This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Avoid wraparound when casting unsigned size_t to signed ssize_t.
[perl5.git] / utf8.h
CommitLineData
a0ed51b3
LW
1/* utf8.h
2 *
2eee27d7
SS
3 * Copyright (C) 2000, 2001, 2002, 2005, 2006, 2007, 2009,
4 * 2010, 2011 by Larry Wall and others
a0ed51b3
LW
5 *
6 * You may distribute under the terms of either the GNU General Public
7 * License or the Artistic License, as specified in the README file.
8 *
9 */
10
57f0e7e2
KW
11#ifndef H_UTF8 /* Guard against recursive inclusion */
12#define H_UTF8 1
13
39e02b42 14/* Use UTF-8 as the default script encoding?
1e54db1a 15 * Turning this on will break scripts having non-UTF-8 binary
39e02b42
JH
16 * data (such as Latin-1) in string literals. */
17#ifdef USE_UTF8_SCRIPTS
18# define USE_UTF8_IN_NAMES (!IN_BYTES)
19#else
20# define USE_UTF8_IN_NAMES (PL_hints & HINT_UTF8)
21#endif
22
051a06d4
KW
23/* For to_utf8_fold_flags, q.v. */
24#define FOLD_FLAGS_LOCALE 0x1
25#define FOLD_FLAGS_FULL 0x2
a0270393 26#define FOLD_FLAGS_NOMIX_ASCII 0x4
051a06d4 27
83199d38
KW
28/* For _core_swash_init(), internal core use only */
29#define _CORE_SWASH_INIT_USER_DEFINED_PROPERTY 0x1
5d3d13d1 30#define _CORE_SWASH_INIT_RETURN_IF_UNDEF 0x2
87367d5f 31#define _CORE_SWASH_INIT_ACCEPT_INVLIST 0x4
83199d38 32
a0270393 33#define to_uni_fold(c, p, lenp) _to_uni_fold_flags(c, p, lenp, FOLD_FLAGS_FULL)
051a06d4
KW
34#define to_utf8_fold(c, p, lenp) _to_utf8_fold_flags(c, p, lenp, \
35 FOLD_FLAGS_FULL, NULL)
36#define to_utf8_lower(a,b,c) _to_utf8_lower_flags(a,b,c,0, NULL)
37#define to_utf8_upper(a,b,c) _to_utf8_upper_flags(a,b,c,0, NULL)
38#define to_utf8_title(a,b,c) _to_utf8_title_flags(a,b,c,0, NULL)
36bb2ab6 39
fd7cb289
RGS
40/* Source backward compatibility. */
41#define uvuni_to_utf8(d, uv) uvuni_to_utf8_flags(d, uv, 0)
42#define is_utf8_string_loc(s, len, ep) is_utf8_string_loclen(s, len, ep, 0)
43
eda9cac1
KW
44#define foldEQ_utf8(s1, pe1, l1, u1, s2, pe2, l2, u2) \
45 foldEQ_utf8_flags(s1, pe1, l1, u1, s2, pe2, l2, u2, 0)
a33c29bc 46#define FOLDEQ_UTF8_NOMIX_ASCII (1 << 0)
5e64d0fa 47#define FOLDEQ_UTF8_LOCALE (1 << 1)
18f762c3
KW
48#define FOLDEQ_S1_ALREADY_FOLDED (1 << 2)
49#define FOLDEQ_S2_ALREADY_FOLDED (1 << 3)
a33c29bc 50
e6226b18
KW
51/*
52=for apidoc ibcmp_utf8
53
54This is a synonym for (! foldEQ_utf8())
55
56=cut
57*/
58#define ibcmp_utf8(s1, pe1, l1, u1, s2, pe2, l2, u2) \
59 cBOOL(! foldEQ_utf8(s1, pe1, l1, u1, s2, pe2, l2, u2))
60
1d72bdf6
NIS
61#ifdef EBCDIC
62/* The equivalent of these macros but implementing UTF-EBCDIC
63 are in the following header file:
64 */
65
66#include "utfebcdic.h"
fd7cb289 67
d06134e5 68#else /* ! EBCDIC */
73c4f7a1
GS
69START_EXTERN_C
70
a0ed51b3 71#ifdef DOINIT
6f06b55f 72EXTCONST unsigned char PL_utf8skip[] = {
b2635aa8
KW
73/* 0x00 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
74/* 0x10 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
75/* 0x20 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
76/* 0x30 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
77/* 0x40 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
78/* 0x50 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
79/* 0x60 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
