Commit | Line | Data |
---|---|---|
a0ed51b3 LW |
1 | /* utf8.c |
2 | * | |
af3babe4 NC |
3 | * Copyright (C) 2000, 2001, 2002, 2003, 2004, 2005 by Larry Wall and |
4 | * 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 | ||
11 | /* | |
12 | * 'What a fix!' said Sam. 'That's the one place in all the lands we've ever | |
13 | * heard of that we don't want to see any closer; and that's the one place | |
14 | * we're trying to get to! And that's just where we can't get, nohow.' | |
15 | * | |
16 | * 'Well do I understand your speech,' he answered in the same language; | |
17 | * 'yet few strangers do so. Why then do you not speak in the Common Tongue, | |
18 | * as is the custom in the West, if you wish to be answered?' | |
19 | * | |
20 | * ...the travellers perceived that the floor was paved with stones of many | |
21 | * hues; branching runes and strange devices intertwined beneath their feet. | |
22 | */ | |
23 | ||
24 | #include "EXTERN.h" | |
864dbfa3 | 25 | #define PERL_IN_UTF8_C |
a0ed51b3 LW |
26 | #include "perl.h" |
27 | ||
901b21bf JH |
28 | static char unees[] = "Malformed UTF-8 character (unexpected end of string)"; |
29 | ||
ccfc67b7 JH |
30 | /* |
31 | =head1 Unicode Support | |
a0ed51b3 | 32 | |
166f8a29 DM |
33 | This file contains various utility functions for manipulating UTF8-encoded |
34 | strings. For the uninitiated, this is a method of representing arbitrary | |
61296642 | 35 | Unicode characters as a variable number of bytes, in such a way that |
56da48f7 DM |
36 | characters in the ASCII range are unmodified, and a zero byte never appears |
37 | within non-zero characters. | |
166f8a29 | 38 | |
b851fbc1 | 39 | =for apidoc A|U8 *|uvuni_to_utf8_flags|U8 *d|UV uv|UV flags |
eebe1485 | 40 | |
1e54db1a | 41 | Adds the UTF-8 representation of the Unicode codepoint C<uv> to the end |
89ebb4a3 | 42 | of the string C<d>; C<d> should be have at least C<UTF8_MAXBYTES+1> free |
eebe1485 | 43 | bytes available. The return value is the pointer to the byte after the |
9041c2e3 | 44 | end of the new character. In other words, |
eebe1485 | 45 | |
b851fbc1 JH |
46 | d = uvuni_to_utf8_flags(d, uv, flags); |
47 | ||
48 | or, in most cases, | |
49 | ||
9041c2e3 | 50 | d = uvuni_to_utf8(d, uv); |
eebe1485 | 51 | |
b851fbc1 JH |
52 | (which is equivalent to) |
53 | ||
54 | d = uvuni_to_utf8_flags(d, uv, 0); | |
55 | ||
eebe1485 SC |
56 | is the recommended Unicode-aware way of saying |
57 | ||
58 | *(d++) = uv; | |
59 | ||
60 | =cut | |
61 | */ | |
62 | ||
dfe13c55 | 63 | U8 * |
b851fbc1 | 64 | Perl_uvuni_to_utf8_flags(pTHX_ U8 *d, UV uv, UV flags) |
a0ed51b3 | 65 | { |
62961d2e | 66 | if (ckWARN(WARN_UTF8)) { |
b851fbc1 JH |
67 | if (UNICODE_IS_SURROGATE(uv) && |
68 | !(flags & UNICODE_ALLOW_SURROGATE)) | |
9014280d | 69 | Perl_warner(aTHX_ packWARN(WARN_UTF8), "UTF-16 surrogate 0x%04"UVxf, uv); |
b851fbc1 JH |
70 | else if ( |
71 | ((uv >= 0xFDD0 && uv <= 0xFDEF && | |
72 | !(flags & UNICODE_ALLOW_FDD0)) | |
73 | || | |
c867b360 | 74 | ((uv & 0xFFFE) == 0xFFFE && /* Either FFFE or FFFF. */ |
b851fbc1 JH |
75 | !(flags & UNICODE_ALLOW_FFFF))) && |
76 | /* UNICODE_ALLOW_SUPER includes | |
2a20b9da | 77 | * FFFEs and FFFFs beyond 0x10FFFF. */ |
b851fbc1 JH |
78 | ((uv <= PERL_UNICODE_MAX) || |
79 | !(flags & UNICODE_ALLOW_SUPER)) | |
80 | ) | |
9014280d | 81 | Perl_warner(aTHX_ packWARN(WARN_UTF8), |
507b9800 JH |
82 | "Unicode character 0x%04"UVxf" is illegal", uv); |
83 | } | |
c4d5f83a | 84 | if (UNI_IS_INVARIANT(uv)) { |
eb160463 | 85 | *d++ = (U8)UTF_TO_NATIVE(uv); |
a0ed51b3 LW |
86 | return d; |
87 | } | |
2d331972 | 88 | #if defined(EBCDIC) |
1d72bdf6 NIS |
89 | else { |
90 | STRLEN len = UNISKIP(uv); | |
91 | U8 *p = d+len-1; | |
92 | while (p > d) { | |
eb160463 | 93 | *p-- = (U8)UTF_TO_NATIVE((uv & UTF_CONTINUATION_MASK) | UTF_CONTINUATION_MARK); |
1d72bdf6 NIS |
94 | uv >>= UTF_ACCUMULATION_SHIFT; |
95 | } | |
eb160463 | 96 | *p = (U8)UTF_TO_NATIVE((uv & UTF_START_MASK(len)) | UTF_START_MARK(len)); |
1d72bdf6 NIS |
97 | return d+len; |
98 | } | |
99 | #else /* Non loop style */ | |
a0ed51b3 | 100 | if (uv < 0x800) { |
eb160463 GS |
101 | *d++ = (U8)(( uv >> 6) | 0xc0); |
102 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
103 | return d; |
104 | } | |
105 | if (uv < 0x10000) { | |
eb160463 GS |
106 | *d++ = (U8)(( uv >> 12) | 0xe0); |
107 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); | |
108 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
109 | return d; |
110 | } | |
111 | if (uv < 0x200000) { | |
eb160463 GS |
112 | *d++ = (U8)(( uv >> 18) | 0xf0); |
113 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); | |
114 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); | |
115 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
116 | return d; |
117 | } | |
118 | if (uv < 0x4000000) { | |
eb160463 GS |
119 | *d++ = (U8)(( uv >> 24) | 0xf8); |
120 | *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80); | |
121 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); | |
122 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); | |
123 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
124 | return d; |
125 | } | |
126 | if (uv < 0x80000000) { | |
eb160463 GS |
127 | *d++ = (U8)(( uv >> 30) | 0xfc); |
128 | *d++ = (U8)(((uv >> 24) & 0x3f) | 0x80); | |
129 | *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80); | |
130 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); | |
131 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); | |
132 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
133 | return d; |
134 | } | |
6b8eaf93 | 135 | #ifdef HAS_QUAD |
d7578b48 | 136 | if (uv < UTF8_QUAD_MAX) |
a0ed51b3 LW |
137 | #endif |
138 | { | |
eb160463 GS |
139 | *d++ = 0xfe; /* Can't match U+FEFF! */ |
140 | *d++ = (U8)(((uv >> 30) & 0x3f) | 0x80); | |
141 | *d++ = (U8)(((uv >> 24) & 0x3f) | 0x80); | |
142 | *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80); | |
143 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); | |
144 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); | |
145 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
146 | return d; |
147 | } | |
6b8eaf93 | 148 | #ifdef HAS_QUAD |
a0ed51b3 | 149 | { |
eb160463 GS |
150 | *d++ = 0xff; /* Can't match U+FFFE! */ |
151 | *d++ = 0x80; /* 6 Reserved bits */ | |
152 | *d++ = (U8)(((uv >> 60) & 0x0f) | 0x80); /* 2 Reserved bits */ | |
153 | *d++ = (U8)(((uv >> 54) & 0x3f) | 0x80); | |
154 | *d++ = (U8)(((uv >> 48) & 0x3f) | 0x80); | |
155 | *d++ = (U8)(((uv >> 42) & 0x3f) | 0x80); | |
156 | *d++ = (U8)(((uv >> 36) & 0x3f) | 0x80); | |
157 | *d++ = (U8)(((uv >> 30) & 0x3f) | 0x80); | |
158 | *d++ = (U8)(((uv >> 24) & 0x3f) | 0x80); | |
159 | *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80); | |
160 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); | |
161 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); | |
162 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
163 | return d; |
164 | } | |
165 | #endif | |
1d72bdf6 | 166 | #endif /* Loop style */ |
a0ed51b3 | 167 | } |
b851fbc1 JH |
168 | |
169 | U8 * | |
170 | Perl_uvuni_to_utf8(pTHX_ U8 *d, UV uv) | |
171 | { | |
172 | return Perl_uvuni_to_utf8_flags(aTHX_ d, uv, 0); | |
173 | } | |
9041c2e3 NIS |
174 | |
175 | ||
176 | /* | |
eebe1485 SC |
177 | =for apidoc A|STRLEN|is_utf8_char|U8 *s |
178 | ||
5da9da9e | 179 | Tests if some arbitrary number of bytes begins in a valid UTF-8 |
82686b01 JH |
180 | character. Note that an INVARIANT (i.e. ASCII) character is a valid |
181 | UTF-8 character. The actual number of bytes in the UTF-8 character | |
182 | will be returned if it is valid, otherwise 0. | |
9041c2e3 | 183 | |
82686b01 | 184 | =cut */ |
067a85ef | 185 | STRLEN |
386d01d6 GS |
186 | Perl_is_utf8_char(pTHX_ U8 *s) |
187 | { | |
188 | U8 u = *s; | |
067a85ef A |
189 | STRLEN slen, len; |
190 | UV uv, ouv; | |
386d01d6 | 191 | |
1d72bdf6 | 192 | if (UTF8_IS_INVARIANT(u)) |
386d01d6 GS |
193 | return 1; |
194 | ||
60006e79 | 195 | if (!UTF8_IS_START(u)) |
386d01d6 GS |
196 | return 0; |
197 | ||
9f07fdcd | 198 | len = UTF8SKIP(s); |
386d01d6 | 199 | |
60006e79 | 200 | if (len < 2 || !UTF8_IS_CONTINUATION(s[1])) |
067a85ef A |
201 | return 0; |
202 | ||
386d01d6 GS |
203 | slen = len - 1; |
204 | s++; | |
6eb6869e | 205 | u &= UTF_START_MASK(len); |
1d72bdf6 | 206 | uv = u; |
067a85ef | 207 | ouv = uv; |
386d01d6 | 208 | while (slen--) { |
60006e79 | 209 | if (!UTF8_IS_CONTINUATION(*s)) |
386d01d6 | 210 | return 0; |
8850bf83 | 211 | uv = UTF8_ACCUMULATE(uv, *s); |
209a9bc1 | 212 | if (uv < ouv) |
067a85ef A |
213 | return 0; |
214 | ouv = uv; | |
386d01d6 GS |
215 | s++; |
216 | } | |
067a85ef | 217 | |
eb160463 | 218 | if ((STRLEN)UNISKIP(uv) < len) |
067a85ef A |
219 | return 0; |
220 | ||
386d01d6 GS |
221 | return len; |
222 | } | |
223 | ||
6662521e | 224 | /* |
eebe1485 | 225 | =for apidoc A|bool|is_utf8_string|U8 *s|STRLEN len |
6662521e | 226 | |
c9ada85f | 227 | Returns true if first C<len> bytes of the given string form a valid |
1e54db1a JH |
228 | UTF-8 string, false otherwise. Note that 'a valid UTF-8 string' does |
229 | not mean 'a string that contains code points above 0x7F encoded in UTF-8' | |
230 | because a valid ASCII string is a valid UTF-8 string. | |
6662521e GS |
231 | |
232 | =cut | |
233 | */ | |
234 | ||
8e84507e | 235 | bool |
6662521e GS |
236 | Perl_is_utf8_string(pTHX_ U8 *s, STRLEN len) |
237 | { | |
067a85ef | 238 | U8* x = s; |
1aa99e6b | 239 | U8* send; |
067a85ef A |
240 | STRLEN c; |
241 | ||
61468b03 | 242 | if (!len && s) |
6cd5fe39 | 243 | len = strlen((char *)s); |
1aa99e6b IH |
244 | send = s + len; |
245 | ||
6662521e | 246 | while (x < send) { |
