[Encode] 1.77 Released
[p5sagit/p5-mst-13.2.git] / utf8.h
CommitLineData
a0ed51b3 1/* utf8.h
2 *
be3c0a43 3 * Copyright (c) 1998-2002, Larry Wall
a0ed51b3 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
39e02b42 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
1d72bdf6 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
73c4f7a1 26START_EXTERN_C
27
a0ed51b3 28#ifdef DOINIT
6f06b55f 29EXTCONST unsigned char PL_utf8skip[] = {
a0ed51b3 301,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
311,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
321,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
331,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
341,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* bogus */
351,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* bogus */
362,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, /* scripts */
3c77ea2b 373,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,4,4,4,4,4,4,4,4,5,5,5,5,6,6, /* cjk etc. */
387,13, /* Perl extended (not UTF-8). Up to 72bit allowed (64-bit + reserved). */
a0ed51b3 39};
40#else
6f06b55f 41EXTCONST unsigned char PL_utf8skip[];
a0ed51b3 42#endif
43
73c4f7a1 44END_EXTERN_C
6f06b55f 45#define UTF8SKIP(s) PL_utf8skip[*(U8*)s]
7e2040f0 46
1d72bdf6 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)
d7578b48 59
2b9d42f0 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
877d9f0d 64/*
9041c2e3 65
8c007b5a 66 The following table is from Unicode 3.2.
877d9f0d 67
68 Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
69
375122d7 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
8c007b5a 75 U+D800..U+DFFF ******* ill-formed *******
375122d7 76 U+E000..U+FFFF EE..EF 80..BF 80..BF
877d9f0d 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
8c007b5a 81Note the A0..BF in U+0800..U+0FFF, the 80..9F in U+D000...U+D7FF,
82the 90..BF in U+10000..U+3FFFF, and the 80...8F in U+100000..U+10FFFF.
37361303 83The "gaps" are caused by legal UTF-8 avoiding non-shortest encodings:
84it is technically possible to UTF-8-encode a single code point in different
85ways, but that is explicitly forbidden, and the shortest possible encoding
86should always be used (and that is what Perl does).
8c007b5a 87
877d9f0d 88 */
89
8c007b5a 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
100As you can see, the continuation bytes all begin with C<10>, and the
101leading bits of the start byte tell how many bytes the are in the
102encoded character.
103
104*/
105
106
c4d5f83a 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))
c512ce4f 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)
db42d148 113#define UTF8_IS_DOWNGRADEABLE_START(c) (((U8)c & 0xfc) == 0xc0)
8850bf83 114
1d72bdf6 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)
cb359b41 121#define UTF8_ACCUMULATE(old, new) (((old) << UTF_ACCUMULATION_SHIFT) | (((U8)new) & UTF_CONTINUATION_MASK))
421a8bf2 122
1d72bdf6 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)
c512ce4f 125
1d68d6cd 126#ifdef HAS_QUAD
5bbb0b5a 127#define UNISKIP(uv) ( (uv) < 0x80 ? 1 : \
1d68d6cd 128 (uv) < 0x800 ? 2 : \
129 (uv) < 0x10000 ? 3 : \
130 (uv) < 0x200000 ? 4 : \
131 (uv) < 0x4000000 ? 5 : \