80/* 0x70 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
81/* 0x80 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* bogus: continuation byte */
82/* 0x90 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* bogus: continuation byte */
83/* 0xA0 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* bogus: continuation byte */
84/* 0xB0 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* bogus: continuation byte */
85/* 0xC0 */ 2,2, /* overlong */
86/* 0xC2 */ 2,2,2,2,2,2,2,2,2,2,2,2,2,2, /* U+0080 to U+03FF */
87/* 0xD0 */ 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, /* U+0400 to U+07FF */
88/* 0xE0 */ 3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3, /* U+0800 to U+FFFF */
89/* 0xF0 */ 4,4,4,4,4,4,4,4,5,5,5,5,6,6, /* above BMP to 2**31 - 1 */
90/* 0xFE */ 7,13, /* Perl extended (never was official UTF-8). Up to 72bit
91 allowed (64-bit + reserved). */
a0ed51b3
LW
92};
93#else
6f06b55f 94EXTCONST unsigned char PL_utf8skip[];
a0ed51b3
LW
95#endif
96
73c4f7a1 97END_EXTERN_C
7e2040f0 98
b96a92fb 99#include "regcharclass.h"
ec34087a
KW
100#include "unicode_constants.h"
101
1d72bdf6
NIS
102/* Native character to iso-8859-1 */
103#define NATIVE_TO_ASCII(ch) (ch)
104#define ASCII_TO_NATIVE(ch) (ch)
105/* Transform after encoding */
106#define NATIVE_TO_UTF(ch) (ch)
428921e5 107#define NATIVE_TO_I8(ch) NATIVE_TO_UTF(ch) /* a clearer synonym */
1d72bdf6 108#define UTF_TO_NATIVE(ch) (ch)
428921e5 109#define I8_TO_NATIVE(ch) UTF_TO_NATIVE(ch)
1d72bdf6
NIS
110/* Transforms in wide UV chars */
111#define UNI_TO_NATIVE(ch) (ch)
112#define NATIVE_TO_UNI(ch) (ch)
113/* Transforms in invariant space */
114#define NATIVE_TO_NEED(enc,ch) (ch)
115#define ASCII_TO_NEED(enc,ch) (ch)
d7578b48 116
d06134e5 117/* As there are no translations, avoid the function wrapper */
1754c1a1 118#define utf8n_to_uvchr utf8n_to_uvuni
6dd9dce9 119#define valid_utf8_to_uvchr valid_utf8_to_uvuni
1754c1a1 120#define uvchr_to_utf8 uvuni_to_utf8
2b9d42f0 121
877d9f0d 122/*
9041c2e3 123
8c007b5a 124 The following table is from Unicode 3.2.
877d9f0d
JH
125
126 Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
127
375122d7 128 U+0000..U+007F 00..7F
e1b711da 129 U+0080..U+07FF * C2..DF 80..BF
37e2e78e 130 U+0800..U+0FFF E0 * A0..BF 80..BF
375122d7 131 U+1000..U+CFFF E1..EC 80..BF 80..BF
e1b711da 132 U+D000..U+D7FF ED 80..9F 80..BF
37e2e78e 133 U+D800..U+DFFF +++++++ utf16 surrogates, not legal utf8 +++++++
375122d7 134 U+E000..U+FFFF EE..EF 80..BF 80..BF
37e2e78e 135 U+10000..U+3FFFF F0 * 90..BF 80..BF 80..BF
877d9f0d
JH
136 U+40000..U+FFFFF F1..F3 80..BF 80..BF 80..BF
137 U+100000..U+10FFFF F4 80..8F 80..BF 80..BF
b2635aa8
KW
138 Below are non-Unicode code points
139 U+110000..U+13FFFF F4 90..BF 80..BF 80..BF
140 U+110000..U+1FFFFF F5..F7 80..BF 80..BF 80..BF
141 U+200000: F8.. * 88..BF 80..BF 80..BF 80..BF
877d9f0d 142
e1b711da 143Note the gaps before several of the byte entries above marked by '*'. These are
37e2e78e
KW
144caused by legal UTF-8 avoiding non-shortest encodings: it is technically
145possible to UTF-8-encode a single code point in different ways, but that is
146explicitly forbidden, and the shortest possible encoding should always be used
147(and that is what Perl does).
