src/objects.h - Issue 11759008: Introduce ENABLE_LATIN_1 compile flag

Unified Diff: src/objects.h

Issue 11759008: Introduce ENABLE_LATIN_1 compile flag (Closed) Base URL: https://v8.googlecode.com/svn/branches/bleeding_edge

Patch Set: Fix FilterASCII Created 7 years, 11 months ago

Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.

Jump to:

View side-by-side diff with in-line comments

Download patch

Index: src/objects.h

diff --git a/src/objects.h b/src/objects.h

index f1fa27accdef209ed0e0653f6bd20b331a3b24e9..ad921251fa2f224028a92d43eaf9d4e4e7bacce7 100644

--- a/src/objects.h

+++ b/src/objects.h

@@ -532,9 +532,11 @@ enum InstanceType {

SYMBOL_TYPE = kTwoByteStringTag | kSymbolTag | kSeqStringTag,

ASCII_SYMBOL_TYPE = kOneByteStringTag | kAsciiDataHintTag | kSymbolTag |

kSeqStringTag,

+ ONE_BYTE_SYMBOL_TYPE = kOneByteStringTag | kSymbolTag | kSeqStringTag,

CONS_SYMBOL_TYPE = kTwoByteStringTag | kSymbolTag | kConsStringTag,

CONS_ASCII_SYMBOL_TYPE = kOneByteStringTag | kAsciiDataHintTag | kSymbolTag |

kConsStringTag,

+ CONS_ONE_BYTE_SYMBOL_TYPE = kOneByteStringTag | kSymbolTag | kConsStringTag,

SHORT_EXTERNAL_SYMBOL_TYPE = kTwoByteStringTag | kSymbolTag |

kExternalStringTag | kShortExternalStringTag,

SHORT_EXTERNAL_SYMBOL_WITH_ASCII_DATA_TYPE =

@@ -543,19 +545,27 @@ enum InstanceType {

SHORT_EXTERNAL_ASCII_SYMBOL_TYPE = kOneByteStringTag | kAsciiDataHintTag |

kExternalStringTag | kSymbolTag |

kShortExternalStringTag,

+ SHORT_EXTERNAL_ONE_BYTE_SYMBOL_TYPE = kOneByteStringTag |

+ kExternalStringTag | kSymbolTag |

+ kShortExternalStringTag,

EXTERNAL_SYMBOL_TYPE = kTwoByteStringTag | kSymbolTag | kExternalStringTag,

EXTERNAL_SYMBOL_WITH_ASCII_DATA_TYPE =

kTwoByteStringTag | kSymbolTag | kExternalStringTag | kAsciiDataHintTag,

EXTERNAL_ASCII_SYMBOL_TYPE =

kOneByteStringTag | kAsciiDataHintTag | kSymbolTag | kExternalStringTag,

+ EXTERNAL_ONE_BYTE_SYMBOL_TYPE =

+ kOneByteStringTag | kSymbolTag | kExternalStringTag,

STRING_TYPE = kTwoByteStringTag | kSeqStringTag,

ASCII_STRING_TYPE = kOneByteStringTag | kAsciiDataHintTag | kSeqStringTag,

+ ONE_BYTE_STRING_TYPE = kOneByteStringTag | kSeqStringTag,

CONS_STRING_TYPE = kTwoByteStringTag | kConsStringTag,

CONS_ASCII_STRING_TYPE =

kOneByteStringTag | kAsciiDataHintTag | kConsStringTag,

+ CONS_ONE_BYTE_STRING_TYPE = kOneByteStringTag | kConsStringTag,

SLICED_STRING_TYPE = kTwoByteStringTag | kSlicedStringTag,

SLICED_ASCII_STRING_TYPE =

kOneByteStringTag | kAsciiDataHintTag | kSlicedStringTag,

+ SLICED_ONE_BYTE_STRING_TYPE = kOneByteStringTag | kSlicedStringTag,

SHORT_EXTERNAL_STRING_TYPE =

kTwoByteStringTag | kExternalStringTag | kShortExternalStringTag,

SHORT_EXTERNAL_STRING_WITH_ASCII_DATA_TYPE =

@@ -564,12 +574,15 @@ enum InstanceType {

SHORT_EXTERNAL_ASCII_STRING_TYPE =

kOneByteStringTag | kAsciiDataHintTag |

kExternalStringTag | kShortExternalStringTag,

+ SHORT_EXTERNAL_ONE_BYTE_STRING_TYPE =

+ kOneByteStringTag | kExternalStringTag | kShortExternalStringTag,

EXTERNAL_STRING_TYPE = kTwoByteStringTag | kExternalStringTag,

EXTERNAL_STRING_WITH_ASCII_DATA_TYPE =

kTwoByteStringTag | kExternalStringTag | kAsciiDataHintTag,

- // LAST_STRING_TYPE

EXTERNAL_ASCII_STRING_TYPE =

kOneByteStringTag | kAsciiDataHintTag | kExternalStringTag,

+ // LAST_STRING_TYPE

+ EXTERNAL_ONE_BYTE_STRING_TYPE = kOneByteStringTag | kExternalStringTag,

PRIVATE_EXTERNAL_ASCII_STRING_TYPE = EXTERNAL_ASCII_STRING_TYPE,

// Objects allocated in their own spaces (never in new space).

