Commit | Line | Data |
---|---|---|
a0ed51b3 LW |
1 | /* utf8.c |
2 | * | |
4bb101f2 | 3 | * Copyright (C) 2000, 2001, 2002, 2003, by Larry Wall and others |
a0ed51b3 LW |
4 | * |
5 | * You may distribute under the terms of either the GNU General Public | |
6 | * License or the Artistic License, as specified in the README file. | |
7 | * | |
8 | */ | |
9 | ||
10 | /* | |
11 | * 'What a fix!' said Sam. 'That's the one place in all the lands we've ever | |
12 | * heard of that we don't want to see any closer; and that's the one place | |
13 | * we're trying to get to! And that's just where we can't get, nohow.' | |
14 | * | |
15 | * 'Well do I understand your speech,' he answered in the same language; | |
16 | * 'yet few strangers do so. Why then do you not speak in the Common Tongue, | |
17 | * as is the custom in the West, if you wish to be answered?' | |
18 | * | |
19 | * ...the travellers perceived that the floor was paved with stones of many | |
20 | * hues; branching runes and strange devices intertwined beneath their feet. | |
21 | */ | |
22 | ||
23 | #include "EXTERN.h" | |
864dbfa3 | 24 | #define PERL_IN_UTF8_C |
a0ed51b3 LW |
25 | #include "perl.h" |
26 | ||
901b21bf JH |
27 | static char unees[] = "Malformed UTF-8 character (unexpected end of string)"; |
28 | ||
ccfc67b7 JH |
29 | /* |
30 | =head1 Unicode Support | |
a0ed51b3 | 31 | |
b851fbc1 | 32 | =for apidoc A|U8 *|uvuni_to_utf8_flags|U8 *d|UV uv|UV flags |
eebe1485 SC |
33 | |
34 | Adds the UTF8 representation of the Unicode codepoint C<uv> to the end | |
35 | of the string C<d>; C<d> should be have at least C<UTF8_MAXLEN+1> free | |
36 | bytes available. The return value is the pointer to the byte after the | |
9041c2e3 | 37 | end of the new character. In other words, |
eebe1485 | 38 | |
b851fbc1 JH |
39 | d = uvuni_to_utf8_flags(d, uv, flags); |
40 | ||
41 | or, in most cases, | |
42 | ||
9041c2e3 | 43 | d = uvuni_to_utf8(d, uv); |
eebe1485 | 44 | |
b851fbc1 JH |
45 | (which is equivalent to) |
46 | ||
47 | d = uvuni_to_utf8_flags(d, uv, 0); | |
48 | ||
eebe1485 SC |
49 | is the recommended Unicode-aware way of saying |
50 | ||
51 | *(d++) = uv; | |
52 | ||
53 | =cut | |
54 | */ | |
55 | ||
dfe13c55 | 56 | U8 * |
b851fbc1 | 57 | Perl_uvuni_to_utf8_flags(pTHX_ U8 *d, UV uv, UV flags) |
a0ed51b3 | 58 | { |
62961d2e | 59 | if (ckWARN(WARN_UTF8)) { |
b851fbc1 JH |
60 | if (UNICODE_IS_SURROGATE(uv) && |
61 | !(flags & UNICODE_ALLOW_SURROGATE)) | |
9014280d | 62 | Perl_warner(aTHX_ packWARN(WARN_UTF8), "UTF-16 surrogate 0x%04"UVxf, uv); |
b851fbc1 JH |
63 | else if ( |
64 | ((uv >= 0xFDD0 && uv <= 0xFDEF && | |
65 | !(flags & UNICODE_ALLOW_FDD0)) | |
66 | || | |
c867b360 | 67 | ((uv & 0xFFFE) == 0xFFFE && /* Either FFFE or FFFF. */ |
b851fbc1 JH |
68 | !(flags & UNICODE_ALLOW_FFFF))) && |
69 | /* UNICODE_ALLOW_SUPER includes | |
2a20b9da | 70 | * FFFEs and FFFFs beyond 0x10FFFF. */ |
b851fbc1 JH |
71 | ((uv <= PERL_UNICODE_MAX) || |
72 | !(flags & UNICODE_ALLOW_SUPER)) | |
73 | ) | |
9014280d | 74 | Perl_warner(aTHX_ packWARN(WARN_UTF8), |
507b9800 JH |
75 | "Unicode character 0x%04"UVxf" is illegal", uv); |
76 | } | |
c4d5f83a | 77 | if (UNI_IS_INVARIANT(uv)) { |
eb160463 | 78 | *d++ = (U8)UTF_TO_NATIVE(uv); |
a0ed51b3 LW |
79 | return d; |
80 | } | |
2d331972 | 81 | #if defined(EBCDIC) |
1d72bdf6 NIS |
82 | else { |
83 | STRLEN len = UNISKIP(uv); | |
84 | U8 *p = d+len-1; | |
85 | while (p > d) { | |
eb160463 | 86 | *p-- = (U8)UTF_TO_NATIVE((uv & UTF_CONTINUATION_MASK) | UTF_CONTINUATION_MARK); |
1d72bdf6 NIS |
87 | uv >>= UTF_ACCUMULATION_SHIFT; |
88 | } | |
eb160463 | 89 | *p = (U8)UTF_TO_NATIVE((uv & UTF_START_MASK(len)) | UTF_START_MARK(len)); |
1d72bdf6 NIS |
90 | return d+len; |
91 | } | |
92 | #else /* Non loop style */ | |
a0ed51b3 | 93 | if (uv < 0x800) { |
eb160463 GS |
94 | *d++ = (U8)(( uv >> 6) | 0xc0); |
95 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
96 | return d; |
97 | } | |
98 | if (uv < 0x10000) { | |
eb160463 GS |
99 | *d++ = (U8)(( uv >> 12) | 0xe0); |
100 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); | |
101 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