8c007b5a 148
877d9f0d
JH
149 */
150
8c007b5a
JH
151/*
152 Another way to look at it, as bits:
153
b2635aa8 154 Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
8c007b5a 155
b2635aa8
KW
156 0aaa aaaa 0aaa aaaa
157 0000 0bbb bbaa aaaa 110b bbbb 10aa aaaa
158 cccc bbbb bbaa aaaa 1110 cccc 10bb bbbb 10aa aaaa
159 00 000d ddcc cccc bbbb bbaa aaaa 1111 0ddd 10cc cccc 10bb bbbb 10aa aaaa
8c007b5a
JH
160
161As you can see, the continuation bytes all begin with C<10>, and the
e1b711da 162leading bits of the start byte tell how many bytes there are in the
8c007b5a
JH
163encoded character.
164
65ab9279
TC
165Perl's extended UTF-8 means we can have start bytes up to FF.
166
8c007b5a
JH
167*/
168
c4d5f83a 169#define UNI_IS_INVARIANT(c) (((UV)c) < 0x80)
0447e8df 170#define UTF8_IS_START(c) (((U8)c) >= 0xc2)
e021c6e6 171#define UTF8_IS_CONTINUATION(c) ((((U8)c) & 0xC0) == 0x80)
c512ce4f 172#define UTF8_IS_CONTINUED(c) (((U8)c) & 0x80)
0ae1fa71 173
4ab10950
KW
174/* Use UTF8_IS_NEXT_CHAR_DOWNGRADEABLE() instead if the input isn't known to
175 * be well-formed. Masking with 0xfe allows low bit to be 0 or 1; thus this
176 * matches 0xc[23]. */
559c7f10 177#define UTF8_IS_DOWNGRADEABLE_START(c) (((U8)(c) & 0xfe) == 0xc2)
4ab10950 178
7b4252f4 179#define UTF8_IS_ABOVE_LATIN1(c) ((U8)(c) >= 0xc4)
8850bf83 180
22901f30 181#define UTF_START_MARK(len) (((len) > 7) ? 0xFF : (0xFE << (7-(len))))
b2635aa8
KW
182
183/* Masks out the initial one bits in a start byte, leaving the real data ones.
184 * Doesn't work on an invariant byte */
22901f30 185#define UTF_START_MASK(len) (((len) >= 7) ? 0x00 : (0x1F >> ((len)-2)))
1d72bdf6
NIS
186
187#define UTF_CONTINUATION_MARK 0x80
188#define UTF_ACCUMULATION_SHIFT 6
b2635aa8
KW
189
190/* 2**UTF_ACCUMULATION_SHIFT - 1 */
1d72bdf6 191#define UTF_CONTINUATION_MASK ((U8)0x3f)
c512ce4f 192
eb83ed87
KW
193/* This sets the UTF_CONTINUATION_MASK in the upper bits of a word. If a value
194 * is anded with it, and the result is non-zero, then using the original value
195 * in UTF8_ACCUMULATE will overflow, shifting bits off the left */
196#define UTF_ACCUMULATION_OVERFLOW_MASK \
bb88be5f
KW
197 (((UV) UTF_CONTINUATION_MASK) << ((sizeof(UV) * CHARBITS) \
198 - UTF_ACCUMULATION_SHIFT))
eb83ed87 199
1d68d6cd 200#ifdef HAS_QUAD
5bbb0b5a 201#define UNISKIP(uv) ( (uv) < 0x80 ? 1 : \
1d68d6cd
SC
202 (uv) < 0x800 ? 2 : \
203 (uv) < 0x10000 ? 3 : \
204 (uv) < 0x200000 ? 4 : \
205 (uv) < 0x4000000 ? 5 : \
206 (uv) < 0x80000000 ? 6 : \
9041c2e3 207 (uv) < UTF8_QUAD_MAX ? 7 : 13 )
1d68d6cd
SC
208#else
209/* No, I'm not even going to *TRY* putting #ifdef inside a #define */
5bbb0b5a 210#define UNISKIP(uv) ( (uv) < 0x80 ? 1 : \
1d68d6cd
SC
211 (uv) < 0x800 ? 2 : \
212 (uv) < 0x10000 ? 3 : \
213 (uv) < 0x200000 ? 4 : \
214 (uv) < 0x4000000 ? 5 : \
215 (uv) < 0x80000000 ? 6 : 7 )
216#endif
217
03c76984
KW
218/* How wide can a single UTF-8 encoded character become in bytes. */
219/* NOTE: Strictly speaking Perl's UTF-8 should not be called UTF-8 since UTF-8
