| 1 | /* utf8.h |
| 2 | * |
| 3 | * Copyright (c) 1998-2002, Larry Wall |
| 4 | * |
| 5 | * You may distribute under the terms of either the GNU General Public |
| 6 | * License or the Artistic License, as specified in the README file. |
| 7 | * |
| 8 | */ |
| 9 | |
| 10 | /* Use UTF-8 as the default script encoding? |
| 11 | * Turning this on will break scripts having non-UTF8 binary |
| 12 | * data (such as Latin-1) in string literals. */ |
| 13 | #ifdef USE_UTF8_SCRIPTS |
| 14 | # define USE_UTF8_IN_NAMES (!IN_BYTES) |
| 15 | #else |
| 16 | # define USE_UTF8_IN_NAMES (PL_hints & HINT_UTF8) |
| 17 | #endif |
| 18 | |
| 19 | #ifdef EBCDIC |
| 20 | /* The equivalent of these macros but implementing UTF-EBCDIC |
| 21 | are in the following header file: |
| 22 | */ |
| 23 | |
| 24 | #include "utfebcdic.h" |
| 25 | #else |
| 26 | START_EXTERN_C |
| 27 | |
| 28 | #ifdef DOINIT |
| 29 | EXTCONST unsigned char PL_utf8skip[] = { |
| 30 | 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */ |
| 31 | 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */ |
| 32 | 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */ |
| 33 | 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */ |
| 34 | 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* bogus */ |
| 35 | 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* bogus */ |
| 36 | 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, /* scripts */ |
| 37 | 3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,4,4,4,4,4,4,4,4,5,5,5,5,6,6, /* cjk etc. */ |
| 38 | 7,13, /* Perl extended (not UTF-8). Up to 72bit allowed (64-bit + reserved). */ |
| 39 | }; |
| 40 | #else |
| 41 | EXTCONST unsigned char PL_utf8skip[]; |
| 42 | #endif |
| 43 | |
| 44 | END_EXTERN_C |
| 45 | #define UTF8SKIP(s) PL_utf8skip[*(U8*)s] |
| 46 | |
| 47 | /* Native character to iso-8859-1 */ |
| 48 | #define NATIVE_TO_ASCII(ch) (ch) |
| 49 | #define ASCII_TO_NATIVE(ch) (ch) |
| 50 | /* Transform after encoding */ |
| 51 | #define NATIVE_TO_UTF(ch) (ch) |
| 52 | #define UTF_TO_NATIVE(ch) (ch) |
| 53 | /* Transforms in wide UV chars */ |
| 54 | #define UNI_TO_NATIVE(ch) (ch) |
| 55 | #define NATIVE_TO_UNI(ch) (ch) |
| 56 | /* Transforms in invariant space */ |
| 57 | #define NATIVE_TO_NEED(enc,ch) (ch) |
| 58 | #define ASCII_TO_NEED(enc,ch) (ch) |
| 59 | |
| 60 | /* As there are no translations avoid the function wrapper */ |
| 61 | #define Perl_utf8n_to_uvchr Perl_utf8n_to_uvuni |
| 62 | #define Perl_uvchr_to_utf8 Perl_uvuni_to_utf8 |
| 63 | |
| 64 | /* |
| 65 | |
| 66 | The following table is from Unicode 3.2. |
| 67 | |
| 68 | Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte |
| 69 | |
| 70 | U+0000..U+007F 00..7F |
