + +template <> +inline void assertCharIsIn8BitRange(LChar) +{ +} + +template +inline void Lexer::append8(const T* p, size_t length) +{ + size_t currentSize = m_buffer8.size(); + m_buffer8.grow(currentSize + length); + LChar* rawBuffer = m_buffer8.data() + currentSize; + + for (size_t i = 0; i < length; i++) { + T c = p[i]; + assertCharIsIn8BitRange(c); + rawBuffer[i] = c; + } +} + +template +inline void Lexer::append16(const LChar* p, size_t length) +{ + size_t currentSize = m_buffer16.size(); + m_buffer16.grow(currentSize + length); + UChar* rawBuffer = m_buffer16.data() + currentSize; + + for (size_t i = 0; i < length; i++) + rawBuffer[i] = p[i]; +} + +template +inline void Lexer::record16(T c) +{ + m_buffer16.append(c); +} + +template +inline void Lexer::record16(int c) +{ + ASSERT(c >= 0); + ASSERT(c <= USHRT_MAX); + m_buffer16.append(static_cast(c)); +} + +template <> + template ALWAYS_INLINE JSTokenType Lexer::parseIdentifier(JSTokenData* tokenData, unsigned lexerFlags, bool strictMode) +{ + const ptrdiff_t remaining = m_codeEnd - m_code; + if ((remaining >= maxTokenLength) && !(lexerFlags & LexerFlagsIgnoreReservedWords)) { + JSTokenType keyword = parseKeyword(tokenData); + if (keyword != IDENT) { + ASSERT((!shouldCreateIdentifier) || tokenData->ident); + return keyword == RESERVED_IF_STRICT && !strictMode ? IDENT : keyword; + } + } + + const LChar* identifierStart = currentCharacter(); + + while (isIdentPart(m_current)) + shift(); + + if (UNLIKELY(m_current == '\\')) { + setOffsetFromCharOffset(identifierStart); + return parseIdentifierSlowCase(tokenData, lexerFlags, strictMode); + } + + const Identifier* ident = 0; + + if (shouldCreateIdentifier) { + int identifierLength = currentCharacter() - identifierStart; + ident = makeIdentifier(identifierStart, identifierLength); + + tokenData->ident = ident; + } else + tokenData->ident = 0; + + m_delimited = false; + + if (UNLIKELY((remaining < maxTokenLength) && !(lexerFlags & LexerFlagsIgnoreReservedWords))) { + ASSERT(shouldCreateIdentifier); + if (remaining < maxTokenLength) { + const HashEntry* entry = m_globalData->keywords->getKeyword(*ident); + ASSERT((remaining < maxTokenLength) || !entry); + if (!entry) + return IDENT; + JSTokenType token = static_cast(entry->lexerValue()); + return (token != RESERVED_IF_STRICT) || strictMode ? token : IDENT; + } + return IDENT; + } + + return IDENT; +} + +template <> +template ALWAYS_INLINE JSTokenType Lexer::parseIdentifier(JSTokenData* tokenData, unsigned lexerFlags, bool strictMode) +{ + const ptrdiff_t remaining = m_codeEnd - m_code; + if ((remaining >= maxTokenLength) && !(lexerFlags & LexerFlagsIgnoreReservedWords)) { + JSTokenType keyword = parseKeyword(tokenData); + if (keyword != IDENT) { + ASSERT((!shouldCreateIdentifier) || tokenData->ident); + return keyword == RESERVED_IF_STRICT && !strictMode ? IDENT : keyword; + } + } + const UChar* identifierStart = currentCharacter(); + + UChar orAllChars = 0; + + while (isIdentPart(m_current)) { + orAllChars |= m_current; + shift(); + } + + if (UNLIKELY(m_current == '\\')) { + setOffsetFromCharOffset(identifierStart); + return parseIdentifierSlowCase(tokenData, lexerFlags, strictMode); + } + + bool isAll8Bit = false; + + if (!(orAllChars & ~0xff)) + isAll8Bit = true; + + const Identifier* ident = 0; + + if (shouldCreateIdentifier) { + int identifierLength = currentCharacter() - identifierStart; + if (isAll8Bit) + ident = makeIdentifierLCharFromUChar(identifierStart, identifierLength); + else + ident = makeIdentifier(identifierStart, identifierLength); + + tokenData->ident = ident; + } else + tokenData->ident = 0; + + m_delimited = false; + + if (UNLIKELY((remaining < maxTokenLength) && !