220 * is an encoding of Unicode, and Unicode's upper limit, 0x10FFFF, can be
221 * expressed with 4 bytes. However, Perl thinks of UTF-8 as a way to encode
222 * non-negative integers in a binary format, even those above Unicode */
223#define UTF8_MAXBYTES 13
224
225/* The maximum number of UTF-8 bytes a single Unicode character can
226 * uppercase/lowercase/fold into. Unicode guarantees that the maximum
227 * expansion is 3 characters. On ASCIIish platforms, the highest Unicode
228 * character occupies 4 bytes, therefore this number would be 12, but this is
229 * smaller than the maximum width a single above-Unicode character can occupy,
230 * so use that instead */
231#if UTF8_MAXBYTES < 12
232#error UTF8_MAXBYTES must be at least 12
233#endif
234
235#define UTF8_MAXBYTES_CASE UTF8_MAXBYTES
236
d06134e5
KW
237#endif /* EBCDIC vs ASCII */
238
239/* Rest of these are attributes of Unicode and perl's internals rather than the
240 * encoding, or happen to be the same in both ASCII and EBCDIC (at least at
241 * this level; the macros that some of these call may have different
242 * definitions in the two encodings */
243
244#define NATIVE8_TO_UNI(ch) NATIVE_TO_ASCII(ch) /* a clearer synonym */
245
bb88be5f
KW
246#define UTF8_ACCUMULATE(old, new) (((old) << UTF_ACCUMULATION_SHIFT) \
247 | (((U8)new) & UTF_CONTINUATION_MASK))
d06134e5 248
4ab10950
KW
249/* This works in the face of malformed UTF-8. */
250#define UTF8_IS_NEXT_CHAR_DOWNGRADEABLE(s, e) (UTF8_IS_DOWNGRADEABLE_START(*s) \
251 && ( (e) - (s) > 1) \
252 && UTF8_IS_CONTINUATION(*((s)+1)))
253
2950f2a7
KW
254/* Convert a two (not one) byte utf8 character to a unicode code point value.
255 * Needs just one iteration of accumulate. Should not be used unless it is
256 * known that the two bytes are legal: 1) two-byte start, and 2) continuation.
257 * Note that the result can be larger than 255 if the input character is not
258 * downgradable */
259#define TWO_BYTE_UTF8_TO_UNI(HI, LO) \
260 UTF8_ACCUMULATE((NATIVE_TO_UTF(HI) & UTF_START_MASK(2)), \
261 NATIVE_TO_UTF(LO))
262
d06134e5
KW
263#define UTF8SKIP(s) PL_utf8skip[*(const U8*)(s)]
264
265#define UTF8_IS_INVARIANT(c) UNI_IS_INVARIANT(NATIVE_TO_UTF(c))
266#define NATIVE_IS_INVARIANT(c) UNI_IS_INVARIANT(NATIVE8_TO_UNI(c))
267
268#define MAX_PORTABLE_UTF8_TWO_BYTE 0x3FF /* constrained by EBCDIC */
269
270/* The macros in the next sets are used to generate the two utf8 or utfebcdic
271 * bytes from an ordinal that is known to fit into two bytes; it must be less
272 * than 0x3FF to work across both encodings. */
273/* Nocast allows these to be used in the case label of a switch statement */
428921e5
KW
274#define UTF8_TWO_BYTE_HI_nocast(c) NATIVE_TO_I8(((c) \
275 >> UTF_ACCUMULATION_SHIFT) | (0xFF & UTF_START_MARK(2)))
276#define UTF8_TWO_BYTE_LO_nocast(c) NATIVE_TO_I8(((c) & UTF_CONTINUATION_MASK) \
277 | UTF_CONTINUATION_MARK)
d06134e5
KW
278
279#define UTF8_TWO_BYTE_HI(c) ((U8) (UTF8_TWO_BYTE_HI_nocast(c)))
280#define UTF8_TWO_BYTE_LO(c) ((U8) (UTF8_TWO_BYTE_LO_nocast(c)))
281
282/* This name is used when the source is a single byte */