102 | return d; |
103 | } | |
104 | if (uv < 0x200000) { | |
eb160463 GS |
105 | *d++ = (U8)(( uv >> 18) | 0xf0); |
106 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); | |
107 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); | |
108 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
109 | return d; |
110 | } | |
111 | if (uv < 0x4000000) { | |
eb160463 GS |
112 | *d++ = (U8)(( uv >> 24) | 0xf8); |
113 | *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80); | |
114 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); | |
115 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); | |
116 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
117 | return d; |
118 | } | |
119 | if (uv < 0x80000000) { | |
eb160463 GS |
120 | *d++ = (U8)(( uv >> 30) | 0xfc); |
121 | *d++ = (U8)(((uv >> 24) & 0x3f) | 0x80); | |
122 | *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80); | |
123 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); | |
124 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); | |
125 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
126 | return d; |
127 | } | |
6b8eaf93 | 128 | #ifdef HAS_QUAD |
d7578b48 | 129 | if (uv < UTF8_QUAD_MAX) |
a0ed51b3 LW |
130 | #endif |
131 | { | |
eb160463 GS |
132 | *d++ = 0xfe; /* Can't match U+FEFF! */ |
133 | *d++ = (U8)(((uv >> 30) & 0x3f) | 0x80); | |
134 | *d++ = (U8)(((uv >> 24) & 0x3f) | 0x80); | |
135 | *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80); | |
136 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); | |
137 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); | |
138 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
139 | return d; |
140 | } | |
6b8eaf93 | 141 | #ifdef HAS_QUAD |
a0ed51b3 | 142 | { |
eb160463 GS |
143 | *d++ = 0xff; /* Can't match U+FFFE! */ |
144 | *d++ = 0x80; /* 6 Reserved bits */ | |
145 | *d++ = (U8)(((uv >> 60) & 0x0f) | 0x80); /* 2 Reserved bits */ | |
146 | *d++ = (U8)(((uv >> 54) & 0x3f) | 0x80); | |
147 | *d++ = (U8)(((uv >> 48) & 0x3f) | 0x80); | |
148 | *d++ = (U8)(((uv >> 42) & 0x3f) | 0x80); | |
149 | *d++ = (U8)(((uv >> 36) & 0x3f) | 0x80); | |
150 | *d++ = (U8)(((uv >> 30) & 0x3f) | 0x80); | |
151 | *d++ = (U8)(((uv >> 24) & 0x3f) | 0x80); | |
152 | *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80); | |
153 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); | |
154 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); | |
155 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
156 | return d; |
157 | } | |
158 | #endif | |
1d72bdf6 | 159 | #endif /* Loop style */ |
a0ed51b3 | 160 | } |
b851fbc1 JH |
161 | |
162 | U8 * | |
163 | Perl_uvuni_to_utf8(pTHX_ U8 *d, UV uv) | |
164 | { | |
165 | return Perl_uvuni_to_utf8_flags(aTHX_ d, uv, 0); | |
166 | } | |
9041c2e3 NIS |
167 | |
168 | ||
169 | /* | |
eebe1485 SC |
170 | =for apidoc A|STRLEN|is_utf8_char|U8 *s |
171 | ||
5da9da9e | 172 | Tests if some arbitrary number of bytes begins in a valid UTF-8 |
82686b01 JH |
173 | character. Note that an INVARIANT (i.e. ASCII) character is a valid |
174 | UTF-8 character. The actual number of bytes in the UTF-8 character | |
175 | will be returned if it is valid, otherwise 0. | |
9041c2e3 | 176 | |
82686b01 | 177 | =cut */ |
067a85ef | 178 | STRLEN |
386d01d6 GS |
179 | Perl_is_utf8_char(pTHX_ U8 *s) |
180 | { | |
181 | U8 u = *s; | |
067a85ef A |
182 | STRLEN slen, len; |
183 | UV uv, ouv; | |
386d01d6 | 184 | |
1d72bdf6 | 185 | if (UTF8_IS_INVARIANT(u)) |
386d01d6 GS |
186 | return 1; |
187 | ||
60006e79 | 188 | if (!UTF8_IS_START(u)) |
386d01d6 GS |
189 | return 0; |
190 | ||
9f07fdcd | 191 | len = UTF8SKIP(s); |
386d01d6 | 192 | |
60006e79 | 193 | if (len < 2 || !UTF8_IS_CONTINUATION(s[1])) |
067a85ef A |
194 | return 0; |
195 | ||
386d01d6 GS |
196 | slen = len - 1; |
197 | s++; | |
6eb6869e | 198 | u &= UTF_START_MASK(len); |
1d72bdf6 | 199 | uv = u; |
067a85ef | 200 | ouv = uv; |
386d01d6 | 201 | while (slen--) { |
60006e79 | 202 | if (!UTF8_IS_CONTINUATION(*s)) |
386d01d6 | 203 | return 0; |
8850bf83 | 204 | uv = UTF8_ACCUMULATE(uv, *s); |
209a9bc1 | 205 | if (uv < ouv) |
067a85ef A |
206 | return 0; |