(lexerFlags & LexerFlagsIgnoreReservedWords))) { + ASSERT(shouldCreateIdentifier); + if (remaining < maxTokenLength) { + const HashEntry* entry = m_globalData->keywords->getKeyword(*ident); + ASSERT((remaining < maxTokenLength) || !entry); + if (!entry) + return IDENT; + JSTokenType token = static_cast(entry->lexerValue()); + return (token != RESERVED_IF_STRICT) || strictMode ? token : IDENT; + } + return IDENT; + } + + return IDENT; +} + +template +template JSTokenType Lexer::parseIdentifierSlowCase(JSTokenData* tokenData, unsigned lexerFlags, bool strictMode) +{ + const ptrdiff_t remaining = m_codeEnd - m_code; + const T* identifierStart = currentCharacter(); + bool bufferRequired = false; + + while (true) { + if (LIKELY(isIdentPart(m_current))) { + shift(); + continue; + } + if (LIKELY(m_current != '\\')) + break; + + // \uXXXX unicode characters. + bufferRequired = true; + if (identifierStart != currentCharacter()) + m_buffer16.append(identifierStart, currentCharacter() - identifierStart); + shift(); + if (UNLIKELY(m_current != 'u')) + return ERRORTOK; + shift(); + int character = getUnicodeCharacter(); + if (UNLIKELY(character == -1)) + return ERRORTOK; + if (UNLIKELY(m_buffer16.size() ? !isIdentPart(character) : !isIdentStart(character))) + return ERRORTOK; + if (shouldCreateIdentifier) + record16(character); + identifierStart = currentCharacter(); + } + + int identifierLength; + const Identifier* ident = 0; + if (shouldCreateIdentifier) { + if (!bufferRequired) { + identifierLength = currentCharacter() - identifierStart; + ident = makeIdentifier(identifierStart, identifierLength); + } else { + if (identifierStart != currentCharacter()) + m_buffer16.append(identifierStart, currentCharacter() - identifierStart); + ident = makeIdentifier(m_buffer16.data(), m_buffer16.size()); + } + + tokenData->ident = ident; + } else + tokenData->ident = 0; + + m_delimited = false; + + if (LIKELY(!bufferRequired && !(lexerFlags & LexerFlagsIgnoreReservedWords))) { + ASSERT(shouldCreateIdentifier); + // Keywords must not be recognized if there was an \uXXXX in the identifier. + if (remaining < maxTokenLength) { + const HashEntry* entry = m_globalData->keywords->getKeyword(*ident); + ASSERT((remaining < maxTokenLength) || !entry); + if (!entry) + return IDENT; + JSTokenType token = static_cast(entry->lexerValue()); + return (token != RESERVED_IF_STRICT) || strictMode ? token : IDENT; + } + return IDENT; + } + + m_buffer16.resize(0); + return IDENT; +} + +template +template ALWAYS_INLINE bool Lexer::parseString(JSTokenData* tokenData, bool strictMode) +{ + int startingOffset = currentOffset(); + int startingLineNumber = lineNumber(); + int stringQuoteCharacter = m_current; + shift(); + + const T* stringStart = currentCharacter(); + + while (m_current != stringQuoteCharacter) { + if (UNLIKELY((m_current == '\\'))) { + if (stringStart != currentCharacter() && shouldBuildStrings) + append8(stringStart, currentCharacter() - stringStart); + shift(); + + int escape = singleEscape(m_current); + + // Most common escape sequences first + if (escape) { + if (shouldBuildStrings) + record8(escape); + shift(); + } else if (UNLIKELY(isLineTerminator(m_current))) + shiftLineTerminator(); + else if (m_current == 'x') { + shift(); + if (isASCIIHexDigit(m_current) && isASCIIHexDigit(peek(1))) { + int prev = m_current; + shift(); + if (shouldBuildStrings) + record8(convertHex(prev, m_current)); + shift(); + } else if (shouldBuildStrings) + record8('x'); + } else { + setOffset(startingOffset); + setLineNumber(startingLineNumber); + m_buffer8.resize(0); + return parseStringSlowCase(tokenData, strictMode); + } + stringStart = currentCharacter(); + continue; + } + + if (UNLIKELY(((m_current > 0xff) || (m_current < 0xe)))) { + setOffset(startingOffset); + setLineNumber(startingLineNumber); + m_buffer8.resize(0); + return parseStringSlowCase(tokenData, strictMode); + } + + shift(); + } + + if (currentCharacter() != stringStart && shouldBuildStrings) + append8(stringStart, currentCharacter() - stringStart); + if (shouldBuildStrings) { + tokenData->ident = makeIdentifier(m_buffer8.data(), m_buffer8.size()); + m_buffer8.resize(0); + } else + tokenData->ident = 0; + + return true; +} + +template +template bool Lexer::parseStringSlowCase(JSTokenData* tokenData, bool strictMode) +{ + int stringQuoteCharacter = m_current; + shift(); + + const T* stringStart = currentCharacter(); + + while (m_current != stringQuoteCharacter) { + if (UNLIKELY(m_current == '\\')) { + if (stringStart != currentCharacter() && shouldBuildStrings) + append16(stringStart, currentCharacter() - stringStart); + shift(); + + int escape = singleEscape(m_current); + + // Most common escape sequences first + if (escape) { + if (shouldBuildStrings) + record16(escape); + shift(); + } else if (UNLIKELY(isLineTerminator(m_current))) + shiftLineTerminator(); + else if (m_current == 'x') { + shift(); + if (isASCIIHexDigit(m_current) && isASCIIHexDigit(peek(1))) { + int prev = m_current; + shift(); + if (shouldBuildStrings) + record16(convertHex(prev, m_current)); + shift(); + } else if (shouldBuildStrings) + record16('x'); + } else if (m_current == 'u') { + shift(); + int character = getUnicodeCharacter(); + if (character != -1) { + if (shouldBuildStrings) + record16(character); + } else if (m_current == stringQuoteCharacter) { + if (shouldBuildStrings) + record16('u'); + } else { + m_lexErrorMessage = "\\u can only be followed by a Unicode character sequence"; + return false; + } + } else if (strictMode && isASCIIDigit(m_current)) { + // The only valid numeric escape in strict mode is '\0', and this must not be followed by a decimal digit. + int character1 = m_current; + shift(); + if (character1 != '0' || isASCIIDigit(m_current)) { + m_lexErrorMessage = "The only valid numeric escape in strict mode is '\\0'"; + return false; + } + if (shouldBuildStrings) + record16(0); + } else if (!strictMode && isASCIIOctalDigit(m_current)) { + // Octal character sequences + int character1 = m_current; + shift(); + if (isASCIIOctalDigit(m_current)) { + // Two octal characters + int character2 = m_current; + shift(); + if (character1 >= '0' && character1 <= '3' && isASCIIOctalDigit(m_current)) { + if (shouldBuildStrings) + record16((character1 - '0') * 64 + (character2 - '0') * 8 + m_current - '0'); + shift(); + } else { + if (shouldBuildStrings) + record16((character1 - '0') * 8 + character2 - '0'); + } + } else { + if (shouldBuildStrings) + record16(character1 - '0'); + } + } else if (m_current != -1) { + if (shouldBuildStrings) + record16(m_current); + shift(); + } else { + m_lexErrorMessage = "Unterminated string constant"; + return false; + } + + stringStart = currentCharacter(); + continue; + } + // Fast check for characters that require special handling. + // Catches -1, \n, \r, 0x2028, and 0x2029 as efficiently + // as possible, and lets through all common ASCII characters. + if (UNLIKELY(((static_cast(m_current) - 0xE) & 0x2000))) { + // New-line or end of input is not allowed + if (UNLIKELY(isLineTerminator(m_current)) || UNLIKELY(m_current == -1)) { + m_lexErrorMessage = "Unexpected EOF"; + return false; + } + // Anything else is just a normal character + } + shift(); + } + + if (currentCharacter() != stringStart && shouldBuildStrings) + append16(stringStart, currentCharacter() - stringStart); + if (shouldBuildStrings) + tokenData->ident = makeIdentifier(m_buffer16.data(), m_buffer16.size()); + else + tokenData->ident = 0; + + m_buffer16.resize(0); + return true; +} + +template +ALWAYS_INLINE void Lexer::parseHex(double& returnValue) +{ + // Optimization: most hexadecimal values fit into 4 bytes. + uint32_t hexValue = 0; + int maximumDigits = 7; + + // Shift out the 'x' prefix. + shift(); + + do { + hexValue = (hexValue << 4) + toASCIIHexValue(m_current); + shift(); + --maximumDigits; + } while (isASCIIHexDigit(m_current) && maximumDigits >= 0); + + if (maximumDigits >= 0) { + returnValue = hexValue; + return; + } + + // No more place in the hexValue buffer. + // The values are shifted out and placed into the m_buffer8 vector. + for (int i = 0; i < 8; ++i) { + int digit = hexValue >> 28; + if (digit < 10) + record8(digit + '0'); + else + record8(digit - 10 + 'a'); + hexValue <<= 4; + } + + while (isASCIIHexDigit(m_current)) { + record8(m_current); + shift(); + } + + returnValue = parseIntOverflow(m_buffer8.data(), m_buffer8.size(), 16); +} + +template +ALWAYS_INLINE bool Lexer::parseOctal(double& returnValue) +{ + // Optimization: most octal values fit into 4 bytes. + uint32_t octalValue = 0; + int maximumDigits = 9; + // Temporary buffer for the digits. Makes easier + // to reconstruct the input characters when needed. + LChar digits[10]; + + do { + octalValue = octalValue * 8 + (m_current - '0'); + digits[maximumDigits] = m_current; + shift(); + --maximumDigits; + } while (isASCIIOctalDigit(m_current) && maximumDigits >= 0); + + if (!isASCIIDigit(m_current) && maximumDigits >= 0) { + returnValue = octalValue; + return true; + } + + for (int i = 9; i > maximumDigits; --i) + record8(digits[i]); + + while (isASCIIOctalDigit(m_current)) { + record8(m_current); + shift(); + } + + if (isASCIIDigit(m_current)) + return false; + + returnValue = parseIntOverflow(m_buffer8.data(), m_buffer8.size(), 8); + return true; +} + +template +ALWAYS_INLINE bool Lexer::parseDecimal(double& returnValue) +{ + // Optimization: most decimal values fit into 4 bytes. + uint32_t decimalValue = 0; + + // Since parseOctal may be executed before parseDecimal, + // the m_buffer8 may hold ascii digits. + if (!m_buffer8.size()) { + int maximumDigits = 9; + // Temporary buffer for the digits. Makes easier + // to reconstruct the input characters when needed. + LChar digits[10]; + + do { + decimalValue = decimalValue * 10 + (m_current - '0'); + digits[maximumDigits] = m_current; + shift(); + --maximumDigits; + } while (isASCIIDigit(m_current) && maximumDigits >= 0); + + if (maximumDigits >= 0 && m_current != '.' && (m_current | 0x20) != 'e') { + returnValue = decimalValue; + return true; + } + + for (int i = 9; i > maximumDigits; --i) + record8(digits[i]); + } + + while (isASCIIDigit(m_current)) { + record8(m_current); + shift(); + } + + return false; +} + +template +ALWAYS_INLINE void Lexer::parseNumberAfterDecimalPoint() +{ + record8('.'); + while (isASCIIDigit(m_current)) { + record8(m_current); + shift(); + } +} + +template +ALWAYS_INLINE bool Lexer::parseNumberAfterExponentIndicator() +{ + record8('e'); + shift(); + if (m_current == '+' || m_current == '-') { + record8(m_current); + shift(); + } + + if (!isASCIIDigit(m_current)) + return false; + + do { + record8(m_current); + shift(); + } while (isASCIIDigit(m_current)); + return true; +} + +template +ALWAYS_INLINE bool Lexer::parseMultilineComment() +{ + while (true) { + while (UNLIKELY(m_current == '*')) { + shift(); + if (m_current == '/') { + shift(); + return true; + } + } + + if (UNLIKELY(m_current == -1)) + return false; + + if (isLineTerminator(m_current)) { + shiftLineTerminator(); + m_terminator = true; + } else + shift(); + } +} + +template +bool Lexer::nextTokenIsColon() +{ + const T* code = m_code; + while (code < m_codeEnd && (isWhiteSpace(*code) || isLineTerminator(*code))) + code++; + + return code < m_codeEnd && *code == ':'; +} + +template +JSTokenType Lexer::lex(JSTokenData* tokenData, JSTokenInfo* tokenInfo, unsigned lexerFlags, bool strictMode) +{ + ASSERT(!m_error); + ASSERT(m_buffer8.isEmpty()); + ASSERT(m_buffer16.isEmpty()); + + JSTokenType token = ERRORTOK; + m_terminator = false; + +start: + while (isWhiteSpace(m_current)) + shift(); + + int startOffset = currentOffset(); + + if (UNLIKELY(m_current == -1)) + return EOFTOK; + + m_delimited = false; + + CharacterType type; + if (LIKELY(isASCII(m_current))) + type = static_cast(typesOfASCIICharacters[m_current]); + else if (isNonASCIIIdentStart(m_current)) + type = CharacterIdentifierStart; + else if (isLineTerminator(m_current)) + type = CharacterLineTerminator; + else + type = CharacterInvalid; + + switch (type) { + case CharacterGreater: + shift(); + if (m_current == '>') { + shift(); + if (m_current == '>') { + shift(); + if (m_current == '=') { + shift(); + token = URSHIFTEQUAL; + break; + } + token = URSHIFT; + break; + } + if (m_current == '=') { + shift(); + token = RSHIFTEQUAL; + break; + } + token = RSHIFT; + break; + } + if (m_current == '=') { + shift(); + token = GE; + break; + } + token = GT; + break; + case CharacterEqual: + shift(); + if (m_current == '=') { + shift(); + if (m_current == '=') { + shift(); + token = STREQ; + break; + } + token = EQEQ; + break; + } + token = EQUAL; + break; + case CharacterLess: + shift(); + if (m_current == '!' && peek(1) == '-' && peek(2) == '-') { + //