1acdb0da JH |
247 | /* Inline the easy bits of is_utf8_char() here for speed... */ |
248 | if (UTF8_IS_INVARIANT(*x)) | |
249 | c = 1; | |
250 | else if (!UTF8_IS_START(*x)) | |
251 | return FALSE; | |
252 | else { | |
253 | /* ... and call is_utf8_char() only if really needed. */ | |
254 | c = is_utf8_char(x); | |
255 | if (!c) | |
256 | return FALSE; | |
257 | } | |
6662521e | 258 | x += c; |
6662521e | 259 | } |
60006e79 JH |
260 | if (x != send) |
261 | return FALSE; | |
067a85ef A |
262 | |
263 | return TRUE; | |
6662521e GS |
264 | } |
265 | ||
67e989fb | 266 | /* |
81cd54e3 JH |
267 | =for apidoc A|bool|is_utf8_string_loc|U8 *s|STRLEN len|U8 **p |
268 | ||
269 | Like is_ut8_string but store the location of the failure in | |
270 | the last argument. | |
271 | ||
272 | =cut | |
273 | */ | |
274 | ||
275 | bool | |
276 | Perl_is_utf8_string_loc(pTHX_ U8 *s, STRLEN len, U8 **p) | |
277 | { | |
278 | U8* x = s; | |
279 | U8* send; | |
280 | STRLEN c; | |
281 | ||
61468b03 | 282 | if (!len && s) |
81cd54e3 JH |
283 | len = strlen((char *)s); |
284 | send = s + len; | |
285 | ||
286 | while (x < send) { | |
287 | /* Inline the easy bits of is_utf8_char() here for speed... */ | |
288 | if (UTF8_IS_INVARIANT(*x)) | |
289 | c = 1; | |
290 | else if (!UTF8_IS_START(*x)) { | |
291 | if (p) | |
292 | *p = x; | |
293 | return FALSE; | |
294 | } | |
295 | else { | |
296 | /* ... and call is_utf8_char() only if really needed. */ | |
297 | c = is_utf8_char(x); | |
298 | if (!c) { | |
299 | if (p) | |
300 | *p = x; | |
301 | return FALSE; | |
302 | } | |
303 | } | |
304 | x += c; | |
305 | } | |
306 | if (x != send) { | |
307 | if (p) | |
308 | *p = x; | |
309 | return FALSE; | |
310 | } | |
311 | ||
312 | return TRUE; | |
313 | } | |
314 | ||
315 | /* | |
9041c2e3 | 316 | =for apidoc A|UV|utf8n_to_uvuni|U8 *s|STRLEN curlen|STRLEN *retlen|U32 flags |
67e989fb | 317 | |
9041c2e3 NIS |
318 | Bottom level UTF-8 decode routine. |
319 | Returns the unicode code point value of the first character in the string C<s> | |
1e54db1a | 320 | which is assumed to be in UTF-8 encoding and no longer than C<curlen>; |
7df053ec | 321 | C<retlen> will be set to the length, in bytes, of that character. |
67e989fb | 322 | |
1e54db1a | 323 | If C<s> does not point to a well-formed UTF-8 character, the behaviour |
dcad2880 JH |
324 | is dependent on the value of C<flags>: if it contains UTF8_CHECK_ONLY, |
325 | it is assumed that the caller will raise a warning, and this function | |
28d3d195 JH |
326 | will silently just set C<retlen> to C<-1> and return zero. If the |
327 | C<flags> does not contain UTF8_CHECK_ONLY, warnings about | |
328 | malformations will be given, C<retlen> will be set to the expected | |
329 | length of the UTF-8 character in bytes, and zero will be returned. | |
330 | ||
331 | The C<flags> can also contain various flags to allow deviations from | |
332 | the strict UTF-8 encoding (see F<utf8.h>). | |
67e989fb | 333 | |
9041c2e3 NIS |
334 | Most code should use utf8_to_uvchr() rather than call this directly. |
335 | ||
37607a96 PK |
336 | =cut |
337 | */ | |
67e989fb | 338 | |
a0ed51b3 | 339 | UV |
37607a96 | 340 | Perl_utf8n_to_uvuni(pTHX_ U8 *s, STRLEN curlen, STRLEN *retlen, U32 flags) |
a0ed51b3 | 341 | { |
097fb8e2 | 342 | U8 *s0 = s; |
9c5ffd7c | 343 | UV uv = *s, ouv = 0; |
ba210ebe JH |
344 | STRLEN len = 1; |
345 | bool dowarn = ckWARN_d(WARN_UTF8); | |
4f849cb6 | 346 | UV startbyte = *s; |
ba210ebe | 347 | STRLEN expectlen = 0; |
a0dbb045 JH |
348 | U32 warning = 0; |
349 | ||
350 | /* This list is a superset of the UTF8_ALLOW_XXX. */ | |
351 | ||
352 | #define UTF8_WARN_EMPTY 1 | |
353 | #define UTF8_WARN_CONTINUATION 2 | |
354 | #define UTF8_WARN_NON_CONTINUATION 3 | |
355 | #define UTF8_WARN_FE_FF 4 | |
356 | #define UTF8_WARN_SHORT 5 | |
357 | #define UTF8_WARN_OVERFLOW 6 | |
358 | #define UTF8_WARN_SURROGATE 7 | |
c867b360 JH |
359 | #define UTF8_WARN_LONG 8 |
360 | #define UTF8_WARN_FFFF 9 /* Also FFFE. */ | |
a0dbb045 JH |
361 | |
362 | if (curlen == 0 && | |
363 | !(flags & UTF8_ALLOW_EMPTY)) { | |
364 | warning = UTF8_WARN_EMPTY; | |
0c443dc2 JH |
365 | goto malformed; |
366 | } | |
367 | ||
1d72bdf6 | 368 | if (UTF8_IS_INVARIANT(uv)) { |
a0ed51b3 LW |
369 | if (retlen) |
370 | *retlen = 1; | |
c4d5f83a | 371 | return (UV) (NATIVE_TO_UTF(*s)); |
a0ed51b3 | 372 | } |
67e989fb | 373 | |
421a8bf2 | 374 | if (UTF8_IS_CONTINUATION(uv) && |
fcc8fcf6 | 375 | !(flags & UTF8_ALLOW_CONTINUATION)) { |
a0dbb045 | 376 | warning = UTF8_WARN_CONTINUATION; |
ba210ebe JH |
377 | goto malformed; |
378 | } | |
379 | ||
421a8bf2 | 380 | if (UTF8_IS_START(uv) && curlen > 1 && !UTF8_IS_CONTINUATION(s[1]) && |
fcc8fcf6 | 381 | !(flags & UTF8_ALLOW_NON_CONTINUATION)) { |
a0dbb045 | 382 | warning = UTF8_WARN_NON_CONTINUATION; |
ba210ebe JH |
383 | goto malformed; |
384 | } | |
9041c2e3 | 385 | |
1d72bdf6 | 386 | #ifdef EBCDIC |
75383841 | 387 | uv = NATIVE_TO_UTF(uv); |
1d72bdf6 | 388 | #else |
fcc8fcf6 JH |
389 | if ((uv == 0xfe || uv == 0xff) && |
390 | !(flags & UTF8_ALLOW_FE_FF)) { | |
a0dbb045 | 391 | warning = UTF8_WARN_FE_FF; |
ba210ebe | 392 | goto malformed; |
a0ed51b3 | 393 | } |
1d72bdf6 NIS |
394 | #endif |
395 | ||
ba210ebe JH |
396 | if (!(uv & 0x20)) { len = 2; uv &= 0x1f; } |
397 | else if (!(uv & 0x10)) { len = 3; uv &= 0x0f; } | |
398 | else if (!(uv & 0x08)) { len = 4; uv &= 0x07; } | |
399 | else if (!(uv & 0x04)) { len = 5; uv &= 0x03; } | |
1d72bdf6 NIS |
400 | #ifdef EBCDIC |
401 | else if (!(uv & 0x02)) { len = 6; uv &= 0x01; } | |
402 | else { len = 7; uv &= 0x01; } | |
403 | #else | |
ba210ebe JH |
404 | else if (!(uv & 0x02)) { len = 6; uv &= 0x01; } |
405 | else if (!(uv & 0x01)) { len = 7; uv = 0; } | |
1d72bdf6 NIS |
406 | else { len = 13; uv = 0; } /* whoa! */ |
407 | #endif | |
408 | ||
a0ed51b3 LW |
409 | if (retlen) |
410 | *retlen = len; | |
9041c2e3 | 411 | |
ba210ebe JH |
412 | expectlen = len; |
413 | ||
fcc8fcf6 JH |
414 | if ((curlen < expectlen) && |
415 | !(flags & UTF8_ALLOW_SHORT)) { | |
a0dbb045 | 416 | warning = UTF8_WARN_SHORT; |
ba210ebe JH |
417 | goto malformed; |
418 | } | |
419 | ||
420 | len--; | |
a0ed51b3 | 421 | s++; |
ba210ebe JH |
422 | ouv = uv; |
423 | ||
a0ed51b3 | 424 | while (len--) { |
421a8bf2 JH |
425 | if (!UTF8_IS_CONTINUATION(*s) && |
426 | !(flags & UTF8_ALLOW_NON_CONTINUATION)) { | |
a0dbb045 JH |
427 | s--; |
428 | warning = UTF8_WARN_NON_CONTINUATION; | |
ba210ebe | 429 | goto malformed; |
a0ed51b3 LW |
430 | } |
431 | else | |
8850bf83 | 432 | uv = UTF8_ACCUMULATE(uv, *s); |
a0dbb045 JH |
433 | if (!(uv > ouv)) { |
434 | /* These cannot be allowed. */ | |
435 | if (uv == ouv) { | |
75dbc644 | 436 | if (expectlen != 13 && !(flags & UTF8_ALLOW_LONG)) { |
a0dbb045 JH |
437 | warning = UTF8_WARN_LONG; |
438 | goto malformed; | |
439 | } | |
440 | } | |
441 | else { /* uv < ouv */ | |
442 | /* This cannot be allowed. */ | |
443 | warning = UTF8_WARN_OVERFLOW; | |
444 | goto malformed; | |
445 | } | |
ba210ebe JH |
446 | } |
447 | s++; | |
448 | ouv = uv; | |
449 | } | |
450 | ||
421a8bf2 | 451 | if (UNICODE_IS_SURROGATE(uv) && |
fcc8fcf6 | 452 | !(flags & UTF8_ALLOW_SURROGATE)) { |
a0dbb045 | 453 | warning = UTF8_WARN_SURROGATE; |
ba210ebe | 454 | goto malformed; |
eb160463 | 455 | } else if ((expectlen > (STRLEN)UNISKIP(uv)) && |
fcc8fcf6 | 456 | !(flags & UTF8_ALLOW_LONG)) { |
a0dbb045 | 457 | warning = UTF8_WARN_LONG; |
ba210ebe | 458 | goto malformed; |
421a8bf2 | 459 | } else if (UNICODE_IS_ILLEGAL(uv) && |
a9917092 | 460 | !(flags & UTF8_ALLOW_FFFF)) { |
a0dbb045 | 461 | warning = UTF8_WARN_FFFF; |
a9917092 | 462 | goto malformed; |
a0ed51b3 | 463 | } |
ba210ebe | 464 | |
a0ed51b3 | 465 | return uv; |
ba210ebe JH |
466 | |
467 | malformed: | |
468 | ||
fcc8fcf6 | 469 | if (flags & UTF8_CHECK_ONLY) { |
ba210ebe | 470 | if (retlen) |
cc366d4b | 471 | *retlen = -1; |
ba210ebe JH |
472 | return 0; |
473 | } | |
474 | ||
a0dbb045 JH |
475 | if (dowarn) { |
476 | SV* sv = sv_2mortal(newSVpv("Malformed UTF-8 character ", 0)); | |
477 | ||
478 | switch (warning) { | |
479 | case 0: /* Intentionally empty. */ break; | |
480 | case UTF8_WARN_EMPTY: | |
54667de8 | 481 | Perl_sv_catpv(aTHX_ sv, "(empty string)"); |
a0dbb045 JH |
482 | break; |
483 | case UTF8_WARN_CONTINUATION: | |
097fb8e2 | 484 | Perl_sv_catpvf(aTHX_ sv, "(unexpected continuation byte 0x%02"UVxf", with no preceding start byte)", uv); |
a0dbb045 JH |
485 | break; |
486 | case UTF8_WARN_NON_CONTINUATION: | |
097fb8e2 JH |
487 | if (s == s0) |
488 | Perl_sv_catpvf(aTHX_ sv, "(unexpected non-continuation byte 0x%02"UVxf", immediately after start byte 0x%02"UVxf")", | |
489 | (UV)s[1], startbyte); | |
490 | else | |
491 | Perl_sv_catpvf(aTHX_ sv, "(unexpected non-continuation byte 0x%02"UVxf", %d byte%s after start byte 0x%02"UVxf", expected %d bytes)", | |
492 | (UV)s[1], s - s0, s - s0 > 1 ? "s" : "", startbyte, expectlen); | |
493 | ||
a0dbb045 JH |
494 | break; |
495 | case UTF8_WARN_FE_FF: | |
496 | Perl_sv_catpvf(aTHX_ sv, "(byte 0x%02"UVxf")", uv); | |
497 | break; | |
498 | case UTF8_WARN_SHORT: | |
097fb8e2 JH |
499 | Perl_sv_catpvf(aTHX_ sv, "(%d byte%s, need %d, after start byte 0x%02"UVxf")", |
500 | curlen, curlen == 1 ? "" : "s", expectlen, startbyte); | |
b31f83c2 | 501 | expectlen = curlen; /* distance for caller to skip */ |
a0dbb045 JH |
502 | break; |
503 | case UTF8_WARN_OVERFLOW: | |
097fb8e2 JH |
504 | Perl_sv_catpvf(aTHX_ sv, "(overflow at 0x%"UVxf", byte 0x%02x, after start byte 0x%02"UVxf")", |
505 | ouv, *s, startbyte); | |
a0dbb045 JH |
506 | break; |
507 | case UTF8_WARN_SURROGATE: | |
508 | Perl_sv_catpvf(aTHX_ sv, "(UTF-16 surrogate 0x%04"UVxf")", uv); | |