207 | ouv = uv; | |
386d01d6 GS |
208 | s++; |
209 | } | |
067a85ef | 210 | |
eb160463 | 211 | if ((STRLEN)UNISKIP(uv) < len) |
067a85ef A |
212 | return 0; |
213 | ||
386d01d6 GS |
214 | return len; |
215 | } | |
216 | ||
6662521e | 217 | /* |
eebe1485 | 218 | =for apidoc A|bool|is_utf8_string|U8 *s|STRLEN len |
6662521e | 219 | |
c9ada85f JH |
220 | Returns true if first C<len> bytes of the given string form a valid |
221 | UTF8 string, false otherwise. Note that 'a valid UTF8 string' does | |
222 | not mean 'a string that contains code points above 0x7F encoded in | |
223 | UTF8' because a valid ASCII string is a valid UTF8 string. | |
6662521e GS |
224 | |
225 | =cut | |
226 | */ | |
227 | ||
8e84507e | 228 | bool |
6662521e GS |
229 | Perl_is_utf8_string(pTHX_ U8 *s, STRLEN len) |
230 | { | |
067a85ef | 231 | U8* x = s; |
1aa99e6b | 232 | U8* send; |
067a85ef A |
233 | STRLEN c; |
234 | ||
1aa99e6b | 235 | if (!len) |
6cd5fe39 | 236 | len = strlen((char *)s); |
1aa99e6b IH |
237 | send = s + len; |
238 | ||
6662521e | 239 | while (x < send) { |
1acdb0da JH |
240 | /* Inline the easy bits of is_utf8_char() here for speed... */ |
241 | if (UTF8_IS_INVARIANT(*x)) | |
242 | c = 1; | |
243 | else if (!UTF8_IS_START(*x)) | |
244 | return FALSE; | |
245 | else { | |
246 | /* ... and call is_utf8_char() only if really needed. */ | |
247 | c = is_utf8_char(x); | |
248 | if (!c) | |
249 | return FALSE; | |
250 | } | |
6662521e | 251 | x += c; |
6662521e | 252 | } |
60006e79 JH |
253 | if (x != send) |
254 | return FALSE; | |
067a85ef A |
255 | |
256 | return TRUE; | |
6662521e GS |
257 | } |
258 | ||
67e989fb | 259 | /* |
9041c2e3 | 260 | =for apidoc A|UV|utf8n_to_uvuni|U8 *s|STRLEN curlen|STRLEN *retlen|U32 flags |
67e989fb | 261 | |
9041c2e3 NIS |
262 | Bottom level UTF-8 decode routine. |
263 | Returns the unicode code point value of the first character in the string C<s> | |
ba210ebe | 264 | which is assumed to be in UTF8 encoding and no longer than C<curlen>; |
7df053ec | 265 | C<retlen> will be set to the length, in bytes, of that character. |
67e989fb JH |
266 | |
267 | If C<s> does not point to a well-formed UTF8 character, the behaviour | |
dcad2880 JH |
268 | is dependent on the value of C<flags>: if it contains UTF8_CHECK_ONLY, |
269 | it is assumed that the caller will raise a warning, and this function | |
28d3d195 JH |
270 | will silently just set C<retlen> to C<-1> and return zero. If the |
271 | C<flags> does not contain UTF8_CHECK_ONLY, warnings about | |
272 | malformations will be given, C<retlen> will be set to the expected | |
273 | length of the UTF-8 character in bytes, and zero will be returned. | |
274 | ||
275 | The C<flags> can also contain various flags to allow deviations from | |
276 | the strict UTF-8 encoding (see F<utf8.h>). | |
67e989fb | 277 | |
9041c2e3 NIS |
278 | Most code should use utf8_to_uvchr() rather than call this directly. |
279 | ||
37607a96 PK |
280 | =cut |
281 | */ | |
67e989fb | 282 | |
a0ed51b3 | 283 | UV |
37607a96 | 284 | Perl_utf8n_to_uvuni(pTHX_ U8 *s, STRLEN curlen, STRLEN *retlen, U32 flags) |
a0ed51b3 | 285 | { |
097fb8e2 | 286 | U8 *s0 = s; |
9c5ffd7c | 287 | UV uv = *s, ouv = 0; |
ba210ebe JH |
288 | STRLEN len = 1; |
289 | bool dowarn = ckWARN_d(WARN_UTF8); | |
4f849cb6 | 290 | UV startbyte = *s; |
ba210ebe | 291 | STRLEN expectlen = 0; |
a0dbb045 JH |
292 | U32 warning = 0; |
293 | ||
294 | /* This list is a superset of the UTF8_ALLOW_XXX. */ | |
295 | ||
296 | #define UTF8_WARN_EMPTY 1 | |
297 | #define UTF8_WARN_CONTINUATION 2 | |
298 | #define UTF8_WARN_NON_CONTINUATION 3 | |
299 | #define UTF8_WARN_FE_FF 4 | |
300 | #define UTF8_WARN_SHORT 5 | |
301 | #define UTF8_WARN_OVERFLOW 6 | |
302 | #define UTF8_WARN_SURROGATE 7 | |
c867b360 JH |
303 | #define UTF8_WARN_LONG 8 |
304 | #define UTF8_WARN_FFFF 9 /* Also FFFE. */ | |
a0dbb045 JH |
305 | |
306 | if (curlen == 0 && | |
307 | !(flags & UTF8_ALLOW_EMPTY)) { | |
308 | warning = UTF8_WARN_EMPTY; | |
0c443dc2 JH |
309 | goto malformed; |
310 | } | |
311 | ||
1d72bdf6 | 312 | if (UTF8_IS_INVARIANT(uv)) { |
a0ed51b3 LW |
313 | if (retlen) |
314 | *retlen = 1; | |