283#define UTF8_EIGHT_BIT_HI(c) UTF8_TWO_BYTE_HI((U8)(c))
284#define UTF8_EIGHT_BIT_LO(c) UTF8_TWO_BYTE_LO((U8)(c))
285
7e2040f0 286/*
e3036cf4 287 * 'UTF' is whether or not p is encoded in UTF8. The names 'foo_lazy_if' stem
20df05f4
KW
288 * from an earlier version of these macros in which they didn't call the
289 * foo_utf8() macros (i.e. were 'lazy') unless they decided that *p is the
290 * beginning of a utf8 character. Now that foo_utf8() determines that itself,
291 * no need to do it again here
7e2040f0 292 */
e3036cf4
KW
293#define isIDFIRST_lazy_if(p,UTF) ((IN_BYTES || !UTF ) \
294 ? isIDFIRST(*(p)) \
295 : isIDFIRST_utf8((const U8*)p))
32636478
KW
296#define isWORDCHAR_lazy_if(p,UTF) ((IN_BYTES || (!UTF )) \
297 ? isWORDCHAR(*(p)) \
298 : isWORDCHAR_utf8((const U8*)p))
299#define isALNUM_lazy_if(p,UTF) isWORDCHAR_lazy_if(p,UTF)
1d72bdf6 300
89ebb4a3
JH
301#define UTF8_MAXLEN UTF8_MAXBYTES
302
8cb75cc8
KW
303/* A Unicode character can fold to up to 3 characters */
304#define UTF8_MAX_FOLD_CHAR_EXPAND 3
305
a98fe34d 306#define IN_BYTES (CopHINTS_get(PL_curcop) & HINT_BYTES)
0064a8a9 307#define DO_UTF8(sv) (SvUTF8(sv) && !IN_BYTES)
66cbab2c 308#define IN_UNI_8_BIT \
b36bf33f
KW
309 (CopHINTS_get(PL_curcop) & (HINT_UNI_8_BIT|HINT_LOCALE_NOT_CHARS) \
310 && ! IN_LOCALE_RUNTIME && ! IN_BYTES)
311
1d72bdf6 312
c76687c5
KW
313#define UTF8_ALLOW_EMPTY 0x0001 /* Allow a zero length string */
314
315/* Allow first byte to be a continuation byte */
1d72bdf6 316#define UTF8_ALLOW_CONTINUATION 0x0002
c76687c5
KW
317
318/* Allow second... bytes to be non-continuation bytes */
1d72bdf6 319#define UTF8_ALLOW_NON_CONTINUATION 0x0004
949cf498
KW
320
321/* expecting more bytes than were available in the string */
322#define UTF8_ALLOW_SHORT 0x0008
323
324/* Overlong sequence; i.e., the code point can be specified in fewer bytes. */
325#define UTF8_ALLOW_LONG 0x0010
326
327#define UTF8_DISALLOW_SURROGATE 0x0020 /* Unicode surrogates */
328#define UTF8_WARN_SURROGATE 0x0040
329
330#define UTF8_DISALLOW_NONCHAR 0x0080 /* Unicode non-character */
331#define UTF8_WARN_NONCHAR 0x0100 /* code points */
332
333#define UTF8_DISALLOW_SUPER 0x0200 /* Super-set of Unicode: code */
334#define UTF8_WARN_SUPER 0x0400 /* points above the legal max */
335
336/* Code points which never were part of the original UTF-8 standard, the first
337 * byte of which is a FE or FF on ASCII platforms. */
338#define UTF8_DISALLOW_FE_FF 0x0800
339#define UTF8_WARN_FE_FF 0x1000
340
341#define UTF8_CHECK_ONLY 0x2000
342
343/* For backwards source compatibility. They do nothing, as the default now
344 * includes what they used to mean. The first one's meaning was to allow the
345 * just the single non-character 0xFFFF */
346#define UTF8_ALLOW_FFFF 0
347#define UTF8_ALLOW_SURROGATE 0
348
33d9abfb 349#define UTF8_DISALLOW_ILLEGAL_INTERCHANGE (UTF8_DISALLOW_SUPER|UTF8_DISALLOW_NONCHAR|UTF8_DISALLOW_SURROGATE|UTF8_DISALLOW_FE_FF)
949cf498 350#define UTF8_WARN_ILLEGAL_INTERCHANGE \