509 | break; | |
a0dbb045 | 510 | case UTF8_WARN_LONG: |
097fb8e2 JH |
511 | Perl_sv_catpvf(aTHX_ sv, "(%d byte%s, need %d, after start byte 0x%02"UVxf")", |
512 | expectlen, expectlen == 1 ? "": "s", UNISKIP(uv), startbyte); | |
a0dbb045 JH |
513 | break; |
514 | case UTF8_WARN_FFFF: | |
515 | Perl_sv_catpvf(aTHX_ sv, "(character 0x%04"UVxf")", uv); | |
516 | break; | |
517 | default: | |
54667de8 | 518 | Perl_sv_catpv(aTHX_ sv, "(unknown reason)"); |
a0dbb045 JH |
519 | break; |
520 | } | |
521 | ||
522 | if (warning) { | |
523 | char *s = SvPVX(sv); | |
524 | ||
525 | if (PL_op) | |
9014280d | 526 | Perl_warner(aTHX_ packWARN(WARN_UTF8), |
53e06cf0 | 527 | "%s in %s", s, OP_DESC(PL_op)); |
a0dbb045 | 528 | else |
9014280d | 529 | Perl_warner(aTHX_ packWARN(WARN_UTF8), "%s", s); |
a0dbb045 JH |
530 | } |
531 | } | |
532 | ||
ba210ebe | 533 | if (retlen) |
28d3d195 | 534 | *retlen = expectlen ? expectlen : len; |
ba210ebe | 535 | |
28d3d195 | 536 | return 0; |
a0ed51b3 LW |
537 | } |
538 | ||
8e84507e | 539 | /* |
37607a96 | 540 | =for apidoc A|UV|utf8_to_uvchr|U8 *s|STRLEN *retlen |
9041c2e3 NIS |
541 | |
542 | Returns the native character value of the first character in the string C<s> | |
1e54db1a | 543 | which is assumed to be in UTF-8 encoding; C<retlen> will be set to the |
9041c2e3 NIS |
544 | length, in bytes, of that character. |
545 | ||
1e54db1a | 546 | If C<s> does not point to a well-formed UTF-8 character, zero is |
9041c2e3 NIS |
547 | returned and retlen is set, if possible, to -1. |
548 | ||
549 | =cut | |
550 | */ | |
551 | ||
552 | UV | |
37607a96 | 553 | Perl_utf8_to_uvchr(pTHX_ U8 *s, STRLEN *retlen) |
9041c2e3 | 554 | { |
89ebb4a3 | 555 | return Perl_utf8n_to_uvchr(aTHX_ s, UTF8_MAXBYTES, retlen, |
872c91ae | 556 | ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY); |
9041c2e3 NIS |
557 | } |
558 | ||
559 | /* | |
37607a96 | 560 | =for apidoc A|UV|utf8_to_uvuni|U8 *s|STRLEN *retlen |
9041c2e3 NIS |
561 | |
562 | Returns the Unicode code point of the first character in the string C<s> | |
1e54db1a | 563 | which is assumed to be in UTF-8 encoding; C<retlen> will be set to the |
9041c2e3 NIS |
564 | length, in bytes, of that character. |
565 | ||
566 | This function should only be used when returned UV is considered | |
567 | an index into the Unicode semantic tables (e.g. swashes). | |
568 | ||
1e54db1a | 569 | If C<s> does not point to a well-formed UTF-8 character, zero is |
ba210ebe | 570 | returned and retlen is set, if possible, to -1. |
8e84507e NIS |
571 | |
572 | =cut | |
573 | */ | |
574 | ||
575 | UV | |
37607a96 | 576 | Perl_utf8_to_uvuni(pTHX_ U8 *s, STRLEN *retlen) |
8e84507e | 577 | { |
9041c2e3 | 578 | /* Call the low level routine asking for checks */ |
89ebb4a3 | 579 | return Perl_utf8n_to_uvuni(aTHX_ s, UTF8_MAXBYTES, retlen, |
872c91ae | 580 | ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY); |
8e84507e NIS |
581 | } |
582 | ||
b76347f2 | 583 | /* |
37607a96 | 584 | =for apidoc A|STRLEN|utf8_length|U8 *s|U8 *e |
b76347f2 JH |
585 | |
586 | Return the length of the UTF-8 char encoded string C<s> in characters. | |
02eb7b47 JH |
587 | Stops at C<e> (inclusive). If C<e E<lt> s> or if the scan would end |
588 | up past C<e>, croaks. | |
b76347f2 JH |
589 | |
590 | =cut | |
591 | */ | |
592 | ||
593 | STRLEN | |
37607a96 | 594 | Perl_utf8_length(pTHX_ U8 *s, U8 *e) |
b76347f2 JH |
595 | { |
596 | STRLEN len = 0; | |
597 | ||
8850bf83 JH |
598 | /* Note: cannot use UTF8_IS_...() too eagerly here since e.g. |
599 | * the bitops (especially ~) can create illegal UTF-8. | |
600 | * In other words: in Perl UTF-8 is not just for Unicode. */ | |
601 | ||
901b21bf JH |
602 | if (e < s) { |
603 | if (ckWARN_d(WARN_UTF8)) { | |
604 | if (PL_op) | |
605 | Perl_warner(aTHX_ packWARN(WARN_UTF8), | |
606 | "%s in %s", unees, OP_DESC(PL_op)); | |
607 | else | |
608 | Perl_warner(aTHX_ packWARN(WARN_UTF8), unees); | |
609 | } | |
610 | return 0; | |
611 | } | |
b76347f2 | 612 | while (s < e) { |
02eb7b47 | 613 | U8 t = UTF8SKIP(s); |
b76347f2 | 614 | |
901b21bf JH |
615 | if (e - s < t) { |
616 | if (ckWARN_d(WARN_UTF8)) { | |
617 | if (PL_op) | |
618 | Perl_warner(aTHX_ packWARN(WARN_UTF8), | |
619 | unees, OP_DESC(PL_op)); | |
620 | else | |
621 | Perl_warner(aTHX_ packWARN(WARN_UTF8), unees); | |
622 | } | |
623 | return len; | |
624 | } | |
b76347f2 JH |
625 | s += t; |
626 | len++; | |
627 | } | |
628 | ||
629 | return len; | |
630 | } | |
631 | ||
b06226ff | 632 | /* |
eebe1485 | 633 | =for apidoc A|IV|utf8_distance|U8 *a|U8 *b |
b06226ff | 634 | |
1e54db1a | 635 | Returns the number of UTF-8 characters between the UTF-8 pointers C<a> |
b06226ff JH |
636 | and C<b>. |
637 | ||
638 | WARNING: use only if you *know* that the pointers point inside the | |
639 | same UTF-8 buffer. | |
640 | ||
37607a96 PK |
641 | =cut |
642 | */ | |
a0ed51b3 | 643 | |
02eb7b47 | 644 | IV |
864dbfa3 | 645 | Perl_utf8_distance(pTHX_ U8 *a, U8 *b) |
a0ed51b3 | 646 | { |
02eb7b47 JH |
647 | IV off = 0; |
648 | ||
8850bf83 JH |
649 | /* Note: cannot use UTF8_IS_...() too eagerly here since e.g. |
650 | * the bitops (especially ~) can create illegal UTF-8. | |
651 | * In other words: in Perl UTF-8 is not just for Unicode. */ | |
652 | ||
a0ed51b3 LW |
653 | if (a < b) { |
654 | while (a < b) { | |
02eb7b47 JH |
655 | U8 c = UTF8SKIP(a); |
656 | ||
901b21bf JH |
657 | if (b - a < c) { |
658 | if (ckWARN_d(WARN_UTF8)) { | |
659 | if (PL_op) | |
660 | Perl_warner(aTHX_ packWARN(WARN_UTF8), | |
661 | "%s in %s", unees, OP_DESC(PL_op)); | |
662 | else | |
663 | Perl_warner(aTHX_ packWARN(WARN_UTF8), unees); | |
664 | } | |
665 | return off; | |
666 | } | |
02eb7b47 | 667 | a += c; |
a0ed51b3 LW |
668 | off--; |
669 | } | |
670 | } | |
671 | else { | |
672 | while (b < a) { | |
02eb7b47 JH |
673 | U8 c = UTF8SKIP(b); |
674 | ||
901b21bf JH |
675 | if (a - b < c) { |
676 | if (ckWARN_d(WARN_UTF8)) { | |
677 | if (PL_op) | |
678 | Perl_warner(aTHX_ packWARN(WARN_UTF8), | |
679 | "%s in %s", unees, OP_DESC(PL_op)); | |
680 | else | |
681 | Perl_warner(aTHX_ packWARN(WARN_UTF8), unees); | |
682 | } | |
683 | return off; | |
684 | } | |
02eb7b47 | 685 | b += c; |
a0ed51b3 LW |
686 | off++; |
687 | } | |
688 | } | |
02eb7b47 | 689 | |
a0ed51b3 LW |
690 | return off; |
691 | } | |
692 | ||
b06226ff | 693 | /* |
37607a96 | 694 | =for apidoc A|U8 *|utf8_hop|U8 *s|I32 off |
b06226ff | 695 | |
8850bf83 JH |
696 | Return the UTF-8 pointer C<s> displaced by C<off> characters, either |
697 | forward or backward. | |
b06226ff JH |
698 | |
699 | WARNING: do not use the following unless you *know* C<off> is within | |
8850bf83 JH |
700 | the UTF-8 data pointed to by C<s> *and* that on entry C<s> is aligned |
701 | on the first byte of character or just after the last byte of a character. | |
b06226ff | 702 | |
37607a96 PK |
703 | =cut |
704 | */ | |
a0ed51b3 LW |
705 | |
706 | U8 * | |
864dbfa3 | 707 | Perl_utf8_hop(pTHX_ U8 *s, I32 off) |
a0ed51b3 | 708 | { |
8850bf83 JH |
709 | /* Note: cannot use UTF8_IS_...() too eagerly here since e.g |
710 | * the bitops (especially ~) can create illegal UTF-8. | |
711 | * In other words: in Perl UTF-8 is not just for Unicode. */ | |
712 | ||
a0ed51b3 LW |
713 | if (off >= 0) { |
714 | while (off--) | |
715 | s += UTF8SKIP(s); | |
716 | } | |
717 | else { | |
718 | while (off++) { | |
719 | s--; | |
8850bf83 JH |
720 | while (UTF8_IS_CONTINUATION(*s)) |
721 | s--; | |
a0ed51b3 LW |
722 | } |
723 | } | |
724 | return s; | |
725 | } | |
726 | ||
6940069f | 727 | /* |
eebe1485 | 728 | =for apidoc A|U8 *|utf8_to_bytes|U8 *s|STRLEN *len |
6940069f | 729 | |
1e54db1a | 730 | Converts a string C<s> of length C<len> from UTF-8 into byte encoding. |
246fae53 MG |
731 | Unlike C<bytes_to_utf8>, this over-writes the original string, and |
732 | updates len to contain the new length. | |
67e989fb | 733 | Returns zero on failure, setting C<len> to -1. |
6940069f GS |
734 | |
735 | =cut | |
736 | */ | |
737 | ||
738 | U8 * | |
37607a96 | 739 | Perl_utf8_to_bytes(pTHX_ U8 *s, STRLEN *len) |
6940069f | 740 | { |
6940069f GS |
741 | U8 *send; |
742 | U8 *d; | |
dcad2880 | 743 | U8 *save = s; |
246fae53 | 744 | |
1e54db1a | 745 | /* ensure valid UTF-8 and chars < 256 before updating string */ |
dcad2880 JH |
746 | for (send = s + *len; s < send; ) { |
747 | U8 c = *s++; | |
748 | ||
1d72bdf6 NIS |
749 | if (!UTF8_IS_INVARIANT(c) && |
750 | (!UTF8_IS_DOWNGRADEABLE_START(c) || (s >= send) | |
751 | || !(c = *s++) || !UTF8_IS_CONTINUATION(c))) { | |
dcad2880 JH |
752 | *len = -1; |
753 | return 0; | |
754 | } | |
246fae53 | 755 | } |
dcad2880 JH |
756 | |
757 | d = s = save; | |
6940069f | 758 | while (s < send) { |
ed646e6e | 759 | STRLEN ulen; |
9041c2e3 | 760 | *d++ = (U8)utf8_to_uvchr(s, &ulen); |
ed646e6e | 761 | s += ulen; |
6940069f GS |
762 | } |
763 | *d = '\0'; | |
246fae53 | 764 | *len = d - save; |
6940069f GS |
765 | return save; |
766 | } | |
767 | ||
768 | /* | |
f9a63242 JH |
769 | =for apidoc A|U8 *|bytes_from_utf8|U8 *s|STRLEN *len|bool *is_utf8 |
770 | ||
1e54db1a | 771 | Converts a string C<s> of length C<len> from UTF-8 into byte encoding. |
f9a63242 | 772 | Unlike <utf8_to_bytes> but like C<bytes_to_utf8>, returns a pointer to |
ef9edfd0 JH |
773 | the newly-created string, and updates C<len> to contain the new |
774 | length. Returns the original string if no conversion occurs, C<len> | |
775 | is unchanged. Do nothing if C<is_utf8> points to 0. Sets C<is_utf8> to | |
776 | 0 if C<s> is converted or contains all 7bit characters. | |
f9a63242 | 777 | |
37607a96 PK |
778 | =cut |
779 | */ | |
f9a63242 JH |
780 | |
781 | U8 * | |