c4d5f83a | 315 | return (UV) (NATIVE_TO_UTF(*s)); |
a0ed51b3 | 316 | } |
67e989fb | 317 | |
421a8bf2 | 318 | if (UTF8_IS_CONTINUATION(uv) && |
fcc8fcf6 | 319 | !(flags & UTF8_ALLOW_CONTINUATION)) { |
a0dbb045 | 320 | warning = UTF8_WARN_CONTINUATION; |
ba210ebe JH |
321 | goto malformed; |
322 | } | |
323 | ||
421a8bf2 | 324 | if (UTF8_IS_START(uv) && curlen > 1 && !UTF8_IS_CONTINUATION(s[1]) && |
fcc8fcf6 | 325 | !(flags & UTF8_ALLOW_NON_CONTINUATION)) { |
a0dbb045 | 326 | warning = UTF8_WARN_NON_CONTINUATION; |
ba210ebe JH |
327 | goto malformed; |
328 | } | |
9041c2e3 | 329 | |
1d72bdf6 | 330 | #ifdef EBCDIC |
75383841 | 331 | uv = NATIVE_TO_UTF(uv); |
1d72bdf6 | 332 | #else |
fcc8fcf6 JH |
333 | if ((uv == 0xfe || uv == 0xff) && |
334 | !(flags & UTF8_ALLOW_FE_FF)) { | |
a0dbb045 | 335 | warning = UTF8_WARN_FE_FF; |
ba210ebe | 336 | goto malformed; |
a0ed51b3 | 337 | } |
1d72bdf6 NIS |
338 | #endif |
339 | ||
ba210ebe JH |
340 | if (!(uv & 0x20)) { len = 2; uv &= 0x1f; } |
341 | else if (!(uv & 0x10)) { len = 3; uv &= 0x0f; } | |
342 | else if (!(uv & 0x08)) { len = 4; uv &= 0x07; } | |
343 | else if (!(uv & 0x04)) { len = 5; uv &= 0x03; } | |
1d72bdf6 NIS |
344 | #ifdef EBCDIC |
345 | else if (!(uv & 0x02)) { len = 6; uv &= 0x01; } | |
346 | else { len = 7; uv &= 0x01; } | |
347 | #else | |
ba210ebe JH |
348 | else if (!(uv & 0x02)) { len = 6; uv &= 0x01; } |
349 | else if (!(uv & 0x01)) { len = 7; uv = 0; } | |
1d72bdf6 NIS |
350 | else { len = 13; uv = 0; } /* whoa! */ |
351 | #endif | |
352 | ||
a0ed51b3 LW |
353 | if (retlen) |
354 | *retlen = len; | |
9041c2e3 | 355 | |
ba210ebe JH |
356 | expectlen = len; |
357 | ||
fcc8fcf6 JH |
358 | if ((curlen < expectlen) && |
359 | !(flags & UTF8_ALLOW_SHORT)) { | |
a0dbb045 | 360 | warning = UTF8_WARN_SHORT; |
ba210ebe JH |
361 | goto malformed; |
362 | } | |
363 | ||
364 | len--; | |
a0ed51b3 | 365 | s++; |
ba210ebe JH |
366 | ouv = uv; |
367 | ||
a0ed51b3 | 368 | while (len--) { |
421a8bf2 JH |
369 | if (!UTF8_IS_CONTINUATION(*s) && |
370 | !(flags & UTF8_ALLOW_NON_CONTINUATION)) { | |
a0dbb045 JH |
371 | s--; |
372 | warning = UTF8_WARN_NON_CONTINUATION; | |
ba210ebe | 373 | goto malformed; |
a0ed51b3 LW |
374 | } |
375 | else | |
8850bf83 | 376 | uv = UTF8_ACCUMULATE(uv, *s); |
a0dbb045 JH |
377 | if (!(uv > ouv)) { |
378 | /* These cannot be allowed. */ | |
379 | if (uv == ouv) { | |
380 | if (!(flags & UTF8_ALLOW_LONG)) { | |
381 | warning = UTF8_WARN_LONG; | |
382 | goto malformed; | |
383 | } | |
384 | } | |
385 | else { /* uv < ouv */ | |
386 | /* This cannot be allowed. */ | |
387 | warning = UTF8_WARN_OVERFLOW; | |
388 | goto malformed; | |
389 | } | |
ba210ebe JH |
390 | } |
391 | s++; | |
392 | ouv = uv; | |
393 | } | |
394 | ||
421a8bf2 | 395 | if (UNICODE_IS_SURROGATE(uv) && |
fcc8fcf6 | 396 | !(flags & UTF8_ALLOW_SURROGATE)) { |
a0dbb045 | 397 | warning = UTF8_WARN_SURROGATE; |
ba210ebe | 398 | goto malformed; |
eb160463 | 399 | } else if ((expectlen > (STRLEN)UNISKIP(uv)) && |
fcc8fcf6 | 400 | !(flags & UTF8_ALLOW_LONG)) { |
a0dbb045 | 401 | warning = UTF8_WARN_LONG; |
ba210ebe | 402 | goto malformed; |
421a8bf2 | 403 | } else if (UNICODE_IS_ILLEGAL(uv) && |
a9917092 | 404 | !(flags & UTF8_ALLOW_FFFF)) { |
a0dbb045 | 405 | warning = UTF8_WARN_FFFF; |
a9917092 | 406 | goto malformed; |
a0ed51b3 | 407 | } |
ba210ebe | 408 | |
a0ed51b3 | 409 | return uv; |
ba210ebe JH |
410 | |
411 | malformed: | |
412 | ||
fcc8fcf6 | 413 | if (flags & UTF8_CHECK_ONLY) { |
ba210ebe | 414 | if (retlen) |
cc366d4b | 415 | *retlen = -1; |
ba210ebe JH |
416 | return 0; |
417 | } | |
418 | ||
a0dbb045 JH |
419 | if (dowarn) { |
420 | SV* sv = sv_2mortal(newSVpv("Malformed UTF-8 character ", 0)); | |
421 | ||
422 | switch (warning) { | |
423 | case 0: /* Intentionally empty. */ break; | |
424 | case UTF8_WARN_EMPTY: | |
425 | Perl_sv_catpvf(aTHX_ sv, "(empty string)"); | |
426 | break; | |
427 | case UTF8_WARN_CONTINUATION: | |
097fb8e2 | 428 | Perl_sv_catpvf(aTHX_ sv, "(unexpected continuation byte 0x%02"UVxf", with no preceding start byte)", uv); |
a0dbb045 JH |
429 | break; |
430 | case UTF8_WARN_NON_CONTINUATION: | |