132 (uv) < 0x80000000 ? 6 : \
9041c2e3 133 (uv) < UTF8_QUAD_MAX ? 7 : 13 )
1d68d6cd 134#else
135/* No, I'm not even going to *TRY* putting #ifdef inside a #define */
5bbb0b5a 136#define UNISKIP(uv) ( (uv) < 0x80 ? 1 : \
1d68d6cd 137 (uv) < 0x800 ? 2 : \
138 (uv) < 0x10000 ? 3 : \
139 (uv) < 0x200000 ? 4 : \
140 (uv) < 0x4000000 ? 5 : \
141 (uv) < 0x80000000 ? 6 : 7 )
142#endif
143
7e2040f0 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 */
0064a8a9 150#define isIDFIRST_lazy_if(p,c) ((IN_BYTES || (!c || (*((U8*)p) < 0xc0))) \
7e2040f0 151 ? isIDFIRST(*(p)) \
152 : isIDFIRST_utf8((U8*)p))
0064a8a9 153#define isALNUM_lazy_if(p,c) ((IN_BYTES || (!c || (*((U8*)p) < 0xc0))) \
7e2040f0 154 ? isALNUM(*(p)) \
155 : isALNUM_utf8((U8*)p))
1d72bdf6 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
7e2040f0 162#define isIDFIRST_lazy(p) isIDFIRST_lazy_if(p,1)
163#define isALNUM_lazy(p) isALNUM_lazy_if(p,1)
3bd709b1 164
e7ae6809 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 */
3ec562b0 168#define UTF8_MAXLEN_UCLC_MULT 3
169#define UTF8_MAXLEN_UCLC (UTF8_MAXLEN*UTF8_MAXLEN_UCLC_MULT)
e7ae6809 170/* how wide a character can become when casefolded */
3ec562b0 171#define UTF8_MAXLEN_FOLD_MULT 3
172#define UTF8_MAXLEN_FOLD (UTF8_MAXLEN*UTF8_MAXLEN_FOLD_MULT)
3bd709b1 173
0064a8a9 174#define IN_BYTES (PL_curcop->op_private & HINT_BYTES)
175#define DO_UTF8(sv) (SvUTF8(sv) && !IN_BYTES)
1d72bdf6 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
c867b360 183#define UTF8_ALLOW_FFFF 0x0040 /* Allows also FFFE. */
184#define UTF8_ALLOW_LONG 0x0080
1d72bdf6 185#define UTF8_ALLOW_ANYUV (UTF8_ALLOW_EMPTY|UTF8_ALLOW_FE_FF|\
c867b360 186 UTF8_ALLOW_SURROGATE|\
1d72bdf6 187 UTF8_ALLOW_FFFF|UTF8_ALLOW_LONG)
c867b360 188#define UTF8_ALLOW_ANY 0x00FF
e83d50c9 189#define UTF8_CHECK_ONLY 0x0200
1d72bdf6 190
c867b360 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
1d72bdf6 196
b851fbc1 197/* Though our UTF-8 encoding can go beyond this,
872c91ae 198 * let's be conservative and do as Unicode 3.2 says. */
b851fbc1 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 */
c867b360 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
b851fbc1 206
1d72bdf6 207#define UNICODE_IS_SURROGATE(c) ((c) >= UNICODE_SURROGATE_FIRST && \
208 (c) <= UNICODE_SURROGATE_LAST)
a10ec373 209#define UNICODE_IS_REPLACEMENT(c) ((c) == UNICODE_REPLACEMENT)
872c91ae 210#define UNICODE_IS_BYTE_ORDER_MARK(c) ((c) == UNICODE_BYTE_ORDER_MARK)
1d72bdf6 211#define UNICODE_IS_ILLEGAL(c) ((c) == UNICODE_ILLEGAL)
212
f067b878 213#ifdef HAS_QUAD
214# define UTF8_QUAD_MAX UINT64_C(0x1000000000)
215#endif
3bd709b1 216
1d72bdf6 217#define UTF8_IS_ASCII(c) UTF8_IS_INVARIANT(c)
09091399 218
e0f9d4a8 219#define UNICODE_LATIN_SMALL_LETTER_SHARP_S 0x00DF
09091399 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
ebc501f0 224#define EBCDIC_LATIN_SMALL_LETTER_SHARP_S 0x0059
225
9e55ce06 226#define UNI_DISPLAY_ISPRINT 0x0001
c728cb41 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)
9e55ce06 230
ebc501f0 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) \
07b6858f 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')
ebc501f0 247#endif
248#define SHARP_S_SKIP 2