37607a96 | 782 | Perl_bytes_from_utf8(pTHX_ U8 *s, STRLEN *len, bool *is_utf8) |
f9a63242 | 783 | { |
f9a63242 JH |
784 | U8 *d; |
785 | U8 *start = s; | |
db42d148 | 786 | U8 *send; |
f9a63242 JH |
787 | I32 count = 0; |
788 | ||
789 | if (!*is_utf8) | |
790 | return start; | |
791 | ||
1e54db1a | 792 | /* ensure valid UTF-8 and chars < 256 before converting string */ |
f9a63242 JH |
793 | for (send = s + *len; s < send;) { |
794 | U8 c = *s++; | |
1d72bdf6 | 795 | if (!UTF8_IS_INVARIANT(c)) { |
db42d148 NIS |
796 | if (UTF8_IS_DOWNGRADEABLE_START(c) && s < send && |
797 | (c = *s++) && UTF8_IS_CONTINUATION(c)) | |
798 | count++; | |
799 | else | |
f9a63242 | 800 | return start; |
db42d148 | 801 | } |
f9a63242 JH |
802 | } |
803 | ||
804 | *is_utf8 = 0; | |
805 | ||
f9a63242 | 806 | Newz(801, d, (*len) - count + 1, U8); |
ef9edfd0 | 807 | s = start; start = d; |
f9a63242 JH |
808 | while (s < send) { |
809 | U8 c = *s++; | |
c4d5f83a NIS |
810 | if (!UTF8_IS_INVARIANT(c)) { |
811 | /* Then it is two-byte encoded */ | |
812 | c = UTF8_ACCUMULATE(NATIVE_TO_UTF(c), *s++); | |
813 | c = ASCII_TO_NATIVE(c); | |
814 | } | |
815 | *d++ = c; | |
f9a63242 JH |
816 | } |
817 | *d = '\0'; | |
818 | *len = d - start; | |
819 | return start; | |
820 | } | |
821 | ||
822 | /* | |
eebe1485 | 823 | =for apidoc A|U8 *|bytes_to_utf8|U8 *s|STRLEN *len |
6940069f | 824 | |
1e54db1a | 825 | Converts a string C<s> of length C<len> from ASCII into UTF-8 encoding. |
6662521e GS |
826 | Returns a pointer to the newly-created string, and sets C<len> to |
827 | reflect the new length. | |
6940069f | 828 | |
1e54db1a | 829 | If you want to convert to UTF-8 from other encodings than ASCII, |
c9ada85f JH |
830 | see sv_recode_to_utf8(). |
831 | ||
497711e7 | 832 | =cut |
6940069f GS |
833 | */ |
834 | ||
835 | U8* | |
37607a96 | 836 | Perl_bytes_to_utf8(pTHX_ U8 *s, STRLEN *len) |
6940069f | 837 | { |
6940069f GS |
838 | U8 *send; |
839 | U8 *d; | |
840 | U8 *dst; | |
6662521e | 841 | send = s + (*len); |
6940069f | 842 | |
6662521e | 843 | Newz(801, d, (*len) * 2 + 1, U8); |
6940069f GS |
844 | dst = d; |
845 | ||
846 | while (s < send) { | |
db42d148 | 847 | UV uv = NATIVE_TO_ASCII(*s++); |
c4d5f83a | 848 | if (UNI_IS_INVARIANT(uv)) |
eb160463 | 849 | *d++ = (U8)UTF_TO_NATIVE(uv); |
6940069f | 850 | else { |
eb160463 GS |
851 | *d++ = (U8)UTF8_EIGHT_BIT_HI(uv); |
852 | *d++ = (U8)UTF8_EIGHT_BIT_LO(uv); | |
6940069f GS |
853 | } |
854 | } | |
855 | *d = '\0'; | |
6662521e | 856 | *len = d-dst; |
6940069f GS |
857 | return dst; |
858 | } | |
859 | ||
a0ed51b3 | 860 | /* |
dea0fc0b | 861 | * Convert native (big-endian) or reversed (little-endian) UTF-16 to UTF-8. |
a0ed51b3 LW |
862 | * |
863 | * Destination must be pre-extended to 3/2 source. Do not use in-place. | |
864 | * We optimize for native, for obvious reasons. */ | |
865 | ||
866 | U8* | |
dea0fc0b | 867 | Perl_utf16_to_utf8(pTHX_ U8* p, U8* d, I32 bytelen, I32 *newlen) |
a0ed51b3 | 868 | { |
dea0fc0b JH |
869 | U8* pend; |
870 | U8* dstart = d; | |
871 | ||
1de9afcd RGS |
872 | if (bytelen == 1 && p[0] == 0) { /* Be understanding. */ |
873 | d[0] = 0; | |
874 | *newlen = 1; | |
875 | return d; | |
876 | } | |
877 | ||
dea0fc0b | 878 | if (bytelen & 1) |
014ead4b | 879 | Perl_croak(aTHX_ "panic: utf16_to_utf8: odd bytelen %"UVf, (UV)bytelen); |
dea0fc0b JH |
880 | |
881 | pend = p + bytelen; | |
882 | ||
a0ed51b3 | 883 | while (p < pend) { |
dea0fc0b JH |
884 | UV uv = (p[0] << 8) + p[1]; /* UTF-16BE */ |
885 | p += 2; | |
a0ed51b3 | 886 | if (uv < 0x80) { |
eb160463 | 887 | *d++ = (U8)uv; |
a0ed51b3 LW |
888 | continue; |
889 | } | |
890 | if (uv < 0x800) { | |
eb160463 GS |
891 | *d++ = (U8)(( uv >> 6) | 0xc0); |
892 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
893 | continue; |
894 | } | |
895 | if (uv >= 0xd800 && uv < 0xdbff) { /* surrogates */ | |
30f84f9e TD |
896 | UV low = (p[0] << 8) + p[1]; |
897 | p += 2; | |
dea0fc0b JH |
898 | if (low < 0xdc00 || low >= 0xdfff) |
899 | Perl_croak(aTHX_ "Malformed UTF-16 surrogate"); | |
a0ed51b3 LW |
900 | uv = ((uv - 0xd800) << 10) + (low - 0xdc00) + 0x10000; |
901 | } | |
902 | if (uv < 0x10000) { | |
eb160463 GS |
903 | *d++ = (U8)(( uv >> 12) | 0xe0); |
904 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); | |
905 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
906 | continue; |
907 | } | |
908 | else { | |
eb160463 GS |
909 | *d++ = (U8)(( uv >> 18) | 0xf0); |
910 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); | |
911 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); | |
912 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
913 | continue; |
914 | } | |
915 | } | |
dea0fc0b | 916 | *newlen = d - dstart; |
a0ed51b3 LW |
917 | return d; |
918 | } | |
919 | ||
920 | /* Note: this one is slightly destructive of the source. */ | |
921 | ||
922 | U8* | |
dea0fc0b | 923 | Perl_utf16_to_utf8_reversed(pTHX_ U8* p, U8* d, I32 bytelen, I32 *newlen) |
a0ed51b3 LW |
924 | { |
925 | U8* s = (U8*)p; | |
926 | U8* send = s + bytelen; | |
927 | while (s < send) { | |
928 | U8 tmp = s[0]; | |
929 | s[0] = s[1]; | |
930 | s[1] = tmp; | |
931 | s += 2; | |
932 | } | |
dea0fc0b | 933 | return utf16_to_utf8(p, d, bytelen, newlen); |
a0ed51b3 LW |
934 | } |
935 | ||
936 | /* for now these are all defined (inefficiently) in terms of the utf8 versions */ | |
937 | ||
938 | bool | |
84afefe6 | 939 | Perl_is_uni_alnum(pTHX_ UV c) |
a0ed51b3 | 940 | { |
89ebb4a3 | 941 | U8 tmpbuf[UTF8_MAXBYTES+1]; |
230880c1 | 942 | uvchr_to_utf8(tmpbuf, c); |
a0ed51b3 LW |
943 | return is_utf8_alnum(tmpbuf); |
944 | } | |
945 | ||
946 | bool | |
84afefe6 | 947 | Perl_is_uni_alnumc(pTHX_ UV c) |
b8c5462f | 948 | { |
89ebb4a3 | 949 | U8 tmpbuf[UTF8_MAXBYTES+1]; |
230880c1 | 950 | uvchr_to_utf8(tmpbuf, c); |
b8c5462f JH |
951 | return is_utf8_alnumc(tmpbuf); |
952 | } | |
953 | ||
954 | bool | |
84afefe6 | 955 | Perl_is_uni_idfirst(pTHX_ UV c) |
a0ed51b3 | 956 | { |
89ebb4a3 | 957 | U8 tmpbuf[UTF8_MAXBYTES+1]; |
230880c1 | 958 | uvchr_to_utf8(tmpbuf, c); |
a0ed51b3 LW |
959 | return is_utf8_idfirst(tmpbuf); |
960 | } | |
961 | ||
962 | bool | |
84afefe6 | 963 | Perl_is_uni_alpha(pTHX_ UV c) |
a0ed51b3 | 964 | { |
89ebb4a3 | 965 | U8 tmpbuf[UTF8_MAXBYTES+1]; |
230880c1 | 966 | uvchr_to_utf8(tmpbuf, c); |
a0ed51b3 LW |
967 | return is_utf8_alpha(tmpbuf); |
968 | } | |
969 | ||
970 | bool | |
84afefe6 | 971 | Perl_is_uni_ascii(pTHX_ UV c) |
4d61ec05 | 972 | { |
89ebb4a3 | 973 | U8 tmpbuf[UTF8_MAXBYTES+1]; |
230880c1 | 974 | uvchr_to_utf8(tmpbuf, c); |
4d61ec05 GS |
975 | return is_utf8_ascii(tmpbuf); |
976 | } | |
977 | ||
978 | bool | |
84afefe6 | 979 | Perl_is_uni_space(pTHX_ UV c) |
a0ed51b3 | 980 | { |
89ebb4a3 | 981 | U8 tmpbuf[UTF8_MAXBYTES+1]; |
230880c1 | 982 | uvchr_to_utf8(tmpbuf, c); |
a0ed51b3 LW |
983 | return is_utf8_space(tmpbuf); |
984 | } | |
985 | ||
986 | bool | |
84afefe6 | 987 | Perl_is_uni_digit(pTHX_ UV c) |
a0ed51b3 | 988 | { |
89ebb4a3 | 989 | U8 tmpbuf[UTF8_MAXBYTES+1]; |
230880c1 | 990 | uvchr_to_utf8(tmpbuf, c); |
a0ed51b3 LW |
991 | return is_utf8_digit(tmpbuf); |
992 | } | |
993 | ||
994 | bool | |
84afefe6 | 995 | Perl_is_uni_upper(pTHX_ UV c) |
a0ed51b3 | 996 | { |
89ebb4a3 | 997 | U8 tmpbuf[UTF8_MAXBYTES+1]; |
230880c1 | 998 | uvchr_to_utf8(tmpbuf, c); |
a0ed51b3 LW |
999 | return is_utf8_upper(tmpbuf); |
1000 | } | |
1001 | ||
1002 | bool | |
84afefe6 | 1003 | Perl_is_uni_lower(pTHX_ UV c) |
a0ed51b3 | 1004 | { |
89ebb4a3 | 1005 | U8 tmpbuf[UTF8_MAXBYTES+1]; |
230880c1 | 1006 | uvchr_to_utf8(tmpbuf, c); |
a0ed51b3 LW |
1007 | return is_utf8_lower(tmpbuf); |
1008 | } | |
1009 | ||
1010 | bool | |
84afefe6 | 1011 | Perl_is_uni_cntrl(pTHX_ UV c) |
b8c5462f | 1012 | { |
89ebb4a3 | 1013 | U8 tmpbuf[UTF8_MAXBYTES+1]; |
230880c1 | 1014 | uvchr_to_utf8(tmpbuf, c); |
b8c5462f JH |
1015 | return is_utf8_cntrl(tmpbuf); |
1016 | } | |
1017 | ||
1018 | bool | |
84afefe6 | 1019 | Perl_is_uni_graph(pTHX_ UV c) |
b8c5462f | 1020 | { |
89ebb4a3 | 1021 | U8 tmpbuf[UTF8_MAXBYTES+1]; |
230880c1 | 1022 | uvchr_to_utf8(tmpbuf, c); |
b8c5462f JH |
1023 | return is_utf8_graph(tmpbuf); |
1024 | } | |
1025 | ||
1026 | bool | |
84afefe6 | 1027 | Perl_is_uni_print(pTHX_ UV c) |
a0ed51b3 | 1028 | { |
89ebb4a3 | 1029 | U8 tmpbuf[UTF8_MAXBYTES+1]; |
230880c1 | 1030 | uvchr_to_utf8(tmpbuf, c); |
a0ed51b3 LW |
1031 | return is_utf8_print(tmpbuf); |
1032 | } | |
1033 | ||
b8c5462f | 1034 | bool |
84afefe6 | 1035 | Perl_is_uni_punct(pTHX_ UV c) |
b8c5462f | 1036 | { |
89ebb4a3 | 1037 | U8 tmpbuf[UTF8_MAXBYTES+1]; |
230880c1 | 1038 | uvchr_to_utf8(tmpbuf, c); |
b8c5462f JH |
1039 | return is_utf8_punct(tmpbuf); |
1040 | } | |
1041 | ||
4d61ec05 | 1042 | bool |
84afefe6 | 1043 | Perl_is_uni_xdigit(pTHX_ UV c) |
4d61ec05 | 1044 | { |
89ebb4a3 | 1045 | U8 tmpbuf[UTF8_MAXBYTES_CASE+1]; |
230880c1 | 1046 | uvchr_to_utf8(tmpbuf, c); |
4d61ec05 GS |
1047 | return is_utf8_xdigit(tmpbuf); |
1048 | } | |
1049 | ||
84afefe6 JH |
1050 | UV |
1051 | Perl_to_uni_upper(pTHX_ UV c, U8* p, STRLEN *lenp) | |
a0ed51b3 | 1052 | { |
0ebc6274 JH |
1053 | uvchr_to_utf8(p, c); |
1054 | return to_utf8_upper(p, p, lenp); | |
a0ed51b3 LW |
1055 | } |
1056 | ||
84afefe6 JH |
1057 | UV |
1058 | Perl_to_uni_title(pTHX_ UV c, U8* p, STRLEN *lenp) | |
a0ed51b3 | 1059 | { |
0ebc6274 JH |
1060 | uvchr_to_utf8(p, c); |
1061 | return to_utf8_title(p, p, lenp); | |
a0ed51b3 LW |
1062 | } |
1063 | ||
84afefe6 JH |
1064 | UV |
1065 | Perl_to_uni_lower(pTHX_ UV c, U8* p, STRLEN *lenp) | |
a0ed51b3 | 1066 | { |
0ebc6274 JH |
1067 | uvchr_to_utf8(p, c); |
1068 | return to_utf8_lower(p, p, lenp); | |
a0ed51b3 LW |
1069 | } |
1070 | ||
84afefe6 JH |
1071 | UV |
1072 | Perl_to_uni_fold(pTHX_ UV c, U8* p, STRLEN *lenp) | |