097fb8e2 JH |
431 | if (s == s0) |
432 | Perl_sv_catpvf(aTHX_ sv, "(unexpected non-continuation byte 0x%02"UVxf", immediately after start byte 0x%02"UVxf")", | |
433 | (UV)s[1], startbyte); | |
434 | else | |
435 | Perl_sv_catpvf(aTHX_ sv, "(unexpected non-continuation byte 0x%02"UVxf", %d byte%s after start byte 0x%02"UVxf", expected %d bytes)", | |
436 | (UV)s[1], s - s0, s - s0 > 1 ? "s" : "", startbyte, expectlen); | |
437 | ||
a0dbb045 JH |
438 | break; |
439 | case UTF8_WARN_FE_FF: | |
440 | Perl_sv_catpvf(aTHX_ sv, "(byte 0x%02"UVxf")", uv); | |
441 | break; | |
442 | case UTF8_WARN_SHORT: | |
097fb8e2 JH |
443 | Perl_sv_catpvf(aTHX_ sv, "(%d byte%s, need %d, after start byte 0x%02"UVxf")", |
444 | curlen, curlen == 1 ? "" : "s", expectlen, startbyte); | |
b31f83c2 | 445 | expectlen = curlen; /* distance for caller to skip */ |
a0dbb045 JH |
446 | break; |
447 | case UTF8_WARN_OVERFLOW: | |
097fb8e2 JH |
448 | Perl_sv_catpvf(aTHX_ sv, "(overflow at 0x%"UVxf", byte 0x%02x, after start byte 0x%02"UVxf")", |
449 | ouv, *s, startbyte); | |
a0dbb045 JH |
450 | break; |
451 | case UTF8_WARN_SURROGATE: | |
452 | Perl_sv_catpvf(aTHX_ sv, "(UTF-16 surrogate 0x%04"UVxf")", uv); | |
453 | break; | |
a0dbb045 | 454 | case UTF8_WARN_LONG: |
097fb8e2 JH |
455 | Perl_sv_catpvf(aTHX_ sv, "(%d byte%s, need %d, after start byte 0x%02"UVxf")", |
456 | expectlen, expectlen == 1 ? "": "s", UNISKIP(uv), startbyte); | |
a0dbb045 JH |
457 | break; |
458 | case UTF8_WARN_FFFF: | |
459 | Perl_sv_catpvf(aTHX_ sv, "(character 0x%04"UVxf")", uv); | |
460 | break; | |
461 | default: | |
462 | Perl_sv_catpvf(aTHX_ sv, "(unknown reason)"); | |
463 | break; | |
464 | } | |
465 | ||
466 | if (warning) { | |
467 | char *s = SvPVX(sv); | |
468 | ||
469 | if (PL_op) | |
9014280d | 470 | Perl_warner(aTHX_ packWARN(WARN_UTF8), |
53e06cf0 | 471 | "%s in %s", s, OP_DESC(PL_op)); |
a0dbb045 | 472 | else |
9014280d | 473 | Perl_warner(aTHX_ packWARN(WARN_UTF8), "%s", s); |
a0dbb045 JH |
474 | } |
475 | } | |
476 | ||
ba210ebe | 477 | if (retlen) |
28d3d195 | 478 | *retlen = expectlen ? expectlen : len; |
ba210ebe | 479 | |
28d3d195 | 480 | return 0; |
a0ed51b3 LW |
481 | } |
482 | ||
8e84507e | 483 | /* |
37607a96 | 484 | =for apidoc A|UV|utf8_to_uvchr|U8 *s|STRLEN *retlen |
9041c2e3 NIS |
485 | |
486 | Returns the native character value of the first character in the string C<s> | |
487 | which is assumed to be in UTF8 encoding; C<retlen> will be set to the | |
488 | length, in bytes, of that character. | |
489 | ||
490 | If C<s> does not point to a well-formed UTF8 character, zero is | |
491 | returned and retlen is set, if possible, to -1. | |
492 | ||
493 | =cut | |
494 | */ | |
495 | ||
496 | UV | |
37607a96 | 497 | Perl_utf8_to_uvchr(pTHX_ U8 *s, STRLEN *retlen) |
9041c2e3 | 498 | { |
872c91ae JH |
499 | return Perl_utf8n_to_uvchr(aTHX_ s, UTF8_MAXLEN, retlen, |
500 | ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY); | |
9041c2e3 NIS |
501 | } |
502 | ||
503 | /* | |
37607a96 | 504 | =for apidoc A|UV|utf8_to_uvuni|U8 *s|STRLEN *retlen |
9041c2e3 NIS |
505 | |
506 | Returns the Unicode code point of the first character in the string C<s> | |
507 | which is assumed to be in UTF8 encoding; C<retlen> will be set to the | |
508 | length, in bytes, of that character. | |
509 | ||
510 | This function should only be used when returned UV is considered | |
511 | an index into the Unicode semantic tables (e.g. swashes). | |
512 | ||
ba210ebe JH |
513 | If C<s> does not point to a well-formed UTF8 character, zero is |
514 | returned and retlen is set, if possible, to -1. | |
8e84507e NIS |
515 | |
516 | =cut | |
517 | */ | |
518 | ||
519 | UV | |
37607a96 | 520 | Perl_utf8_to_uvuni(pTHX_ U8 *s, STRLEN *retlen) |
8e84507e | 521 | { |
9041c2e3 | 522 | /* Call the low level routine asking for checks */ |
872c91ae JH |
523 | return Perl_utf8n_to_uvuni(aTHX_ s, UTF8_MAXLEN, retlen, |
524 | ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY); | |
8e84507e NIS |
525 | } |
526 | ||
b76347f2 | 527 | /* |
37607a96 | 528 | =for apidoc A|STRLEN|utf8_length|U8 *s|U8 *e |
b76347f2 JH |
529 | |
