char-predicates-inl.h 4.16 KB
Newer Older
1
// Copyright 2011 the V8 project authors. All rights reserved.
2 3
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
4

5 6
#ifndef V8_STRINGS_CHAR_PREDICATES_INL_H_
#define V8_STRINGS_CHAR_PREDICATES_INL_H_
7

8
#include "src/strings/char-predicates.h"
9

10 11
namespace v8 {
namespace internal {
12

13 14 15
// If c is in 'A'-'Z' or 'a'-'z', return its lower-case.
// Else, return something outside of 'A'-'Z' and 'a'-'z'.
// Note: it ignores LOCALE.
16
inline constexpr int AsciiAlphaToLower(uc32 c) { return c | 0x20; }
17

18
inline constexpr bool IsCarriageReturn(uc32 c) { return c == 0x000D; }
19

20
inline constexpr bool IsLineFeed(uc32 c) { return c == 0x000A; }
21

22
inline constexpr bool IsAsciiIdentifier(uc32 c) {
23 24 25
  return IsAlphaNumeric(c) || c == '$' || c == '_';
}

26
inline constexpr bool IsAlphaNumeric(uc32 c) {
27 28
  return IsInRange(AsciiAlphaToLower(c), 'a', 'z') || IsDecimalDigit(c);
}
29

30
inline constexpr bool IsDecimalDigit(uc32 c) {
31
  // ECMA-262, 3rd, 7.8.3 (p 16)
32
  return IsInRange(c, '0', '9');
33 34
}

35
inline constexpr bool IsHexDigit(uc32 c) {
36
  // ECMA-262, 3rd, 7.6 (p 15)
37
  return IsDecimalDigit(c) || IsInRange(AsciiAlphaToLower(c), 'a', 'f');
38 39
}

40
inline constexpr bool IsOctalDigit(uc32 c) {
41 42 43 44
  // ECMA-262, 6th, 7.8.3
  return IsInRange(c, '0', '7');
}

45 46 47
inline constexpr bool IsNonOctalDecimalDigit(uc32 c) {
  return IsInRange(c, '8', '9');
}
48

49
inline constexpr bool IsBinaryDigit(uc32 c) {
50 51 52 53
  // ECMA-262, 6th, 7.8.3
  return c == '0' || c == '1';
}

54 55 56 57 58 59 60 61 62 63 64 65
inline constexpr bool IsAsciiLower(uc32 c) { return IsInRange(c, 'a', 'z'); }

inline constexpr bool IsAsciiUpper(uc32 c) { return IsInRange(c, 'A', 'Z'); }

inline constexpr uc32 ToAsciiUpper(uc32 c) {
  return c & ~(IsAsciiLower(c) << 5);
}

inline constexpr uc32 ToAsciiLower(uc32 c) {
  return c | (IsAsciiUpper(c) << 5);
}

66
inline constexpr bool IsRegExpWord(uc16 c) {
67 68
  return IsInRange(AsciiAlphaToLower(c), 'a', 'z') || IsDecimalDigit(c) ||
         (c == '_');
69 70
}

71 72 73
inline constexpr bool IsRegExpNewline(uc16 c) {
  //          CR             LF             LS             PS
  return c != 0x000A && c != 0x000D && c != 0x2028 && c != 0x2029;
74 75
}

76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132
// Constexpr cache table for character flags.
enum AsciiCharFlags {
  kIsIdentifierStart = 1 << 0,
  kIsIdentifierPart = 1 << 1,
  kIsWhiteSpace = 1 << 2,
  kIsWhiteSpaceOrLineTerminator = 1 << 3
};
constexpr uint8_t BuildAsciiCharFlags(uc32 c) {
  // clang-format off
  return
    (IsAsciiIdentifier(c) || c == '\\') ? (
      kIsIdentifierPart | (!IsDecimalDigit(c) ? kIsIdentifierStart : 0)) : 0 |
    (c == ' ' || c == '\t' || c == '\v' || c == '\f') ?
      kIsWhiteSpace | kIsWhiteSpaceOrLineTerminator : 0 |
    (c == '\r' || c == '\n') ? kIsWhiteSpaceOrLineTerminator : 0;
  // clang-format on
}
const constexpr uint8_t kAsciiCharFlags[128] = {
#define BUILD_CHAR_FLAGS(N) BuildAsciiCharFlags(N),
    INT_0_TO_127_LIST(BUILD_CHAR_FLAGS)
#undef BUILD_CHAR_FLAGS
};

bool IsIdentifierStart(uc32 c) {
  if (!IsInRange(c, 0, 127)) return IsIdentifierStartSlow(c);
  DCHECK_EQ(IsIdentifierStartSlow(c),
            static_cast<bool>(kAsciiCharFlags[c] & kIsIdentifierStart));
  return kAsciiCharFlags[c] & kIsIdentifierStart;
}

bool IsIdentifierPart(uc32 c) {
  if (!IsInRange(c, 0, 127)) return IsIdentifierPartSlow(c);
  DCHECK_EQ(IsIdentifierPartSlow(c),
            static_cast<bool>(kAsciiCharFlags[c] & kIsIdentifierPart));
  return kAsciiCharFlags[c] & kIsIdentifierPart;
}

bool IsWhiteSpace(uc32 c) {
  if (!IsInRange(c, 0, 127)) return IsWhiteSpaceSlow(c);
  DCHECK_EQ(IsWhiteSpaceSlow(c),
            static_cast<bool>(kAsciiCharFlags[c] & kIsWhiteSpace));
  return kAsciiCharFlags[c] & kIsWhiteSpace;
}

bool IsWhiteSpaceOrLineTerminator(uc32 c) {
  if (!IsInRange(c, 0, 127)) return IsWhiteSpaceOrLineTerminatorSlow(c);
  DCHECK_EQ(
      IsWhiteSpaceOrLineTerminatorSlow(c),
      static_cast<bool>(kAsciiCharFlags[c] & kIsWhiteSpaceOrLineTerminator));
  return kAsciiCharFlags[c] & kIsWhiteSpaceOrLineTerminator;
}

bool IsLineTerminatorSequence(uc32 c, uc32 next) {
  if (!unibrow::IsLineTerminator(c)) return false;
  if (c == 0x000d && next == 0x000a) return false;  // CR with following LF.
  return true;
}
133

134 135
}  // namespace internal
}  // namespace v8
136

137
#endif  // V8_STRINGS_CHAR_PREDICATES_INL_H_