1073 | { | |
0ebc6274 JH |
1074 | uvchr_to_utf8(p, c); |
1075 | return to_utf8_fold(p, p, lenp); | |
84afefe6 JH |
1076 | } |
1077 | ||
a0ed51b3 LW |
1078 | /* for now these all assume no locale info available for Unicode > 255 */ |
1079 | ||
1080 | bool | |
84afefe6 | 1081 | Perl_is_uni_alnum_lc(pTHX_ UV c) |
a0ed51b3 LW |
1082 | { |
1083 | return is_uni_alnum(c); /* XXX no locale support yet */ | |
1084 | } | |
1085 | ||
1086 | bool | |
84afefe6 | 1087 | Perl_is_uni_alnumc_lc(pTHX_ UV c) |
b8c5462f JH |
1088 | { |
1089 | return is_uni_alnumc(c); /* XXX no locale support yet */ | |
1090 | } | |
1091 | ||
1092 | bool | |
84afefe6 | 1093 | Perl_is_uni_idfirst_lc(pTHX_ UV c) |
a0ed51b3 LW |
1094 | { |
1095 | return is_uni_idfirst(c); /* XXX no locale support yet */ | |
1096 | } | |
1097 | ||
1098 | bool | |
84afefe6 | 1099 | Perl_is_uni_alpha_lc(pTHX_ UV c) |
a0ed51b3 LW |
1100 | { |
1101 | return is_uni_alpha(c); /* XXX no locale support yet */ | |
1102 | } | |
1103 | ||
1104 | bool | |
84afefe6 | 1105 | Perl_is_uni_ascii_lc(pTHX_ UV c) |
4d61ec05 GS |
1106 | { |
1107 | return is_uni_ascii(c); /* XXX no locale support yet */ | |
1108 | } | |
1109 | ||
1110 | bool | |
84afefe6 | 1111 | Perl_is_uni_space_lc(pTHX_ UV c) |
a0ed51b3 LW |
1112 | { |
1113 | return is_uni_space(c); /* XXX no locale support yet */ | |
1114 | } | |
1115 | ||
1116 | bool | |
84afefe6 | 1117 | Perl_is_uni_digit_lc(pTHX_ UV c) |
a0ed51b3 LW |
1118 | { |
1119 | return is_uni_digit(c); /* XXX no locale support yet */ | |
1120 | } | |
1121 | ||
1122 | bool | |
84afefe6 | 1123 | Perl_is_uni_upper_lc(pTHX_ UV c) |
a0ed51b3 LW |
1124 | { |
1125 | return is_uni_upper(c); /* XXX no locale support yet */ | |
1126 | } | |
1127 | ||
1128 | bool | |
84afefe6 | 1129 | Perl_is_uni_lower_lc(pTHX_ UV c) |
a0ed51b3 LW |
1130 | { |
1131 | return is_uni_lower(c); /* XXX no locale support yet */ | |
1132 | } | |
1133 | ||
1134 | bool | |
84afefe6 | 1135 | Perl_is_uni_cntrl_lc(pTHX_ UV c) |
b8c5462f JH |
1136 | { |
1137 | return is_uni_cntrl(c); /* XXX no locale support yet */ | |
1138 | } | |
1139 | ||
1140 | bool | |
84afefe6 | 1141 | Perl_is_uni_graph_lc(pTHX_ UV c) |
b8c5462f JH |
1142 | { |
1143 | return is_uni_graph(c); /* XXX no locale support yet */ | |
1144 | } | |
1145 | ||
1146 | bool | |
84afefe6 | 1147 | Perl_is_uni_print_lc(pTHX_ UV c) |
a0ed51b3 LW |
1148 | { |
1149 | return is_uni_print(c); /* XXX no locale support yet */ | |
1150 | } | |
1151 | ||
b8c5462f | 1152 | bool |
84afefe6 | 1153 | Perl_is_uni_punct_lc(pTHX_ UV c) |
b8c5462f JH |
1154 | { |
1155 | return is_uni_punct(c); /* XXX no locale support yet */ | |
1156 | } | |
1157 | ||
4d61ec05 | 1158 | bool |
84afefe6 | 1159 | Perl_is_uni_xdigit_lc(pTHX_ UV c) |
4d61ec05 GS |
1160 | { |
1161 | return is_uni_xdigit(c); /* XXX no locale support yet */ | |
1162 | } | |
1163 | ||
b7ac61fa JH |
1164 | U32 |
1165 | Perl_to_uni_upper_lc(pTHX_ U32 c) | |
1166 | { | |
ee099d14 JH |
1167 | /* XXX returns only the first character -- do not use XXX */ |
1168 | /* XXX no locale support yet */ | |
1169 | STRLEN len; | |
89ebb4a3 | 1170 | U8 tmpbuf[UTF8_MAXBYTES_CASE+1]; |
ee099d14 | 1171 | return (U32)to_uni_upper(c, tmpbuf, &len); |
b7ac61fa JH |
1172 | } |
1173 | ||
1174 | U32 | |
1175 | Perl_to_uni_title_lc(pTHX_ U32 c) | |
1176 | { | |
ee099d14 JH |
1177 | /* XXX returns only the first character XXX -- do not use XXX */ |
1178 | /* XXX no locale support yet */ | |
1179 | STRLEN len; | |
89ebb4a3 | 1180 | U8 tmpbuf[UTF8_MAXBYTES_CASE+1]; |
ee099d14 | 1181 | return (U32)to_uni_title(c, tmpbuf, &len); |
b7ac61fa JH |
1182 | } |
1183 | ||
1184 | U32 | |
1185 | Perl_to_uni_lower_lc(pTHX_ U32 c) | |
1186 | { | |
ee099d14 JH |
1187 | /* XXX returns only the first character -- do not use XXX */ |
1188 | /* XXX no locale support yet */ | |
1189 | STRLEN len; | |
89ebb4a3 | 1190 | U8 tmpbuf[UTF8_MAXBYTES_CASE+1]; |
ee099d14 | 1191 | return (U32)to_uni_lower(c, tmpbuf, &len); |
b7ac61fa JH |
1192 | } |
1193 | ||
a0ed51b3 | 1194 | bool |
864dbfa3 | 1195 | Perl_is_utf8_alnum(pTHX_ U8 *p) |
a0ed51b3 | 1196 | { |
386d01d6 GS |
1197 | if (!is_utf8_char(p)) |
1198 | return FALSE; | |
a0ed51b3 | 1199 | if (!PL_utf8_alnum) |
289d4f09 ML |
1200 | /* NOTE: "IsWord", not "IsAlnum", since Alnum is a true |
1201 | * descendant of isalnum(3), in other words, it doesn't | |
1202 | * contain the '_'. --jhi */ | |
1203 | PL_utf8_alnum = swash_init("utf8", "IsWord", &PL_sv_undef, 0, 0); | |
eb160463 | 1204 | return swash_fetch(PL_utf8_alnum, p, TRUE) != 0; |
a0ed51b3 LW |
1205 | /* return *p == '_' || is_utf8_alpha(p) || is_utf8_digit(p); */ |
1206 | #ifdef SURPRISINGLY_SLOWER /* probably because alpha is usually true */ | |
1207 | if (!PL_utf8_alnum) | |
1208 | PL_utf8_alnum = swash_init("utf8", "", | |
1209 | sv_2mortal(newSVpv("+utf8::IsAlpha\n+utf8::IsDigit\n005F\n",0)), 0, 0); | |
eb160463 | 1210 | return swash_fetch(PL_utf8_alnum, p, TRUE) != 0; |
a0ed51b3 LW |
1211 | #endif |
1212 | } | |
1213 | ||
1214 | bool | |
b8c5462f JH |
1215 | Perl_is_utf8_alnumc(pTHX_ U8 *p) |
1216 | { | |
386d01d6 GS |
1217 | if (!is_utf8_char(p)) |
1218 | return FALSE; | |
b8c5462f JH |
1219 | if (!PL_utf8_alnum) |
1220 | PL_utf8_alnum = swash_init("utf8", "IsAlnumC", &PL_sv_undef, 0, 0); | |
eb160463 | 1221 | return swash_fetch(PL_utf8_alnum, p, TRUE) != 0; |
b8c5462f JH |
1222 | /* return is_utf8_alpha(p) || is_utf8_digit(p); */ |
1223 | #ifdef SURPRISINGLY_SLOWER /* probably because alpha is usually true */ | |
1224 | if (!PL_utf8_alnum) | |
1225 | PL_utf8_alnum = swash_init("utf8", "", | |
1226 | sv_2mortal(newSVpv("+utf8::IsAlpha\n+utf8::IsDigit\n005F\n",0)), 0, 0); | |
eb160463 | 1227 | return swash_fetch(PL_utf8_alnum, p, TRUE) != 0; |
b8c5462f JH |
1228 | #endif |
1229 | } | |
1230 | ||
1231 | bool | |
82686b01 | 1232 | Perl_is_utf8_idfirst(pTHX_ U8 *p) /* The naming is historical. */ |
a0ed51b3 | 1233 | { |
82686b01 JH |
1234 | if (*p == '_') |
1235 | return TRUE; | |
1236 | if (!is_utf8_char(p)) | |
1237 | return FALSE; | |
1238 | if (!PL_utf8_idstart) /* is_utf8_idstart would be more logical. */ | |
1239 | PL_utf8_idstart = swash_init("utf8", "IdStart", &PL_sv_undef, 0, 0); | |
eb160463 | 1240 | return swash_fetch(PL_utf8_idstart, p, TRUE) != 0; |
82686b01 JH |
1241 | } |
1242 | ||
1243 | bool | |
1244 | Perl_is_utf8_idcont(pTHX_ U8 *p) | |
1245 | { | |
1246 | if (*p == '_') | |
1247 | return TRUE; | |
1248 | if (!is_utf8_char(p)) | |
1249 | return FALSE; | |
1250 | if (!PL_utf8_idcont) | |
1251 | PL_utf8_idcont = swash_init("utf8", "IdContinue", &PL_sv_undef, 0, 0); | |
eb160463 | 1252 | return swash_fetch(PL_utf8_idcont, p, TRUE) != 0; |
a0ed51b3 LW |
1253 | } |
1254 | ||
1255 | bool | |
864dbfa3 | 1256 | Perl_is_utf8_alpha(pTHX_ U8 *p) |
a0ed51b3 | 1257 | { |
386d01d6 GS |
1258 | if (!is_utf8_char(p)) |
1259 | return FALSE; | |
a0ed51b3 | 1260 | if (!PL_utf8_alpha) |
e24b16f9 | 1261 | PL_utf8_alpha = swash_init("utf8", "IsAlpha", &PL_sv_undef, 0, 0); |
eb160463 | 1262 | return swash_fetch(PL_utf8_alpha, p, TRUE) != 0; |
a0ed51b3 LW |
1263 | } |
1264 | ||
1265 | bool | |
b8c5462f JH |
1266 | Perl_is_utf8_ascii(pTHX_ U8 *p) |
1267 | { | |
386d01d6 GS |
1268 | if (!is_utf8_char(p)) |
1269 | return FALSE; | |
b8c5462f JH |
1270 | if (!PL_utf8_ascii) |
1271 | PL_utf8_ascii = swash_init("utf8", "IsAscii", &PL_sv_undef, 0, 0); | |
eb160463 | 1272 | return swash_fetch(PL_utf8_ascii, p, TRUE) != 0; |
b8c5462f JH |
1273 | } |
1274 | ||
1275 | bool | |
864dbfa3 | 1276 | Perl_is_utf8_space(pTHX_ U8 *p) |
a0ed51b3 | 1277 | { |
386d01d6 GS |
1278 | if (!is_utf8_char(p)) |
1279 | return FALSE; | |
a0ed51b3 | 1280 | if (!PL_utf8_space) |
3bec3564 | 1281 | PL_utf8_space = swash_init("utf8", "IsSpacePerl", &PL_sv_undef, 0, 0); |
eb160463 | 1282 | return swash_fetch(PL_utf8_space, p, TRUE) != 0; |
a0ed51b3 LW |
1283 | } |
1284 | ||
1285 | bool | |
864dbfa3 | 1286 | Perl_is_utf8_digit(pTHX_ U8 *p) |
a0ed51b3 | 1287 | { |
386d01d6 GS |
1288 | if (!is_utf8_char(p)) |
1289 | return FALSE; | |
a0ed51b3 | 1290 | if (!PL_utf8_digit) |
e24b16f9 | 1291 | PL_utf8_digit = swash_init("utf8", "IsDigit", &PL_sv_undef, 0, 0); |
eb160463 | 1292 | return swash_fetch(PL_utf8_digit, p, TRUE) != 0; |
a0ed51b3 LW |
1293 | } |
1294 | ||
1295 | bool | |
864dbfa3 | 1296 | Perl_is_utf8_upper(pTHX_ U8 *p) |
a0ed51b3 | 1297 | { |
386d01d6 GS |
1298 | if (!is_utf8_char(p)) |
1299 | return FALSE; | |
a0ed51b3 | 1300 | if (!PL_utf8_upper) |
c65e4d19 | 1301 | PL_utf8_upper = swash_init("utf8", "IsUppercase", &PL_sv_undef, 0, 0); |
eb160463 | 1302 | return swash_fetch(PL_utf8_upper, p, TRUE) != 0; |
a0ed51b3 LW |
1303 | } |
1304 | ||
1305 | bool | |
864dbfa3 | 1306 | Perl_is_utf8_lower(pTHX_ U8 *p) |
a0ed51b3 | 1307 | { |
386d01d6 GS |
1308 | if (!is_utf8_char(p)) |
1309 | return FALSE; | |
a0ed51b3 | 1310 | if (!PL_utf8_lower) |
c65e4d19 | 1311 | PL_utf8_lower = swash_init("utf8", "IsLowercase", &PL_sv_undef, 0, 0); |
eb160463 | 1312 | return swash_fetch(PL_utf8_lower, p, TRUE) != 0; |
a0ed51b3 LW |
1313 | } |
1314 | ||
1315 | bool | |
b8c5462f JH |
1316 | Perl_is_utf8_cntrl(pTHX_ U8 *p) |
1317 | { | |
386d01d6 GS |
1318 | if (!is_utf8_char(p)) |
1319 | return FALSE; | |
b8c5462f JH |
1320 | if (!PL_utf8_cntrl) |
1321 | PL_utf8_cntrl = swash_init("utf8", "IsCntrl", &PL_sv_undef, 0, 0); | |
eb160463 | 1322 | return swash_fetch(PL_utf8_cntrl, p, TRUE) != 0; |
b8c5462f JH |
1323 | } |
1324 | ||
1325 | bool | |
1326 | Perl_is_utf8_graph(pTHX_ U8 *p) | |
1327 | { | |
386d01d6 GS |
1328 | if (!is_utf8_char(p)) |
1329 | return FALSE; | |
b8c5462f JH |
1330 | if (!PL_utf8_graph) |
1331 | PL_utf8_graph = swash_init("utf8", "IsGraph", &PL_sv_undef, 0, 0); | |