33d9abfb 351 (UTF8_WARN_SUPER|UTF8_WARN_NONCHAR|UTF8_WARN_SURROGATE|UTF8_WARN_FE_FF)
949cf498
KW
352#define UTF8_ALLOW_ANY \
353 (~(UTF8_DISALLOW_ILLEGAL_INTERCHANGE|UTF8_WARN_ILLEGAL_INTERCHANGE))
354#define UTF8_ALLOW_ANYUV \
355 (UTF8_ALLOW_EMPTY \
356 & ~(UTF8_DISALLOW_ILLEGAL_INTERCHANGE|UTF8_WARN_ILLEGAL_INTERCHANGE))
9f7f3913
ST
357#define UTF8_ALLOW_DEFAULT (ckWARN(WARN_UTF8) ? 0 : \
358 UTF8_ALLOW_ANYUV)
1d72bdf6 359
7131f24d
KW
360/* Surrogates, non-character code points and above-Unicode code points are
361 * problematic in some contexts. This allows code that needs to check for
362 * those to to quickly exclude the vast majority of code points it will
363 * encounter */
b96a92fb
KW
364#define UTF8_FIRST_PROBLEMATIC_CODE_POINT_FIRST_BYTE \
365 FIRST_SURROGATE_UTF8_FIRST_BYTE
7131f24d 366
b96a92fb
KW
367#define UTF8_IS_SURROGATE(s) cBOOL(is_SURROGATE_utf8(s))
368#define UTF8_IS_REPLACEMENT(s, send) cBOOL(is_REPLACEMENT_utf8_safe(s,send))
7131f24d
KW
369
370/* ASCII EBCDIC I8
371 * U+10FFFF: \xF4\x8F\xBF\xBF \xF9\xA1\xBF\xBF\xBF max legal Unicode
372 * U+110000: \xF4\x90\x80\x80 \xF9\xA2\xA0\xA0\xA0
373 * U+110001: \xF4\x90\x80\x81 \xF9\xA2\xA0\xA0\xA1
374 */
375#ifdef EBCDIC /* Both versions assume well-formed UTF8 */
0404c92a
KW
376# define UTF8_IS_SUPER(s) (NATIVE_TO_I8(* (U8*) (s)) >= 0xF9 \
377 && (NATIVE_TO_I8(* (U8*) (s)) > 0xF9 \
378 || (NATIVE_TO_I8(* (U8*) ((s)) + 1 >= 0xA2))))
7131f24d 379#else
0404c92a
KW
380# define UTF8_IS_SUPER(s) (*(U8*) (s) >= 0xF4 \
381 && (*(U8*) (s) > 0xF4 || (*((U8*) (s) + 1) >= 0x90)))
7131f24d
KW
382#endif
383
b96a92fb
KW
384/* These are now machine generated, and the 'given' clause is no longer
385 * applicable */
386#define UTF8_IS_NONCHAR_GIVEN_THAT_NON_SUPER_AND_GE_PROBLEMATIC(s) \
387 cBOOL(is_NONCHAR_utf8(s))
388#define UTF8_IS_NONCHAR_(s) \
389 UTF8_IS_NONCHAR_GIVEN_THAT_NON_SUPER_AND_GE_PROBLEMATIC(s)
7131f24d 390
c867b360
JH
391#define UNICODE_SURROGATE_FIRST 0xD800
392#define UNICODE_SURROGATE_LAST 0xDFFF
393#define UNICODE_REPLACEMENT 0xFFFD
394#define UNICODE_BYTE_ORDER_MARK 0xFEFF
1d72bdf6 395
b851fbc1 396/* Though our UTF-8 encoding can go beyond this,
c76687c5 397 * let's be conservative and do as Unicode says. */
b851fbc1
JH
398#define PERL_UNICODE_MAX 0x10FFFF
399
949cf498
KW
400#define UNICODE_WARN_SURROGATE 0x0001 /* UTF-16 surrogates */
401#define UNICODE_WARN_NONCHAR 0x0002 /* Non-char code points */
402#define UNICODE_WARN_SUPER 0x0004 /* Above 0x10FFFF */
403#define UNICODE_WARN_FE_FF 0x0008 /* Above 0x10FFFF */
404#define UNICODE_DISALLOW_SURROGATE 0x0010
405#define UNICODE_DISALLOW_NONCHAR 0x0020
406#define UNICODE_DISALLOW_SUPER 0x0040
407#define UNICODE_DISALLOW_FE_FF 0x0080
bb88be5f
KW
408#define UNICODE_WARN_ILLEGAL_INTERCHANGE \
409 (UNICODE_WARN_SURROGATE|UNICODE_WARN_NONCHAR|UNICODE_WARN_SUPER)
410#define UNICODE_DISALLOW_ILLEGAL_INTERCHANGE \
411 (UNICODE_DISALLOW_SURROGATE|UNICODE_DISALLOW_NONCHAR|UNICODE_DISALLOW_SUPER)
949cf498
KW
412
413/* For backward source compatibility, as are now the default */
414#define UNICODE_ALLOW_SURROGATE 0