@@ -3028,7 +3041,7 @@ class SymbolTable: public HashTable<SymbolTableShape, HashTableKey*> {

// pointer *s is set to the symbol found.

MUST_USE_RESULT MaybeObject* LookupUtf8Symbol(Vector<const char> str,

Object** s);

- MUST_USE_RESULT MaybeObject* LookupOneByteSymbol(Vector<const char> str,

+ MUST_USE_RESULT MaybeObject* LookupOneByteSymbol(Vector<const uint8_t> str,

Object** s);

MUST_USE_RESULT MaybeObject* LookupSubStringOneByteSymbol(

Handle<SeqOneByteString> str,

@@ -7057,12 +7070,19 @@ class String: public HeapObject {

// Returns true if the structure contains two-byte content.

bool IsTwoByte() { return state_ == TWO_BYTE; }

+ // TODO(dcarney): Remove this function.

// Return the ASCII content of the string. Only use if IsAscii() returns

// true.

Vector<const char> ToAsciiVector() {

ASSERT_EQ(ASCII, state_);

return Vector<const char>::cast(buffer_);

}

+ // Return the one byte content of the string. Only use if IsAscii() returns

+ // true.

+ Vector<const uint8_t> ToOneByteVector() {

+ ASSERT_EQ(ASCII, state_);

+ return buffer_;

+ }

// Return the two-byte content of the string. Only use if IsTwoByte()

// returns true.

Vector<const uc16> ToUC16Vector() {

@@ -7113,6 +7133,8 @@ class String: public HeapObject {

// possible.

inline bool HasOnlyAsciiChars();

+ inline bool IsOneByteConvertible();

// Get and set individual two byte chars in the string.

inline void Set(int index, uint16_t value);

// Get individual two byte char in the string. Repeated calls

@@ -7163,8 +7185,8 @@ class String: public HeapObject {

// String equality operations.

inline bool Equals(String* other);

- bool IsEqualTo(Vector<const char> str);

- bool IsAsciiEqualTo(Vector<const char> str);

+ bool IsUtf8EqualTo(Vector<const char> str);

+ bool IsOneByteEqualTo(Vector<const uint8_t> str);

bool IsTwoByteEqualTo(Vector<const uc16> str);

// Return a UTF8 representation of the string. The string is null

@@ -7240,9 +7262,9 @@ class String: public HeapObject {

// value into an array index.

static const int kMaxArrayIndexSize = 10;

- // Max ASCII char code.

- static const int kMaxAsciiCharCode = unibrow::Utf8::kMaxOneByteChar;

- static const unsigned kMaxAsciiCharCodeU = unibrow::Utf8::kMaxOneByteChar;

+ // Max char codes.

+ static const int32_t kMaxOneByteCharCode = unibrow::Latin1::kMaxChar;

+ static const uint32_t kMaxOneByteCharCodeU = unibrow::Latin1::kMaxChar;

static const int kMaxUtf16CodeUnit = 0xffff;

// Mask constant for checking if a string has a computed hash code

@@ -7327,7 +7349,7 @@ class String: public HeapObject {

const char* start = chars;

const char* limit = chars + length;

#ifdef V8_HOST_CAN_READ_UNALIGNED

- ASSERT(kMaxAsciiCharCode == 0x7F);

+ ASSERT(unibrow::Utf8::kMaxOneByteChar == 0x7F);

const uintptr_t non_ascii_mask = kUintptrAllBitsSet / 0xFF * 0x80;

while (chars + sizeof(uintptr_t) <= limit) {

if (*reinterpret_cast<const uintptr_t*>(chars) & non_ascii_mask) {

@@ -7337,7 +7359,7 @@ class String: public HeapObject {

}

#endif

while (chars < limit) {

- if (static_cast<uint8_t>(*chars) > kMaxAsciiCharCodeU) {

+ if (static_cast<uint8_t>(*chars) > unibrow::Utf8::kMaxOneByteChar) {

return static_cast<int>(chars - start);

}

++chars;

@@ -7349,18 +7371,18 @@ class String: public HeapObject {

return NonAsciiStart(chars, length) >= length;

}

- static inline int NonAsciiStart(const uc16* chars, int length) {

+ static inline int NonOneByteStart(const uc16* chars, int length) {

const uc16* limit = chars + length;

const uc16* start = chars;

while (chars < limit) {

- if (*chars > kMaxAsciiCharCodeU) return static_cast<int>(chars - start);

+ if (*chars > kMaxOneByteCharCodeU) return static_cast<int>(chars - start);

++chars;

}

return static_cast<int>(chars - start);

}

- static inline bool IsAscii(const uc16* chars, int length) {

- return NonAsciiStart(chars, length) >= length;

+ static inline bool IsOneByte(const uc16* chars, int length) {

+ return NonOneByteStart(chars, length) >= length;

}

template<class Visitor, class ConsOp>

@@ -7425,7 +7447,9 @@ class SeqOneByteString: public SeqString {

// Get the address of the characters in this string.

inline Address GetCharsAddress();

+ // TODO(dcarney): remove GetChars and rename GetCharsU to GetChars.

inline char* GetChars();

+ inline uint8_t* GetCharsU();

// Casting

static inline SeqOneByteString* cast(Object* obj);

« no previous file with comments | « src/mips/regexp-macro-assembler-mips.cc ('k') | src/objects.cc » ('j') | no next file with comments »