eb160463 | 1332 | return swash_fetch(PL_utf8_graph, p, TRUE) != 0; |
b8c5462f JH |
1333 | } |
1334 | ||
1335 | bool | |
864dbfa3 | 1336 | Perl_is_utf8_print(pTHX_ U8 *p) |
a0ed51b3 | 1337 | { |
386d01d6 GS |
1338 | if (!is_utf8_char(p)) |
1339 | return FALSE; | |
a0ed51b3 | 1340 | if (!PL_utf8_print) |
e24b16f9 | 1341 | PL_utf8_print = swash_init("utf8", "IsPrint", &PL_sv_undef, 0, 0); |
eb160463 | 1342 | return swash_fetch(PL_utf8_print, p, TRUE) != 0; |
a0ed51b3 LW |
1343 | } |
1344 | ||
1345 | bool | |
b8c5462f JH |
1346 | Perl_is_utf8_punct(pTHX_ U8 *p) |
1347 | { | |
386d01d6 GS |
1348 | if (!is_utf8_char(p)) |
1349 | return FALSE; | |
b8c5462f JH |
1350 | if (!PL_utf8_punct) |
1351 | PL_utf8_punct = swash_init("utf8", "IsPunct", &PL_sv_undef, 0, 0); | |
eb160463 | 1352 | return swash_fetch(PL_utf8_punct, p, TRUE) != 0; |
b8c5462f JH |
1353 | } |
1354 | ||
1355 | bool | |
1356 | Perl_is_utf8_xdigit(pTHX_ U8 *p) | |
1357 | { | |
386d01d6 GS |
1358 | if (!is_utf8_char(p)) |
1359 | return FALSE; | |
b8c5462f JH |
1360 | if (!PL_utf8_xdigit) |
1361 | PL_utf8_xdigit = swash_init("utf8", "IsXDigit", &PL_sv_undef, 0, 0); | |
eb160463 | 1362 | return swash_fetch(PL_utf8_xdigit, p, TRUE) != 0; |
b8c5462f JH |
1363 | } |
1364 | ||
1365 | bool | |
864dbfa3 | 1366 | Perl_is_utf8_mark(pTHX_ U8 *p) |
a0ed51b3 | 1367 | { |
386d01d6 GS |
1368 | if (!is_utf8_char(p)) |
1369 | return FALSE; | |
a0ed51b3 | 1370 | if (!PL_utf8_mark) |
e24b16f9 | 1371 | PL_utf8_mark = swash_init("utf8", "IsM", &PL_sv_undef, 0, 0); |
eb160463 | 1372 | return swash_fetch(PL_utf8_mark, p, TRUE) != 0; |
a0ed51b3 LW |
1373 | } |
1374 | ||
6b5c0936 JH |
1375 | /* |
1376 | =for apidoc A|UV|to_utf8_case|U8 *p|U8* ustrp|STRLEN *lenp|SV **swash|char *normal|char *special | |
1377 | ||
1378 | The "p" contains the pointer to the UTF-8 string encoding | |
1379 | the character that is being converted. | |
1380 | ||
1381 | The "ustrp" is a pointer to the character buffer to put the | |
1382 | conversion result to. The "lenp" is a pointer to the length | |
1383 | of the result. | |
1384 | ||
0134edef | 1385 | The "swashp" is a pointer to the swash to use. |
6b5c0936 | 1386 | |
0134edef JH |
1387 | Both the special and normal mappings are stored lib/unicore/To/Foo.pl, |
1388 | and loaded by SWASHGET, using lib/utf8_heavy.pl. The special (usually, | |
1389 | but not always, a multicharacter mapping), is tried first. | |
6b5c0936 | 1390 | |
0134edef JH |
1391 | The "special" is a string like "utf8::ToSpecLower", which means the |
1392 | hash %utf8::ToSpecLower. The access to the hash is through | |
1393 | Perl_to_utf8_case(). | |
6b5c0936 | 1394 | |
0134edef JH |
1395 | The "normal" is a string like "ToLower" which means the swash |
1396 | %utf8::ToLower. | |
1397 | ||
1398 | =cut */ | |
6b5c0936 | 1399 | |
2104c8d9 | 1400 | UV |
a6872d42 | 1401 | Perl_to_utf8_case(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp, SV **swashp, char *normal, char *special) |
a0ed51b3 | 1402 | { |
0134edef | 1403 | UV uv0, uv1; |
89ebb4a3 | 1404 | U8 tmpbuf[UTF8_MAXBYTES_CASE+1]; |
0134edef | 1405 | STRLEN len = 0; |
a0ed51b3 | 1406 | |
1feea2c7 JH |
1407 | uv0 = utf8_to_uvchr(p, 0); |
1408 | /* The NATIVE_TO_UNI() and UNI_TO_NATIVE() mappings | |
1409 | * are necessary in EBCDIC, they are redundant no-ops | |
1410 | * in ASCII-ish platforms, and hopefully optimized away. */ | |
1411 | uv1 = NATIVE_TO_UNI(uv0); | |
1412 | uvuni_to_utf8(tmpbuf, uv1); | |
0134edef JH |
1413 | |
1414 | if (!*swashp) /* load on-demand */ | |
1415 | *swashp = swash_init("utf8", normal, &PL_sv_undef, 4, 0); | |
1416 | ||
b08cf34e JH |
1417 | /* The 0xDF is the only special casing Unicode code point below 0x100. */ |
1418 | if (special && (uv1 == 0xDF || uv1 > 0xFF)) { | |
0134edef | 1419 | /* It might be "special" (sometimes, but not always, |
2a37f04d | 1420 | * a multicharacter mapping) */ |
983ffd37 | 1421 | HV *hv; |
b08cf34e JH |
1422 | SV **svp; |
1423 | ||
1424 | if ((hv = get_hv(special, FALSE)) && | |
1425 | (svp = hv_fetch(hv, (const char*)tmpbuf, UNISKIP(uv1), FALSE)) && | |
1426 | (*svp)) { | |
1427 | char *s; | |
47654450 | 1428 | |
b08cf34e | 1429 | s = SvPV(*svp, len); |
47654450 JH |
1430 | if (len == 1) |
1431 | len = uvuni_to_utf8(ustrp, NATIVE_TO_UNI(*(U8*)s)) - ustrp; | |
2a37f04d | 1432 | else { |
2f9475ad JH |
1433 | #ifdef EBCDIC |
1434 | /* If we have EBCDIC we need to remap the characters | |
1435 | * since any characters in the low 256 are Unicode | |
1436 | * code points, not EBCDIC. */ | |
7cda7a3d | 1437 | U8 *t = (U8*)s, *tend = t + len, *d; |
2f9475ad JH |
1438 | |
1439 | d = tmpbuf; | |
b08cf34e | 1440 | if (SvUTF8(*svp)) { |
2f9475ad JH |
1441 | STRLEN tlen = 0; |
1442 | ||
1443 | while (t < tend) { | |
1444 | UV c = utf8_to_uvchr(t, &tlen); | |
1445 | if (tlen > 0) { | |
1446 | d = uvchr_to_utf8(d, UNI_TO_NATIVE(c)); | |
1447 | t += tlen; | |
1448 | } | |
1449 | else | |
1450 | break; | |
1451 | } | |
1452 | } | |
1453 | else { | |
36fec512 JH |
1454 | while (t < tend) { |
1455 | d = uvchr_to_utf8(d, UNI_TO_NATIVE(*t)); | |
1456 | t++; | |
1457 | } | |
2f9475ad JH |
1458 | } |
1459 | len = d - tmpbuf; | |
1460 | Copy(tmpbuf, ustrp, len, U8); | |
1461 | #else | |
d2dcd0fb | 1462 | Copy(s, ustrp, len, U8); |
2f9475ad | 1463 | #endif |
29e98929 | 1464 | } |
983ffd37 | 1465 | } |
0134edef JH |
1466 | } |
1467 | ||
1468 | if (!len && *swashp) { | |
1469 | UV uv2 = swash_fetch(*swashp, tmpbuf, TRUE); | |
1470 | ||
1471 | if (uv2) { | |
1472 | /* It was "normal" (a single character mapping). */ | |
1473 | UV uv3 = UNI_TO_NATIVE(uv2); | |
1474 | ||
e9101d72 | 1475 | len = uvchr_to_utf8(ustrp, uv3) - ustrp; |
2a37f04d JH |
1476 | } |
1477 | } | |
1feea2c7 | 1478 | |
0134edef JH |
1479 | if (!len) /* Neither: just copy. */ |
1480 | len = uvchr_to_utf8(ustrp, uv0) - ustrp; | |
1481 | ||
2a37f04d JH |
1482 | if (lenp) |
1483 | *lenp = len; | |
1484 | ||
0134edef | 1485 | return len ? utf8_to_uvchr(ustrp, 0) : 0; |
a0ed51b3 LW |
1486 | } |
1487 | ||
d3e79532 JH |
1488 | /* |
1489 | =for apidoc A|UV|to_utf8_upper|U8 *p|U8 *ustrp|STRLEN *lenp | |
1490 | ||
1491 | Convert the UTF-8 encoded character at p to its uppercase version and | |
1492 | store that in UTF-8 in ustrp and its length in bytes in lenp. Note | |
89ebb4a3 JH |
1493 | that the ustrp needs to be at least UTF8_MAXBYTES_CASE+1 bytes since |
1494 | the uppercase version may be longer than the original character. | |
d3e79532 JH |
1495 | |
1496 | The first character of the uppercased version is returned | |
1497 | (but note, as explained above, that there may be more.) | |
1498 | ||
1499 | =cut */ | |
1500 | ||
2104c8d9 | 1501 | UV |
983ffd37 | 1502 | Perl_to_utf8_upper(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp) |
a0ed51b3 | 1503 | { |
983ffd37 | 1504 | return Perl_to_utf8_case(aTHX_ p, ustrp, lenp, |
b4e400f9 | 1505 | &PL_utf8_toupper, "ToUpper", "utf8::ToSpecUpper"); |
983ffd37 | 1506 | } |
a0ed51b3 | 1507 | |
d3e79532 JH |
1508 | /* |
1509 | =for apidoc A|UV|to_utf8_title|U8 *p|U8 *ustrp|STRLEN *lenp | |
1510 | ||
1511 | Convert the UTF-8 encoded character at p to its titlecase version and | |
1512 | store that in UTF-8 in ustrp and its length in bytes in lenp. Note | |
89ebb4a3 JH |
1513 | that the ustrp needs to be at least UTF8_MAXBYTES_CASE+1 bytes since the |
1514 | titlecase version may be longer than the original character. | |
d3e79532 JH |
1515 | |
1516 | The first character of the titlecased version is returned | |
1517 | (but note, as explained above, that there may be more.) | |
1518 | ||
1519 | =cut */ | |
1520 | ||
983ffd37 JH |
1521 | UV |
1522 | Perl_to_utf8_title(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp) | |
1523 | { | |
1524 | return Perl_to_utf8_case(aTHX_ p, ustrp, lenp, | |
b4e400f9 | 1525 | &PL_utf8_totitle, "ToTitle", "utf8::ToSpecTitle"); |
a0ed51b3 LW |
1526 | } |
1527 | ||
d3e79532 JH |
1528 | /* |
1529 | =for apidoc A|UV|to_utf8_lower|U8 *p|U8 *ustrp|STRLEN *lenp | |
1530 | ||
1531 | Convert the UTF-8 encoded character at p to its lowercase version and | |
1532 | store that in UTF-8 in ustrp and its length in bytes in lenp. Note | |
89ebb4a3 JH |
1533 | that the ustrp needs to be at least UTF8_MAXBYTES_CASE+1 bytes since the |
1534 | lowercase version may be longer than the original character. | |
d3e79532 JH |
1535 | |
1536 | The first character of the lowercased version is returned | |
1537 | (but note, as explained above, that there may be more.) | |
1538 | ||
1539 | =cut */ | |
1540 | ||
2104c8d9 | 1541 | UV |
a2a2844f | 1542 | Perl_to_utf8_lower(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp) |
a0ed51b3 | 1543 | { |
983ffd37 | 1544 | return Perl_to_utf8_case(aTHX_ p, ustrp, lenp, |
b4e400f9 JH |
1545 | &PL_utf8_tolower, "ToLower", "utf8::ToSpecLower"); |
1546 | } | |
1547 | ||
d3e79532 JH |
1548 | /* |
1549 | =for apidoc A|UV|to_utf8_fold|U8 *p|U8 *ustrp|STRLEN *lenp | |
1550 | ||
1551 | Convert the UTF-8 encoded character at p to its foldcase version and | |
1552 | store that in UTF-8 in ustrp and its length in bytes in lenp. Note | |
89ebb4a3 | 1553 | that the ustrp needs to be at least UTF8_MAXBYTES_CASE+1 bytes since the |
d3e79532 JH |
1554 | foldcase version may be longer than the original character (up to |
1555 | three characters). | |
1556 | ||
1557 | The first character of the foldcased version is returned | |
1558 | (but note, as explained above, that there may be more.) | |
1559 | ||
1560 | =cut */ | |
1561 | ||
b4e400f9 JH |
1562 | UV |
1563 | Perl_to_utf8_fold(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp) | |
1564 | { | |
1565 | return Perl_to_utf8_case(aTHX_ p, ustrp, lenp, | |
1566 | &PL_utf8_tofold, "ToFold", "utf8::ToSpecFold"); | |