415#define UNICODE_ALLOW_SUPER 0
416#define UNICODE_ALLOW_ANY 0
b851fbc1 417
1d72bdf6
NIS
418#define UNICODE_IS_SURROGATE(c) ((c) >= UNICODE_SURROGATE_FIRST && \
419 (c) <= UNICODE_SURROGATE_LAST)
a10ec373 420#define UNICODE_IS_REPLACEMENT(c) ((c) == UNICODE_REPLACEMENT)
872c91ae 421#define UNICODE_IS_BYTE_ORDER_MARK(c) ((c) == UNICODE_BYTE_ORDER_MARK)
7131f24d
KW
422#define UNICODE_IS_NONCHAR(c) ((c >= 0xFDD0 && c <= 0xFDEF) \
423 /* The other noncharacters end in FFFE or FFFF, which \
424 * the mask below catches both of, but beyond the last \
425 * official unicode code point, they aren't \
426 * noncharacters, since those aren't Unicode \
427 * characters at all */ \
428 || ((((c & 0xFFFE) == 0xFFFE)) && ! UNICODE_IS_SUPER(c)))
429#define UNICODE_IS_SUPER(c) ((c) > PERL_UNICODE_MAX)
430#define UNICODE_IS_FE_FF(c) ((c) > 0x7FFFFFFF)
1d72bdf6 431
f067b878
NA
432#ifdef HAS_QUAD
433# define UTF8_QUAD_MAX UINT64_C(0x1000000000)
434#endif
3bd709b1 435
ec34087a
KW
436#define LATIN_SMALL_LETTER_SHARP_S LATIN_SMALL_LETTER_SHARP_S_NATIVE
437#define LATIN_SMALL_LETTER_Y_WITH_DIAERESIS \
438 LATIN_SMALL_LETTER_Y_WITH_DIAERESIS_NATIVE
439#define MICRO_SIGN MICRO_SIGN_NATIVE
440#define LATIN_CAPITAL_LETTER_A_WITH_RING_ABOVE \
441 LATIN_CAPITAL_LETTER_A_WITH_RING_ABOVE_NATIVE
442#define LATIN_SMALL_LETTER_A_WITH_RING_ABOVE \
443 LATIN_SMALL_LETTER_A_WITH_RING_ABOVE_NATIVE
09091399
JH
444#define UNICODE_GREEK_CAPITAL_LETTER_SIGMA 0x03A3
445#define UNICODE_GREEK_SMALL_LETTER_FINAL_SIGMA 0x03C2
446#define UNICODE_GREEK_SMALL_LETTER_SIGMA 0x03C3
9dcbe121 447#define GREEK_SMALL_LETTER_MU 0x03BC
78a0d3cc
KW
448#define GREEK_CAPITAL_LETTER_MU 0x039C /* Upper and title case of MICRON */
449#define LATIN_CAPITAL_LETTER_Y_WITH_DIAERESIS 0x0178 /* Also is title case */
97298f37 450#define LATIN_CAPITAL_LETTER_SHARP_S 0x1E9E
8d64d87f
KW
451#define LATIN_SMALL_LETTER_LONG_S 0x017F
452#define KELVIN_SIGN 0x212A
453#define ANGSTROM_SIGN 0x212B
09091399 454
9e55ce06 455#define UNI_DISPLAY_ISPRINT 0x0001
c728cb41
JH
456#define UNI_DISPLAY_BACKSLASH 0x0002
457#define UNI_DISPLAY_QQ (UNI_DISPLAY_ISPRINT|UNI_DISPLAY_BACKSLASH)
458#define UNI_DISPLAY_REGEX (UNI_DISPLAY_ISPRINT|UNI_DISPLAY_BACKSLASH)
9e55ce06 459
5cd46e1f
KW
460#define ANYOF_FOLD_SHARP_S(node, input, end) \
461 (ANYOF_BITMAP_TEST(node, LATIN_SMALL_LETTER_SHARP_S) && \
137165a6 462 (ANYOF_NONBITMAP(node)) && \
39065660 463 (ANYOF_FLAGS(node) & ANYOF_LOC_NONBITMAP_FOLD) && \
07b6858f
JH
464 ((end) > (input) + 1) && \
465 toLOWER((input)[0]) == 's' && \
466 toLOWER((input)[1]) == 's')
ebc501f0 467#define SHARP_S_SKIP 2
3b0fc154 468
a4f7a67c
KW
469/* If you want to exclude surrogates, and beyond legal Unicode, see the blame
470 * log for earlier versions which gave details for these */
4d646140
KW
471
472#ifndef EBCDIC
473/* This was generated by regen/regcharclass.pl, and then moved here. The lines
474 * that generated it were then commented out. This was done solely because it
475 * takes on the order of 10 minutes to generate, and is never going to change.