| 71 | U+0080..U+07FF C2..DF 80..BF |
| 72 | U+0800..U+0FFF E0 A0..BF 80..BF |
| 73 | U+1000..U+CFFF E1..EC 80..BF 80..BF |
| 74 | U+D000..U+D7FF ED 80..9F 80..BF |
| 75 | U+D800..U+DFFF ******* ill-formed ******* |
| 76 | U+E000..U+FFFF EE..EF 80..BF 80..BF |
| 77 | U+10000..U+3FFFF F0 90..BF 80..BF 80..BF |
| 78 | U+40000..U+FFFFF F1..F3 80..BF 80..BF 80..BF |
| 79 | U+100000..U+10FFFF F4 80..8F 80..BF 80..BF |
| 80 | |
| 81 | Note the A0..BF in U+0800..U+0FFF, the 80..9F in U+D000...U+D7FF, |
| 82 | the 90..BF in U+10000..U+3FFFF, and the 80...8F in U+100000..U+10FFFF. |
| 83 | The "gaps" are caused by legal UTF-8 avoiding non-shortest encodings: |
| 84 | it is technically possible to UTF-8-encode a single code point in different |
| 85 | ways, but that is explicitly forbidden, and the shortest possible encoding |
| 86 | should always be used (and that is what Perl does). |
| 87 | |
| 88 | */ |
| 89 | |
| 90 | /* |
| 91 | Another way to look at it, as bits: |
| 92 | |
| 93 | Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte |
| 94 | |
| 95 | 0aaaaaaa 0aaaaaaa |
| 96 | 00000bbbbbaaaaaa 110bbbbb 10aaaaaa |
| 97 | ccccbbbbbbaaaaaa 1110cccc 10bbbbbb 10aaaaaa |
| 98 | 00000dddccccccbbbbbbaaaaaa 11110ddd 10cccccc 10bbbbbb 10aaaaaa |
| 99 | |
| 100 | As you can see, the continuation bytes all begin with C<10>, and the |
| 101 | leading bits of the start byte tell how many bytes the are in the |
| 102 | encoded character. |
| 103 | |
| 104 | */ |
| 105 | |
| 106 | |
| 107 | #define UNI_IS_INVARIANT(c) (((UV)c) < 0x80) |
| 108 | #define UTF8_IS_INVARIANT(c) UNI_IS_INVARIANT(NATIVE_TO_UTF(c)) |
| 109 | #define NATIVE_IS_INVARIANT(c) UNI_IS_INVARIANT(NATIVE_TO_ASCII(c)) |
| 110 | #define UTF8_IS_START(c) (((U8)c) >= 0xc0 && (((U8)c) <= 0xfd)) |
| 111 | #define UTF8_IS_CONTINUATION(c) (((U8)c) >= 0x80 && (((U8)c) <= 0xbf)) |
| 112 | #define UTF8_IS_CONTINUED(c) (((U8)c) & 0x80) |
| 113 | #define UTF8_IS_DOWNGRADEABLE_START(c) (((U8)c & 0xfc) == 0xc0) |
| 114 | |
| 115 | #define UTF_START_MARK(len) ((len > 7) ? 0xFF : (0xFE << (7-len))) |
| 116 | #define UTF_START_MASK(len) ((len >= 7) ? 0x00 : (0x1F >> (len-2))) |
| 117 | |
| 118 | #define UTF_CONTINUATION_MARK 0x80 |
| 119 | #define UTF_ACCUMULATION_SHIFT 6 |
| 120 | #define UTF_CONTINUATION_MASK ((U8)0x3f) |
| 121 | #define UTF8_ACCUMULATE(old, new) (((old) << UTF_ACCUMULATION_SHIFT) | (((U8)new) & UTF_CONTINUATION_MASK)) |
| 122 | |
| 123 | #define UTF8_EIGHT_BIT_HI(c) ((((U8)(c))>>UTF_ACCUMULATION_SHIFT)|UTF_START_MARK(2)) |
| 124 | #define UTF8_EIGHT_BIT_LO(c) (((((U8)(c)))&UTF_CONTINUATION_MASK)|UTF_CONTINUATION_MARK) |
| 125 | |
| 126 | #ifdef HAS_QUAD |
| 127 | #define UNISKIP(uv) ( (uv) < 0x80 ? 1 : \ |