a0ed51b3 LW |
1567 | } |
1568 | ||
1569 | /* a "swash" is a swatch hash */ | |
1570 | ||
1571 | SV* | |
864dbfa3 | 1572 | Perl_swash_init(pTHX_ char* pkg, char* name, SV *listsv, I32 minbits, I32 none) |
a0ed51b3 LW |
1573 | { |
1574 | SV* retval; | |
243b1711 | 1575 | SV* tokenbufsv = sv_newmortal(); |
8e84507e | 1576 | dSP; |
71bed85a NC |
1577 | size_t pkg_len = strlen(pkg); |
1578 | size_t name_len = strlen(name); | |
1579 | HV *stash = gv_stashpvn(pkg, pkg_len, FALSE); | |
f8be5cf0 | 1580 | SV* errsv_save; |
ce3b816e | 1581 | |
1b026014 | 1582 | if (!gv_fetchmeth(stash, "SWASHNEW", 8, -1)) { /* demand load utf8 */ |
ce3b816e | 1583 | ENTER; |
f8be5cf0 | 1584 | errsv_save = newSVsv(ERRSV); |
71bed85a NC |
1585 | Perl_load_module(aTHX_ PERL_LOADMOD_NOIMPORT, newSVpvn(pkg,pkg_len), |
1586 | Nullsv); | |
f8be5cf0 JH |
1587 | if (!SvTRUE(ERRSV)) |
1588 | sv_setsv(ERRSV, errsv_save); | |
1589 | SvREFCNT_dec(errsv_save); | |
ce3b816e GS |
1590 | LEAVE; |
1591 | } | |
1592 | SPAGAIN; | |
a0ed51b3 LW |
1593 | PUSHSTACKi(PERLSI_MAGIC); |
1594 | PUSHMARK(SP); | |
1595 | EXTEND(SP,5); | |
71bed85a NC |
1596 | PUSHs(sv_2mortal(newSVpvn(pkg, pkg_len))); |
1597 | PUSHs(sv_2mortal(newSVpvn(name, name_len))); | |
a0ed51b3 LW |
1598 | PUSHs(listsv); |
1599 | PUSHs(sv_2mortal(newSViv(minbits))); | |
1600 | PUSHs(sv_2mortal(newSViv(none))); | |
1601 | PUTBACK; | |
1602 | ENTER; | |
1603 | SAVEI32(PL_hints); | |
1604 | PL_hints = 0; | |
1605 | save_re_context(); | |
923e4eb5 | 1606 | if (IN_PERL_COMPILETIME) { |
bf1fed83 | 1607 | /* XXX ought to be handled by lex_start */ |
82686b01 | 1608 | SAVEI32(PL_in_my); |
2b4bd638 | 1609 | PL_in_my = 0; |
bf1fed83 | 1610 | sv_setpv(tokenbufsv, PL_tokenbuf); |
82686b01 | 1611 | } |
f8be5cf0 | 1612 | errsv_save = newSVsv(ERRSV); |
864dbfa3 | 1613 | if (call_method("SWASHNEW", G_SCALAR)) |
8e84507e | 1614 | retval = newSVsv(*PL_stack_sp--); |
a0ed51b3 | 1615 | else |
e24b16f9 | 1616 | retval = &PL_sv_undef; |
f8be5cf0 JH |
1617 | if (!SvTRUE(ERRSV)) |
1618 | sv_setsv(ERRSV, errsv_save); | |
1619 | SvREFCNT_dec(errsv_save); | |
a0ed51b3 LW |
1620 | LEAVE; |
1621 | POPSTACK; | |
923e4eb5 | 1622 | if (IN_PERL_COMPILETIME) { |
bf1fed83 JH |
1623 | STRLEN len; |
1624 | char* pv = SvPV(tokenbufsv, len); | |
1625 | ||
1626 | Copy(pv, PL_tokenbuf, len+1, char); | |
eb160463 | 1627 | PL_curcop->op_private = (U8)(PL_hints & HINT_PRIVATE_MASK); |
a0ed51b3 | 1628 | } |
bc45ce41 JH |
1629 | if (!SvROK(retval) || SvTYPE(SvRV(retval)) != SVt_PVHV) { |
1630 | if (SvPOK(retval)) | |
35c1215d NC |
1631 | Perl_croak(aTHX_ "Can't find Unicode property definition \"%"SVf"\"", |
1632 | retval); | |
cea2e8a9 | 1633 | Perl_croak(aTHX_ "SWASHNEW didn't return an HV ref"); |
bc45ce41 | 1634 | } |
a0ed51b3 LW |
1635 | return retval; |
1636 | } | |
1637 | ||
035d37be JH |
1638 | |
1639 | /* This API is wrong for special case conversions since we may need to | |
1640 | * return several Unicode characters for a single Unicode character | |
1641 | * (see lib/unicore/SpecCase.txt) The SWASHGET in lib/utf8_heavy.pl is | |
1642 | * the lower-level routine, and it is similarly broken for returning | |
1643 | * multiple values. --jhi */ | |
a0ed51b3 | 1644 | UV |
3568d838 | 1645 | Perl_swash_fetch(pTHX_ SV *sv, U8 *ptr, bool do_utf8) |
a0ed51b3 LW |
1646 | { |
1647 | HV* hv = (HV*)SvRV(sv); | |
3568d838 JH |
1648 | U32 klen; |
1649 | U32 off; | |
a0ed51b3 | 1650 | STRLEN slen; |
7d85a32c | 1651 | STRLEN needents; |
4ea42e7f | 1652 | U8 *tmps = NULL; |
a0ed51b3 LW |
1653 | U32 bit; |
1654 | SV *retval; | |
3568d838 JH |
1655 | U8 tmputf8[2]; |
1656 | UV c = NATIVE_TO_ASCII(*ptr); | |
1657 | ||
1658 | if (!do_utf8 && !UNI_IS_INVARIANT(c)) { | |
eb160463 GS |
1659 | tmputf8[0] = (U8)UTF8_EIGHT_BIT_HI(c); |
1660 | tmputf8[1] = (U8)UTF8_EIGHT_BIT_LO(c); | |
3568d838 JH |
1661 | ptr = tmputf8; |
1662 | } | |
1663 | /* Given a UTF-X encoded char 0xAA..0xYY,0xZZ | |
1664 | * then the "swatch" is a vec() for al the chars which start | |
1665 | * with 0xAA..0xYY | |
1666 | * So the key in the hash (klen) is length of encoded char -1 | |
1667 | */ | |
1668 | klen = UTF8SKIP(ptr) - 1; | |
1669 | off = ptr[klen]; | |
a0ed51b3 | 1670 | |
7d85a32c JH |
1671 | if (klen == 0) |
1672 | { | |
1673 | /* If char in invariant then swatch is for all the invariant chars | |
1e54db1a | 1674 | * In both UTF-8 and UTF-8-MOD that happens to be UTF_CONTINUATION_MARK |
7d85a32c JH |
1675 | */ |
1676 | needents = UTF_CONTINUATION_MARK; | |
1677 | off = NATIVE_TO_UTF(ptr[klen]); | |
1678 | } | |
1679 | else | |
1680 | { | |
1681 | /* If char is encoded then swatch is for the prefix */ | |
1682 | needents = (1 << UTF_ACCUMULATION_SHIFT); | |
1683 | off = NATIVE_TO_UTF(ptr[klen]) & UTF_CONTINUATION_MASK; | |
1684 | } | |
1685 | ||
a0ed51b3 LW |
1686 | /* |
1687 | * This single-entry cache saves about 1/3 of the utf8 overhead in test | |
1688 | * suite. (That is, only 7-8% overall over just a hash cache. Still, | |
1689 | * it's nothing to sniff at.) Pity we usually come through at least | |
1690 | * two function calls to get here... | |
1691 | * | |
1692 | * NB: this code assumes that swatches are never modified, once generated! | |
1693 | */ | |
1694 | ||
3568d838 | 1695 | if (hv == PL_last_swash_hv && |
a0ed51b3 | 1696 | klen == PL_last_swash_klen && |
3568d838 | 1697 | (!klen || memEQ((char *)ptr, (char *)PL_last_swash_key, klen)) ) |
a0ed51b3 LW |
1698 | { |
1699 | tmps = PL_last_swash_tmps; | |
1700 | slen = PL_last_swash_slen; | |
1701 | } | |
1702 | else { | |
1703 | /* Try our second-level swatch cache, kept in a hash. */ | |
dfe13c55 | 1704 | SV** svp = hv_fetch(hv, (char*)ptr, klen, FALSE); |
a0ed51b3 LW |
1705 | |
1706 | /* If not cached, generate it via utf8::SWASHGET */ | |
dfe13c55 | 1707 | if (!svp || !SvPOK(*svp) || !(tmps = (U8*)SvPV(*svp, slen))) { |
a0ed51b3 | 1708 | dSP; |
2b9d42f0 NIS |
1709 | /* We use utf8n_to_uvuni() as we want an index into |
1710 | Unicode tables, not a native character number. | |
1711 | */ | |
89ebb4a3 | 1712 | UV code_point = utf8n_to_uvuni(ptr, UTF8_MAXBYTES, 0, |
872c91ae JH |
1713 | ckWARN(WARN_UTF8) ? |
1714 | 0 : UTF8_ALLOW_ANY); | |
f8be5cf0 | 1715 | SV *errsv_save; |
a0ed51b3 LW |
1716 | ENTER; |
1717 | SAVETMPS; | |
1718 | save_re_context(); | |
1719 | PUSHSTACKi(PERLSI_MAGIC); | |
1720 | PUSHMARK(SP); | |
1721 | EXTEND(SP,3); | |
1722 | PUSHs((SV*)sv); | |
ffbc6a93 | 1723 | /* On EBCDIC & ~(0xA0-1) isn't a useful thing to do */ |
3568d838 JH |
1724 | PUSHs(sv_2mortal(newSViv((klen) ? |
1725 | (code_point & ~(needents - 1)) : 0))); | |
a0ed51b3 LW |
1726 | PUSHs(sv_2mortal(newSViv(needents))); |
1727 | PUTBACK; | |
f8be5cf0 | 1728 | errsv_save = newSVsv(ERRSV); |
864dbfa3 | 1729 | if (call_method("SWASHGET", G_SCALAR)) |
8e84507e | 1730 | retval = newSVsv(*PL_stack_sp--); |
a0ed51b3 | 1731 | else |
e24b16f9 | 1732 | retval = &PL_sv_undef; |
f8be5cf0 JH |
1733 | if (!SvTRUE(ERRSV)) |
1734 | sv_setsv(ERRSV, errsv_save); | |
1735 | SvREFCNT_dec(errsv_save); | |
a0ed51b3 LW |
1736 | POPSTACK; |
1737 | FREETMPS; | |
1738 | LEAVE; | |
923e4eb5 | 1739 | if (IN_PERL_COMPILETIME) |
eb160463 | 1740 | PL_curcop->op_private = (U8)(PL_hints & HINT_PRIVATE_MASK); |
a0ed51b3 | 1741 | |
dfe13c55 | 1742 | svp = hv_store(hv, (char*)ptr, klen, retval, 0); |
a0ed51b3 | 1743 | |
7d85a32c | 1744 | if (!svp || !(tmps = (U8*)SvPV(*svp, slen)) || (slen << 3) < needents) |
cea2e8a9 | 1745 | Perl_croak(aTHX_ "SWASHGET didn't return result of proper length"); |
a0ed51b3 LW |
1746 | } |
1747 | ||
1748 | PL_last_swash_hv = hv; | |
1749 | PL_last_swash_klen = klen; | |
1750 | PL_last_swash_tmps = tmps; | |
1751 | PL_last_swash_slen = slen; | |
1752 | if (klen) | |
1753 | Copy(ptr, PL_last_swash_key, klen, U8); | |
1754 | } | |
1755 | ||
9faf8d75 | 1756 | switch ((int)((slen << 3) / needents)) { |
a0ed51b3 LW |
1757 | case 1: |
1758 | bit = 1 << (off & 7); | |
1759 | off >>= 3; | |
1760 | return (tmps[off] & bit) != 0; | |
1761 | case 8: | |
1762 | return tmps[off]; | |
1763 | case 16: | |
1764 | off <<= 1; | |
1765 | return (tmps[off] << 8) + tmps[off + 1] ; | |
1766 | case 32: | |
1767 | off <<= 2; | |
1768 | return (tmps[off] << 24) + (tmps[off+1] << 16) + (tmps[off+2] << 8) + tmps[off + 3] ; | |
1769 | } | |
cea2e8a9 | 1770 | Perl_croak(aTHX_ "panic: swash_fetch"); |
a0ed51b3 LW |
1771 | return 0; |
1772 | } | |
2b9d42f0 NIS |
1773 | |
1774 | ||
1775 | /* | |
37607a96 | 1776 | =for apidoc A|U8 *|uvchr_to_utf8|U8 *d|UV uv |
2b9d42f0 | 1777 | |
1e54db1a | 1778 | Adds the UTF-8 representation of the Native codepoint C<uv> to the end |
89ebb4a3 | 1779 | of the string C<d>; C<d> should be have at least C<UTF8_MAXBYTES+1> free |
2b9d42f0 NIS |
1780 | bytes available. The return value is the pointer to the byte after the |
1781 | end of the new character. In other words, | |
1782 | ||
1783 | d = uvchr_to_utf8(d, uv); | |
1784 | ||
1785 | is the recommended wide native character-aware way of saying | |
1786 | ||
1787 | *(d++) = uv; | |
1788 | ||
1789 | =cut | |
1790 | */ | |
1791 | ||
1792 | /* On ASCII machines this is normally a macro but we want a | |
1793 | real function in case XS code wants it | |
1794 | */ | |
1795 | #undef Perl_uvchr_to_utf8 | |
1796 | U8 * | |
1797 | Perl_uvchr_to_utf8(pTHX_ U8 *d, UV uv) | |
1798 | { | |
b851fbc1 | 1799 | return Perl_uvuni_to_utf8_flags(aTHX_ d, NATIVE_TO_UNI(uv), 0); |
2b9d42f0 NIS |
1800 | } |
1801 | ||
b851fbc1 JH |
1802 | U8 * |
1803 | Perl_uvchr_to_utf8_flags(pTHX_ U8 *d, UV uv, UV flags) | |
1804 | { | |
1805 | return Perl_uvuni_to_utf8_flags(aTHX_ d, NATIVE_TO_UNI(uv), flags); | |
1806 | } | |
2b9d42f0 NIS |
1807 | |
1808 | /* | |