530 | Return the length of the UTF-8 char encoded string C<s> in characters. | |
02eb7b47 JH |
531 | Stops at C<e> (inclusive). If C<e E<lt> s> or if the scan would end |
532 | up past C<e>, croaks. | |
b76347f2 JH |
533 | |
534 | =cut | |
535 | */ | |
536 | ||
537 | STRLEN | |
37607a96 | 538 | Perl_utf8_length(pTHX_ U8 *s, U8 *e) |
b76347f2 JH |
539 | { |
540 | STRLEN len = 0; | |
541 | ||
8850bf83 JH |
542 | /* Note: cannot use UTF8_IS_...() too eagerly here since e.g. |
543 | * the bitops (especially ~) can create illegal UTF-8. | |
544 | * In other words: in Perl UTF-8 is not just for Unicode. */ | |
545 | ||
901b21bf JH |
546 | if (e < s) { |
547 | if (ckWARN_d(WARN_UTF8)) { | |
548 | if (PL_op) | |
549 | Perl_warner(aTHX_ packWARN(WARN_UTF8), | |
550 | "%s in %s", unees, OP_DESC(PL_op)); | |
551 | else | |
552 | Perl_warner(aTHX_ packWARN(WARN_UTF8), unees); | |
553 | } | |
554 | return 0; | |
555 | } | |
b76347f2 | 556 | while (s < e) { |
02eb7b47 | 557 | U8 t = UTF8SKIP(s); |
b76347f2 | 558 | |
901b21bf JH |
559 | if (e - s < t) { |
560 | if (ckWARN_d(WARN_UTF8)) { | |
561 | if (PL_op) | |
562 | Perl_warner(aTHX_ packWARN(WARN_UTF8), | |
563 | unees, OP_DESC(PL_op)); | |
564 | else | |
565 | Perl_warner(aTHX_ packWARN(WARN_UTF8), unees); | |
566 | } | |
567 | return len; | |
568 | } | |
b76347f2 JH |
569 | s += t; |
570 | len++; | |
571 | } | |
572 | ||
573 | return len; | |
574 | } | |
575 | ||
b06226ff | 576 | /* |
eebe1485 | 577 | =for apidoc A|IV|utf8_distance|U8 *a|U8 *b |
b06226ff JH |
578 | |
579 | Returns the number of UTF8 characters between the UTF-8 pointers C<a> | |
580 | and C<b>. | |
581 | ||
582 | WARNING: use only if you *know* that the pointers point inside the | |
583 | same UTF-8 buffer. | |
584 | ||
37607a96 PK |
585 | =cut |
586 | */ | |
a0ed51b3 | 587 | |
02eb7b47 | 588 | IV |
864dbfa3 | 589 | Perl_utf8_distance(pTHX_ U8 *a, U8 *b) |
a0ed51b3 | 590 | { |
02eb7b47 JH |
591 | IV off = 0; |
592 | ||
8850bf83 JH |
593 | /* Note: cannot use UTF8_IS_...() too eagerly here since e.g. |
594 | * the bitops (especially ~) can create illegal UTF-8. | |
595 | * In other words: in Perl UTF-8 is not just for Unicode. */ | |
596 | ||
a0ed51b3 LW |
597 | if (a < b) { |
598 | while (a < b) { | |
02eb7b47 JH |
599 | U8 c = UTF8SKIP(a); |
600 | ||
901b21bf JH |
601 | if (b - a < c) { |
602 | if (ckWARN_d(WARN_UTF8)) { | |
603 | if (PL_op) | |
604 | Perl_warner(aTHX_ packWARN(WARN_UTF8), | |
605 | "%s in %s", unees, OP_DESC(PL_op)); | |
606 | else | |
607 | Perl_warner(aTHX_ packWARN(WARN_UTF8), unees); | |
608 | } | |
609 | return off; | |
610 | } | |
02eb7b47 | 611 | a += c; |
a0ed51b3 LW |
612 | off--; |
613 | } | |
614 | } | |
615 | else { | |
616 | while (b < a) { | |
02eb7b47 JH |
617 | U8 c = UTF8SKIP(b); |
618 | ||
901b21bf JH |
619 | if (a - b < c) { |
620 | if (ckWARN_d(WARN_UTF8)) { | |
621 | if (PL_op) | |
622 | Perl_warner(aTHX_ packWARN(WARN_UTF8), | |
623 | "%s in %s", unees, OP_DESC(PL_op)); | |
624 | else | |
625 | Perl_warner(aTHX_ packWARN(WARN_UTF8), unees); | |
626 | } | |
627 | return off; | |
628 | } | |
02eb7b47 | 629 | b += c; |
a0ed51b3 LW |
630 | off++; |
631 | } | |
632 | } | |
02eb7b47 | 633 | |
a0ed51b3 LW |
634 | return off; |
635 | } | |
636 | ||
b06226ff | 637 | /* |
37607a96 | 638 | =for apidoc A|U8 *|utf8_hop|U8 *s|I32 off |
b06226ff | 639 | |
8850bf83 JH |
640 | Return the UTF-8 pointer C<s> displaced by C<off> characters, either |
641 | forward or backward. | |
b06226ff JH |
642 | |
643 | WARNING: do not use the following unless you *know* C<off> is within | |
8850bf83 JH |
644 | the UTF-8 data pointed to by C<s> *and* that on entry C<s> is aligned |
645 | on the first byte of character or just after the last byte of a character. | |
b06226ff | 646 | |
37607a96 PK |
647 | =cut |
648 | */ | |
a0ed51b3 LW |
649 | |
650 | U8 * | |
864dbfa3 | 651 | Perl_utf8_hop(pTHX_ U8 *s, I32 off) |
a0ed51b3 | 652 | { |
8850bf83 JH |
653 | /* Note: cannot use UTF8_IS_...() too eagerly here since e.g |
654 | * the bitops (especially ~) can create illegal UTF-8. | |
655 | * In other words: in Perl UTF-8 is not just for Unicode. */ | |
656 | ||
a0ed51b3 LW |