| 128 | (uv) < 0x800 ? 2 : \ |
| 129 | (uv) < 0x10000 ? 3 : \ |
| 130 | (uv) < 0x200000 ? 4 : \ |
| 131 | (uv) < 0x4000000 ? 5 : \ |
| 132 | (uv) < 0x80000000 ? 6 : \ |
| 133 | (uv) < UTF8_QUAD_MAX ? 7 : 13 ) |
| 134 | #else |
| 135 | /* No, I'm not even going to *TRY* putting #ifdef inside a #define */ |
| 136 | #define UNISKIP(uv) ( (uv) < 0x80 ? 1 : \ |
| 137 | (uv) < 0x800 ? 2 : \ |
| 138 | (uv) < 0x10000 ? 3 : \ |
| 139 | (uv) < 0x200000 ? 4 : \ |
| 140 | (uv) < 0x4000000 ? 5 : \ |
| 141 | (uv) < 0x80000000 ? 6 : 7 ) |
| 142 | #endif |
| 143 | |
| 144 | /* |
| 145 | * Note: we try to be careful never to call the isXXX_utf8() functions |
| 146 | * unless we're pretty sure we've seen the beginning of a UTF-8 character |
| 147 | * (that is, the two high bits are set). Otherwise we risk loading in the |
| 148 | * heavy-duty SWASHINIT and SWASHGET routines unnecessarily. |
| 149 | */ |
| 150 | #define isIDFIRST_lazy_if(p,c) ((IN_BYTES || (!c || (*((U8*)p) < 0xc0))) \ |
| 151 | ? isIDFIRST(*(p)) \ |
| 152 | : isIDFIRST_utf8((U8*)p)) |
| 153 | #define isALNUM_lazy_if(p,c) ((IN_BYTES || (!c || (*((U8*)p) < 0xc0))) \ |
| 154 | ? isALNUM(*(p)) \ |
| 155 | : isALNUM_utf8((U8*)p)) |
| 156 | |
| 157 | |
| 158 | #endif /* EBCDIC vs ASCII */ |
| 159 | |
| 160 | /* Rest of these are attributes of Unicode and perl's internals rather than the encoding */ |
| 161 | |
| 162 | #define isIDFIRST_lazy(p) isIDFIRST_lazy_if(p,1) |
| 163 | #define isALNUM_lazy(p) isALNUM_lazy_if(p,1) |
| 164 | |
| 165 | /* how wide can a single UTF8 encoded character become */ |
| 166 | #define UTF8_MAXLEN 13 |
| 167 | /* how wide a character can become when upper/lowercased */ |
| 168 | #define UTF8_MAXLEN_UCLC_MULT 3 |
| 169 | #define UTF8_MAXLEN_UCLC (UTF8_MAXLEN*UTF8_MAXLEN_UCLC_MULT) |
| 170 | /* how wide a character can become when casefolded */ |
| 171 | #define UTF8_MAXLEN_FOLD_MULT 3 |
| 172 | #define UTF8_MAXLEN_FOLD (UTF8_MAXLEN*UTF8_MAXLEN_FOLD_MULT) |
| 173 | |
| 174 | #define IN_BYTES (PL_curcop->op_private & HINT_BYTES) |
| 175 | #define DO_UTF8(sv) (SvUTF8(sv) && !IN_BYTES) |
| 176 | |
| 177 | #define UTF8_ALLOW_EMPTY 0x0001 |
| 178 | #define UTF8_ALLOW_CONTINUATION 0x0002 |
| 179 | #define UTF8_ALLOW_NON_CONTINUATION 0x0004 |
| 180 | #define UTF8_ALLOW_FE_FF 0x0008 |
| 181 | #define UTF8_ALLOW_SHORT 0x0010 |
| 182 | #define UTF8_ALLOW_SURROGATE 0x0020 |
| 183 | #define UTF8_ALLOW_FFFF 0x0040 /* Allows also FFFE. */ |
| 184 | #define UTF8_ALLOW_LONG 0x0080 |
| 185 | #define UTF8_ALLOW_ANYUV (UTF8_ALLOW_EMPTY|UTF8_ALLOW_FE_FF|\ |
| 186 | UTF8_ALLOW_SURROGATE|\ |
| 187 | UTF8_ALLOW_FFFF|UTF8_ALLOW_LONG) |
| 188 | #define UTF8_ALLOW_ANY 0x00FF |
| 189 | #define UTF8_CHECK_ONLY 0x0200 |