476 * The EBCDIC equivalent hasn't been commented out in regcharclass.pl, so it
477 * should generate and run the correct stuff */
39a0f513
YO
478/*
479 UTF8_CHAR: Matches utf8 from 1 to 4 bytes
480
481 0x0 - 0x1FFFFF
482*/
4d646140
KW
483/*** GENERATED CODE ***/
484#define is_UTF8_CHAR_utf8_safe(s,e) \
485( ((e)-(s) > 3) ? \
486 ( ( ( ((U8*)s)[0] & 0x80 ) == 0x00 ) ? 1 \
487 : ( 0xC2 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0xDF ) ? \
488 ( ( ( ((U8*)s)[1] & 0xC0 ) == 0x80 ) ? 2 : 0 ) \
489 : ( 0xE0 == ((U8*)s)[0] ) ? \
490 ( ( ( ( ((U8*)s)[1] & 0xE0 ) == 0xA0 ) && ( ( ((U8*)s)[2] & 0xC0 ) == 0x80 ) ) ? 3 : 0 )\
491 : ( 0xE1 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0xEF ) ? \
492 ( ( ( ( ((U8*)s)[1] & 0xC0 ) == 0x80 ) && ( ( ((U8*)s)[2] & 0xC0 ) == 0x80 ) ) ? 3 : 0 )\
493 : ( 0xF0 == ((U8*)s)[0] ) ? \
494 ( ( ( ( 0x90 <= ((U8*)s)[1] && ((U8*)s)[1] <= 0xBF ) && ( ( ((U8*)s)[2] & 0xC0 ) == 0x80 ) ) && ( ( ((U8*)s)[3] & 0xC0 ) == 0x80 ) ) ? 4 : 0 )\
39a0f513 495 : ( ( ( ( 0xF1 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0xF7 ) && ( ( ((U8*)s)[1] & 0xC0 ) == 0x80 ) ) && ( ( ((U8*)s)[2] & 0xC0 ) == 0x80 ) ) && ( ( ((U8*)s)[3] & 0xC0 ) == 0x80 ) ) ? 4 : 0 )\
4d646140
KW
496: ((e)-(s) > 2) ? \
497 ( ( ( ((U8*)s)[0] & 0x80 ) == 0x00 ) ? 1 \
498 : ( 0xC2 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0xDF ) ? \
499 ( ( ( ((U8*)s)[1] & 0xC0 ) == 0x80 ) ? 2 : 0 ) \
500 : ( 0xE0 == ((U8*)s)[0] ) ? \
501 ( ( ( ( ((U8*)s)[1] & 0xE0 ) == 0xA0 ) && ( ( ((U8*)s)[2] & 0xC0 ) == 0x80 ) ) ? 3 : 0 )\
39a0f513 502 : ( ( ( 0xE1 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0xEF ) && ( ( ((U8*)s)[1] & 0xC0 ) == 0x80 ) ) && ( ( ((U8*)s)[2] & 0xC0 ) == 0x80 ) ) ? 3 : 0 )\
4d646140
KW
503: ((e)-(s) > 1) ? \
504 ( ( ( ((U8*)s)[0] & 0x80 ) == 0x00 ) ? 1 \
39a0f513 505 : ( ( 0xC2 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0xDF ) && ( ( ((U8*)s)[1] & 0xC0 ) == 0x80 ) ) ? 2 : 0 )\
4d646140
KW
506: ((e)-(s) > 0) ? \
507 ( ( ((U8*)s)[0] & 0x80 ) == 0x00 ) \
508: 0 )
509#endif
3b0fc154
JH
510
511/* IS_UTF8_CHAR(p) is strictly speaking wrong (not UTF-8) because it
512 * (1) allows UTF-8 encoded UTF-16 surrogates
513 * (2) it allows code points past U+10FFFF.
514 * The Perl_is_utf8_char() full "slow" code will handle the Perl
515 * "extended UTF-8". */
4d646140 516#define IS_UTF8_CHAR(p, n) (is_UTF8_CHAR_utf8_safe(p, (p) + (n)) == n)
2f454f11 517
4d646140
KW
518/* regen/regcharclass.pl generates is_UTF8_CHAR_utf8_safe() macros for up to
519 * these number of bytes. So this has to be coordinated with it */
520#ifdef EBCDIC
2f454f11 521# define IS_UTF8_CHAR_FAST(n) ((n) <= 5)
4d646140
KW
522#else
523# define IS_UTF8_CHAR_FAST(n) ((n) <= 4)
524#endif
e9a8c099 525
57f0e7e2
KW
526#endif /* H_UTF8 */
527
e9a8c099
MHM
528/*
529 * Local variables:
530 * c-indentation-style: bsd
531 * c-basic-offset: 4
14d04a33 532 * indent-tabs-mode: nil
e9a8c099
MHM
533 * End:
534 *
14d04a33 535 * ex: set ts=8 sts=4 sw=4 et:
e9a8c099 536 */