657 | if (off >= 0) { |
658 | while (off--) | |
659 | s += UTF8SKIP(s); | |
660 | } | |
661 | else { | |
662 | while (off++) { | |
663 | s--; | |
8850bf83 JH |
664 | while (UTF8_IS_CONTINUATION(*s)) |
665 | s--; | |
a0ed51b3 LW |
666 | } |
667 | } | |
668 | return s; | |
669 | } | |
670 | ||
6940069f | 671 | /* |
eebe1485 | 672 | =for apidoc A|U8 *|utf8_to_bytes|U8 *s|STRLEN *len |
6940069f | 673 | |
246fae53 MG |
674 | Converts a string C<s> of length C<len> from UTF8 into byte encoding. |
675 | Unlike C<bytes_to_utf8>, this over-writes the original string, and | |
676 | updates len to contain the new length. | |
67e989fb | 677 | Returns zero on failure, setting C<len> to -1. |
6940069f GS |
678 | |
679 | =cut | |
680 | */ | |
681 | ||
682 | U8 * | |
37607a96 | 683 | Perl_utf8_to_bytes(pTHX_ U8 *s, STRLEN *len) |
6940069f | 684 | { |
6940069f GS |
685 | U8 *send; |
686 | U8 *d; | |
dcad2880 | 687 | U8 *save = s; |
246fae53 MG |
688 | |
689 | /* ensure valid UTF8 and chars < 256 before updating string */ | |
dcad2880 JH |
690 | for (send = s + *len; s < send; ) { |
691 | U8 c = *s++; | |
692 | ||
1d72bdf6 NIS |
693 | if (!UTF8_IS_INVARIANT(c) && |
694 | (!UTF8_IS_DOWNGRADEABLE_START(c) || (s >= send) | |
695 | || !(c = *s++) || !UTF8_IS_CONTINUATION(c))) { | |
dcad2880 JH |
696 | *len = -1; |
697 | return 0; | |
698 | } | |
246fae53 | 699 | } |
dcad2880 JH |
700 | |
701 | d = s = save; | |
6940069f | 702 | while (s < send) { |
ed646e6e | 703 | STRLEN ulen; |
9041c2e3 | 704 | *d++ = (U8)utf8_to_uvchr(s, &ulen); |
ed646e6e | 705 | s += ulen; |
6940069f GS |
706 | } |
707 | *d = '\0'; | |
246fae53 | 708 | *len = d - save; |
6940069f GS |
709 | return save; |
710 | } | |
711 | ||
712 | /* | |
f9a63242 JH |
713 | =for apidoc A|U8 *|bytes_from_utf8|U8 *s|STRLEN *len|bool *is_utf8 |
714 | ||
715 | Converts a string C<s> of length C<len> from UTF8 into byte encoding. | |
716 | Unlike <utf8_to_bytes> but like C<bytes_to_utf8>, returns a pointer to | |
ef9edfd0 JH |
717 | the newly-created string, and updates C<len> to contain the new |
718 | length. Returns the original string if no conversion occurs, C<len> | |
719 | is unchanged. Do nothing if C<is_utf8> points to 0. Sets C<is_utf8> to | |
720 | 0 if C<s> is converted or contains all 7bit characters. | |
f9a63242 | 721 | |
37607a96 PK |
722 | =cut |
723 | */ | |
f9a63242 JH |
724 | |
725 | U8 * | |
37607a96 | 726 | Perl_bytes_from_utf8(pTHX_ U8 *s, STRLEN *len, bool *is_utf8) |
f9a63242 | 727 | { |
f9a63242 JH |
728 | U8 *d; |
729 | U8 *start = s; | |
db42d148 | 730 | U8 *send; |
f9a63242 JH |
731 | I32 count = 0; |
732 | ||
733 | if (!*is_utf8) | |
734 | return start; | |
735 | ||
ef9edfd0 | 736 | /* ensure valid UTF8 and chars < 256 before converting string */ |
f9a63242 JH |
737 | for (send = s + *len; s < send;) { |
738 | U8 c = *s++; | |
1d72bdf6 | 739 | if (!UTF8_IS_INVARIANT(c)) { |
db42d148 NIS |
740 | if (UTF8_IS_DOWNGRADEABLE_START(c) && s < send && |
741 | (c = *s++) && UTF8_IS_CONTINUATION(c)) | |
742 | count++; | |
743 | else | |
f9a63242 | 744 | return start; |
db42d148 | 745 | } |
f9a63242 JH |
746 | } |
747 | ||
748 | *is_utf8 = 0; | |
749 | ||
f9a63242 | 750 | Newz(801, d, (*len) - count + 1, U8); |
ef9edfd0 | 751 | s = start; start = d; |
f9a63242 JH |
752 | while (s < send) { |
753 | U8 c = *s++; | |
c4d5f83a NIS |
754 | if (!UTF8_IS_INVARIANT(c)) { |
755 | /* Then it is two-byte encoded */ | |
756 | c = UTF8_ACCUMULATE(NATIVE_TO_UTF(c), *s++); | |
757 | c = ASCII_TO_NATIVE(c); | |
758 | } | |
759 | *d++ = c; | |
f9a63242 JH |
760 | } |
761 | *d = '\0'; | |
762 | *len = d - start; | |
763 | return start; | |
764 | } | |
765 | ||
766 | /* | |
eebe1485 | 767 | =for apidoc A|U8 *|bytes_to_utf8|U8 *s|STRLEN *len |
6940069f GS |
768 | |
769 | Converts a string C<s> of length C<len> from ASCII into UTF8 encoding. | |
6662521e GS |
770 | Returns a pointer to the newly-created string, and sets C<len> to |
771 | reflect the new length. | |
6940069f | 772 | |
c9ada85f JH |
773 | If you want to convert to UTF8 from other encodings than ASCII, |
774 | see sv_recode_to_utf8(). | |
775 | ||
497711e7 | 776 | =cut |