| 190 | |
| 191 | #define UNICODE_SURROGATE_FIRST 0xD800 |
| 192 | #define UNICODE_SURROGATE_LAST 0xDFFF |
| 193 | #define UNICODE_REPLACEMENT 0xFFFD |
| 194 | #define UNICODE_BYTE_ORDER_MARK 0xFEFF |
| 195 | #define UNICODE_ILLEGAL 0xFFFF |
| 196 | |
| 197 | /* Though our UTF-8 encoding can go beyond this, |
| 198 | * let's be conservative and do as Unicode 3.2 says. */ |
| 199 | #define PERL_UNICODE_MAX 0x10FFFF |
| 200 | |
| 201 | #define UNICODE_ALLOW_SURROGATE 0x0001 /* Allow UTF-16 surrogates (EVIL) */ |
| 202 | #define UNICODE_ALLOW_FDD0 0x0002 /* Allow the U+FDD0...U+FDEF */ |
| 203 | #define UNICODE_ALLOW_FFFF 0x0004 /* Allow 0xFFF[EF], 0x1FFF[EF], ... */ |
| 204 | #define UNICODE_ALLOW_SUPER 0x0008 /* Allow past 10xFFFF */ |
| 205 | #define UNICODE_ALLOW_ANY 0x000F |
| 206 | |
| 207 | #define UNICODE_IS_SURROGATE(c) ((c) >= UNICODE_SURROGATE_FIRST && \ |
| 208 | (c) <= UNICODE_SURROGATE_LAST) |
| 209 | #define UNICODE_IS_REPLACEMENT(c) ((c) == UNICODE_REPLACEMENT) |
| 210 | #define UNICODE_IS_BYTE_ORDER_MARK(c) ((c) == UNICODE_BYTE_ORDER_MARK) |
| 211 | #define UNICODE_IS_ILLEGAL(c) ((c) == UNICODE_ILLEGAL) |
| 212 | |
| 213 | #ifdef HAS_QUAD |
| 214 | # define UTF8_QUAD_MAX UINT64_C(0x1000000000) |
| 215 | #endif |
| 216 | |
| 217 | #define UTF8_IS_ASCII(c) UTF8_IS_INVARIANT(c) |
| 218 | |
| 219 | #define UNICODE_LATIN_SMALL_LETTER_SHARP_S 0x00DF |
| 220 | #define UNICODE_GREEK_CAPITAL_LETTER_SIGMA 0x03A3 |
| 221 | #define UNICODE_GREEK_SMALL_LETTER_FINAL_SIGMA 0x03C2 |
| 222 | #define UNICODE_GREEK_SMALL_LETTER_SIGMA 0x03C3 |
| 223 | |
| 224 | #define EBCDIC_LATIN_SMALL_LETTER_SHARP_S 0x0059 |
| 225 | |
| 226 | #define UNI_DISPLAY_ISPRINT 0x0001 |
| 227 | #define UNI_DISPLAY_BACKSLASH 0x0002 |
| 228 | #define UNI_DISPLAY_QQ (UNI_DISPLAY_ISPRINT|UNI_DISPLAY_BACKSLASH) |
| 229 | #define UNI_DISPLAY_REGEX (UNI_DISPLAY_ISPRINT|UNI_DISPLAY_BACKSLASH) |
| 230 | |
| 231 | #ifdef EBCDIC |
| 232 | # define ANYOF_FOLD_SHARP_S(node, input, end) \ |
| 233 | (ANYOF_BITMAP_TEST(node, EBCDIC_LATIN_SMALL_LETTER_SHARP_S) && \ |
| 234 | (ANYOF_FLAGS(node) & ANYOF_UNICODE) && \ |
| 235 | (ANYOF_FLAGS(node) & ANYOF_FOLD) && \ |
| 236 | ((end) > (input) + 1) && \ |
| 237 | toLOWER((input)[0]) == 's' && \ |
| 238 | toLOWER((input)[1]) == 's') |
| 239 | #else |
| 240 | # define ANYOF_FOLD_SHARP_S(node, input, end) \ |
| 241 | (ANYOF_BITMAP_TEST(node, UNICODE_LATIN_SMALL_LETTER_SHARP_S) && \ |
| 242 | (ANYOF_FLAGS(node) & ANYOF_UNICODE) && \ |
| 243 | (ANYOF_FLAGS(node) & ANYOF_FOLD) && \ |
| 244 | ((end) > (input) + 1) && \ |
| 245 | toLOWER((input)[0]) == 's' && \ |
| 246 | toLOWER((input)[1]) == 's') |
| 247 | #endif |
| 248 | #define SHARP_S_SKIP 2 |