37607a96 | 1809 | =for apidoc A|UV|utf8n_to_uvchr|U8 *s|STRLEN curlen|STRLEN *retlen|U32 flags |
2b9d42f0 NIS |
1810 | |
1811 | Returns the native character value of the first character in the string C<s> | |
1e54db1a | 1812 | which is assumed to be in UTF-8 encoding; C<retlen> will be set to the |
2b9d42f0 NIS |
1813 | length, in bytes, of that character. |
1814 | ||
1815 | Allows length and flags to be passed to low level routine. | |
1816 | ||
1817 | =cut | |
1818 | */ | |
0a2ef054 JH |
1819 | /* On ASCII machines this is normally a macro but we want |
1820 | a real function in case XS code wants it | |
2b9d42f0 NIS |
1821 | */ |
1822 | #undef Perl_utf8n_to_uvchr | |
1823 | UV | |
37607a96 | 1824 | Perl_utf8n_to_uvchr(pTHX_ U8 *s, STRLEN curlen, STRLEN *retlen, U32 flags) |
2b9d42f0 NIS |
1825 | { |
1826 | UV uv = Perl_utf8n_to_uvuni(aTHX_ s, curlen, retlen, flags); | |
1827 | return UNI_TO_NATIVE(uv); | |
1828 | } | |
1829 | ||
d2cc3551 JH |
1830 | /* |
1831 | =for apidoc A|char *|pv_uni_display|SV *dsv|U8 *spv|STRLEN len|STRLEN pvlim|UV flags | |
1832 | ||
1833 | Build to the scalar dsv a displayable version of the string spv, | |
1834 | length len, the displayable version being at most pvlim bytes long | |
1835 | (if longer, the rest is truncated and "..." will be appended). | |
0a2ef054 | 1836 | |
9e55ce06 | 1837 | The flags argument can have UNI_DISPLAY_ISPRINT set to display |
00e86452 | 1838 | isPRINT()able characters as themselves, UNI_DISPLAY_BACKSLASH |
0a2ef054 JH |
1839 | to display the \\[nrfta\\] as the backslashed versions (like '\n') |
1840 | (UNI_DISPLAY_BACKSLASH is preferred over UNI_DISPLAY_ISPRINT for \\). | |
1841 | UNI_DISPLAY_QQ (and its alias UNI_DISPLAY_REGEX) have both | |
1842 | UNI_DISPLAY_BACKSLASH and UNI_DISPLAY_ISPRINT turned on. | |
1843 | ||
d2cc3551 JH |
1844 | The pointer to the PV of the dsv is returned. |
1845 | ||
1846 | =cut */ | |
e6b2e755 JH |
1847 | char * |
1848 | Perl_pv_uni_display(pTHX_ SV *dsv, U8 *spv, STRLEN len, STRLEN pvlim, UV flags) | |
1849 | { | |
1850 | int truncated = 0; | |
1851 | char *s, *e; | |
1852 | ||
1853 | sv_setpvn(dsv, "", 0); | |
1854 | for (s = (char *)spv, e = s + len; s < e; s += UTF8SKIP(s)) { | |
1855 | UV u; | |
a49f32c6 NC |
1856 | /* This serves double duty as a flag and a character to print after |
1857 | a \ when flags & UNI_DISPLAY_BACKSLASH is true. | |
1858 | */ | |
1859 | char ok = 0; | |
c728cb41 | 1860 | |
e6b2e755 JH |
1861 | if (pvlim && SvCUR(dsv) >= pvlim) { |
1862 | truncated++; | |
1863 | break; | |
1864 | } | |
1865 | u = utf8_to_uvchr((U8*)s, 0); | |
c728cb41 | 1866 | if (u < 256) { |
a49f32c6 | 1867 | unsigned char c = u & 0xFF; |
c728cb41 | 1868 | if (!ok && (flags & UNI_DISPLAY_BACKSLASH)) { |
a49f32c6 | 1869 | switch (c) { |
c728cb41 | 1870 | case '\n': |
a49f32c6 | 1871 | ok = 'n'; break; |
c728cb41 | 1872 | case '\r': |
a49f32c6 | 1873 | ok = 'r'; break; |
c728cb41 | 1874 | case '\t': |
a49f32c6 | 1875 | ok = 't'; break; |
c728cb41 | 1876 | case '\f': |
a49f32c6 | 1877 | ok = 'f'; break; |
c728cb41 | 1878 | case '\a': |
a49f32c6 | 1879 | ok = 'a'; break; |
c728cb41 | 1880 | case '\\': |
a49f32c6 | 1881 | ok = '\\'; break; |
c728cb41 JH |
1882 | default: break; |
1883 | } | |
a49f32c6 NC |
1884 | if (ok) { |
1885 | Perl_sv_catpvf(aTHX_ dsv, "\\%c", ok); | |
1886 | } | |
c728cb41 | 1887 | } |
00e86452 | 1888 | /* isPRINT() is the locale-blind version. */ |
a49f32c6 NC |
1889 | if (!ok && (flags & UNI_DISPLAY_ISPRINT) && isPRINT(c)) { |
1890 | Perl_sv_catpvf(aTHX_ dsv, "%c", c); | |
1891 | ok = 1; | |
0a2ef054 | 1892 | } |
c728cb41 JH |
1893 | } |
1894 | if (!ok) | |
9e55ce06 | 1895 | Perl_sv_catpvf(aTHX_ dsv, "\\x{%"UVxf"}", u); |
e6b2e755 JH |
1896 | } |
1897 | if (truncated) | |
1898 | sv_catpvn(dsv, "...", 3); | |
1899 | ||
1900 | return SvPVX(dsv); | |
1901 | } | |
2b9d42f0 | 1902 | |
d2cc3551 JH |
1903 | /* |
1904 | =for apidoc A|char *|sv_uni_display|SV *dsv|SV *ssv|STRLEN pvlim|UV flags | |
1905 | ||
1906 | Build to the scalar dsv a displayable version of the scalar sv, | |
0a2ef054 | 1907 | the displayable version being at most pvlim bytes long |
d2cc3551 | 1908 | (if longer, the rest is truncated and "..." will be appended). |
0a2ef054 JH |
1909 | |
1910 | The flags argument is as in pv_uni_display(). | |
1911 | ||
d2cc3551 JH |
1912 | The pointer to the PV of the dsv is returned. |
1913 | ||
1914 | =cut */ | |
e6b2e755 JH |
1915 | char * |
1916 | Perl_sv_uni_display(pTHX_ SV *dsv, SV *ssv, STRLEN pvlim, UV flags) | |
1917 | { | |
701a277b JH |
1918 | return Perl_pv_uni_display(aTHX_ dsv, (U8*)SvPVX(ssv), SvCUR(ssv), |
1919 | pvlim, flags); | |
1920 | } | |
1921 | ||
d2cc3551 | 1922 | /* |
d07ddd77 | 1923 | =for apidoc A|I32|ibcmp_utf8|const char *s1|char **pe1|register UV l1|bool u1|const char *s2|char **pe2|register UV l2|bool u2 |
d2cc3551 JH |
1924 | |
1925 | Return true if the strings s1 and s2 differ case-insensitively, false | |
1926 | if not (if they are equal case-insensitively). If u1 is true, the | |
1927 | string s1 is assumed to be in UTF-8-encoded Unicode. If u2 is true, | |
d07ddd77 JH |
1928 | the string s2 is assumed to be in UTF-8-encoded Unicode. If u1 or u2 |
1929 | are false, the respective string is assumed to be in native 8-bit | |
1930 | encoding. | |
1931 | ||
1932 | If the pe1 and pe2 are non-NULL, the scanning pointers will be copied | |
1933 | in there (they will point at the beginning of the I<next> character). | |
1934 | If the pointers behind pe1 or pe2 are non-NULL, they are the end | |
1935 | pointers beyond which scanning will not continue under any | |
4cdaeff7 | 1936 | circumstances. If the byte lengths l1 and l2 are non-zero, s1+l1 and |
d07ddd77 JH |
1937 | s2+l2 will be used as goal end pointers that will also stop the scan, |
1938 | and which qualify towards defining a successful match: all the scans | |
1939 | that define an explicit length must reach their goal pointers for | |
1940 | a match to succeed). | |
d2cc3551 JH |
1941 | |
1942 | For case-insensitiveness, the "casefolding" of Unicode is used | |
1943 | instead of upper/lowercasing both the characters, see | |
1944 | http://www.unicode.org/unicode/reports/tr21/ (Case Mappings). | |
1945 | ||
1946 | =cut */ | |
701a277b | 1947 | I32 |
d07ddd77 | 1948 | Perl_ibcmp_utf8(pTHX_ const char *s1, char **pe1, register UV l1, bool u1, const char *s2, char **pe2, register UV l2, bool u2) |
332ddc25 | 1949 | { |
5469e704 JH |
1950 | register U8 *p1 = (U8*)s1; |
1951 | register U8 *p2 = (U8*)s2; | |
d07ddd77 JH |
1952 | register U8 *e1 = 0, *f1 = 0, *q1 = 0; |
1953 | register U8 *e2 = 0, *f2 = 0, *q2 = 0; | |
1954 | STRLEN n1 = 0, n2 = 0; | |
89ebb4a3 JH |
1955 | U8 foldbuf1[UTF8_MAXBYTES_CASE+1]; |
1956 | U8 foldbuf2[UTF8_MAXBYTES_CASE+1]; | |
d7f013c8 JH |
1957 | U8 natbuf[1+1]; |
1958 | STRLEN foldlen1, foldlen2; | |
d07ddd77 | 1959 | bool match; |
332ddc25 | 1960 | |
d07ddd77 JH |
1961 | if (pe1) |
1962 | e1 = *(U8**)pe1; | |
eb160463 | 1963 | if (e1 == 0 || (l1 && l1 < (UV)(e1 - (U8*)s1))) |
d07ddd77 JH |
1964 | f1 = (U8*)s1 + l1; |
1965 | if (pe2) | |
1966 | e2 = *(U8**)pe2; | |
eb160463 | 1967 | if (e2 == 0 || (l2 && l2 < (UV)(e2 - (U8*)s2))) |
d07ddd77 JH |
1968 | f2 = (U8*)s2 + l2; |
1969 | ||
1970 | if ((e1 == 0 && f1 == 0) || (e2 == 0 && f2 == 0) || (f1 == 0 && f2 == 0)) | |
1971 | return 1; /* mismatch; possible infinite loop or false positive */ | |
1972 | ||
a6872d42 JH |
1973 | if (!u1 || !u2) |
1974 | natbuf[1] = 0; /* Need to terminate the buffer. */ | |
1975 | ||
d07ddd77 JH |
1976 | while ((e1 == 0 || p1 < e1) && |
1977 | (f1 == 0 || p1 < f1) && | |
1978 | (e2 == 0 || p2 < e2) && | |
1979 | (f2 == 0 || p2 < f2)) { | |
1980 | if (n1 == 0) { | |
d7f013c8 JH |
1981 | if (u1) |
1982 | to_utf8_fold(p1, foldbuf1, &foldlen1); | |
1983 | else { | |
f5cee151 | 1984 | natbuf[0] = *p1; |
d7f013c8 JH |
1985 | to_utf8_fold(natbuf, foldbuf1, &foldlen1); |
1986 | } | |
1987 | q1 = foldbuf1; | |
d07ddd77 | 1988 | n1 = foldlen1; |
332ddc25 | 1989 | } |
d07ddd77 | 1990 | if (n2 == 0) { |
d7f013c8 JH |
1991 | if (u2) |
1992 | to_utf8_fold(p2, foldbuf2, &foldlen2); | |
1993 | else { | |
f5cee151 | 1994 | natbuf[0] = *p2; |
d7f013c8 JH |
1995 | to_utf8_fold(natbuf, foldbuf2, &foldlen2); |
1996 | } | |
1997 | q2 = foldbuf2; | |
d07ddd77 | 1998 | n2 = foldlen2; |
332ddc25 | 1999 | } |
d07ddd77 JH |
2000 | while (n1 && n2) { |
2001 | if ( UTF8SKIP(q1) != UTF8SKIP(q2) || | |
2002 | (UTF8SKIP(q1) == 1 && *q1 != *q2) || | |
2003 | memNE((char*)q1, (char*)q2, UTF8SKIP(q1)) ) | |
d7f013c8 | 2004 | return 1; /* mismatch */ |
d07ddd77 | 2005 | n1 -= UTF8SKIP(q1); |
d7f013c8 | 2006 | q1 += UTF8SKIP(q1); |
d07ddd77 | 2007 | n2 -= UTF8SKIP(q2); |
d7f013c8 | 2008 | q2 += UTF8SKIP(q2); |
701a277b | 2009 | } |
d07ddd77 | 2010 | if (n1 == 0) |
d7f013c8 | 2011 | p1 += u1 ? UTF8SKIP(p1) : 1; |
d07ddd77 | 2012 | if (n2 == 0) |
d7f013c8 JH |
2013 | p2 += u2 ? UTF8SKIP(p2) : 1; |
2014 | ||
d2cc3551 | 2015 | } |
5469e704 | 2016 | |
d07ddd77 JH |
2017 | /* A match is defined by all the scans that specified |
2018 | * an explicit length reaching their final goals. */ | |
2019 | match = (f1 == 0 || p1 == f1) && (f2 == 0 || p2 == f2); | |
5469e704 JH |
2020 | |
2021 | if (match) { | |
d07ddd77 JH |
2022 | if (pe1) |
2023 | *pe1 = (char*)p1; | |
2024 | if (pe2) | |
2025 | *pe2 = (char*)p2; | |
5469e704 JH |
2026 | } |
2027 | ||
2028 | return match ? 0 : 1; /* 0 match, 1 mismatch */ | |
e6b2e755 | 2029 | } |
701a277b | 2030 | |
a49f32c6 NC |
2031 | /* |
2032 | * Local variables: | |
2033 | * c-indentation-style: bsd | |
2034 | * c-basic-offset: 4 | |
2035 | * indent-tabs-mode: t | |
2036 | * End: | |
2037 | * | |
2038 | * vim: shiftwidth=4: | |
2039 | */ |