6940069f GS |
777 | */ |
778 | ||
779 | U8* | |
37607a96 | 780 | Perl_bytes_to_utf8(pTHX_ U8 *s, STRLEN *len) |
6940069f | 781 | { |
6940069f GS |
782 | U8 *send; |
783 | U8 *d; | |
784 | U8 *dst; | |
6662521e | 785 | send = s + (*len); |
6940069f | 786 | |
6662521e | 787 | Newz(801, d, (*len) * 2 + 1, U8); |
6940069f GS |
788 | dst = d; |
789 | ||
790 | while (s < send) { | |
db42d148 | 791 | UV uv = NATIVE_TO_ASCII(*s++); |
c4d5f83a | 792 | if (UNI_IS_INVARIANT(uv)) |
eb160463 | 793 | *d++ = (U8)UTF_TO_NATIVE(uv); |
6940069f | 794 | else { |
eb160463 GS |
795 | *d++ = (U8)UTF8_EIGHT_BIT_HI(uv); |
796 | *d++ = (U8)UTF8_EIGHT_BIT_LO(uv); | |
6940069f GS |
797 | } |
798 | } | |
799 | *d = '\0'; | |
6662521e | 800 | *len = d-dst; |
6940069f GS |
801 | return dst; |
802 | } | |
803 | ||
a0ed51b3 | 804 | /* |
dea0fc0b | 805 | * Convert native (big-endian) or reversed (little-endian) UTF-16 to UTF-8. |
a0ed51b3 LW |
806 | * |
807 | * Destination must be pre-extended to 3/2 source. Do not use in-place. | |
808 | * We optimize for native, for obvious reasons. */ | |
809 | ||
810 | U8* | |
dea0fc0b | 811 | Perl_utf16_to_utf8(pTHX_ U8* p, U8* d, I32 bytelen, I32 *newlen) |
a0ed51b3 | 812 | { |
dea0fc0b JH |
813 | U8* pend; |
814 | U8* dstart = d; | |
815 | ||
816 | if (bytelen & 1) | |
a7867d0a | 817 | Perl_croak(aTHX_ "panic: utf16_to_utf8: odd bytelen"); |
dea0fc0b JH |
818 | |
819 | pend = p + bytelen; | |
820 | ||
a0ed51b3 | 821 | while (p < pend) { |
dea0fc0b JH |
822 | UV uv = (p[0] << 8) + p[1]; /* UTF-16BE */ |
823 | p += 2; | |
a0ed51b3 | 824 | if (uv < 0x80) { |
eb160463 | 825 | *d++ = (U8)uv; |
a0ed51b3 LW |
826 | continue; |
827 | } | |
828 | if (uv < 0x800) { | |
eb160463 GS |
829 | *d++ = (U8)(( uv >> 6) | 0xc0); |
830 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
831 | continue; |
832 | } | |
833 | if (uv >= 0xd800 && uv < 0xdbff) { /* surrogates */ | |
dea0fc0b JH |
834 | UV low = *p++; |
835 | if (low < 0xdc00 || low >= 0xdfff) | |
836 | Perl_croak(aTHX_ "Malformed UTF-16 surrogate"); | |
a0ed51b3 LW |
837 | uv = ((uv - 0xd800) << 10) + (low - 0xdc00) + 0x10000; |
838 | } | |
839 | if (uv < 0x10000) { | |
eb160463 GS |
840 | *d++ = (U8)(( uv >> 12) | 0xe0); |
841 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); | |
842 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
843 | continue; |
844 | } | |
845 | else { | |
eb160463 GS |
846 | *d++ = (U8)(( uv >> 18) | 0xf0); |
847 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); | |
848 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); | |
849 | *d++ = (U8)(( uv & 0x3f) | 0x80); | |
a0ed51b3 LW |
850 | continue; |
851 | } | |
852 | } | |
dea0fc0b | 853 | *newlen = d - dstart; |
a0ed51b3 LW |
854 | return d; |
855 | } | |
856 | ||
857 | /* Note: this one is slightly destructive of the source. */ | |
858 | ||
859 | U8* | |
dea0fc0b | 860 | Perl_utf16_to_utf8_reversed(pTHX_ U8* p, U8* d, I32 bytelen, I32 *newlen) |
a0ed51b3 LW |
861 | { |
862 | U8* s = (U8*)p; | |
863 | U8* send = s + bytelen; | |
864 | while (s < send) { | |
865 | U8 tmp = s[0]; | |
866 | s[0] = s[1]; | |
867 | s[1] = tmp; | |
868 | s += 2; | |
869 | } | |
dea0fc0b | 870 | return utf16_to_utf8(p, d, bytelen, newlen); |
a0ed51b3 LW |
871 | } |
872 | ||
873 | /* for now these are all defined (inefficiently) in terms of the utf8 versions */ | |
874 | ||
875 | bool | |
84afefe6 | 876 | Perl_is_uni_alnum(pTHX_ UV c) |
a0ed51b3 | 877 | { |
ad391ad9 | 878 | U8 tmpbuf[UTF8_MAXLEN+1]; |
230880c1 | 879 | uvchr_to_utf8(tmpbuf, c); |
a0ed51b3 LW |
880 | return is_utf8_alnum(tmpbuf); |
881 | } | |
882 | ||
883 | bool | |
84afefe6 | 884 | Perl_is_uni_alnumc(pTHX_ UV c) |
b8c5462f | 885 | { |
ad391ad9 | 886 | U8 tmpbuf[UTF8_MAXLEN+1]; |
230880c1 | 887 | uvchr_to_utf8(tmpbuf, c); |
b8c5462f JH |
888 | return is_utf8_alnumc(tmpbuf); |
889 | } | |
890 | ||
891 | bool | |
84afefe6 | 892 | Perl_is_uni_idfirst(pTHX_ UV c) |
a0ed51b3 | 893 | { |
ad391ad9 | 894 | U8 tmpbuf[UTF8_MAXLEN+1]; |
230880c1 | 895 | uvchr_to_utf8(tmpbuf, c); |
a0ed51b3 LW |
896 | return is_utf8_idfirst(tmpbuf); |
897 | } | |
898 | ||
899 | bool | |
84afefe6 | 900 | Perl_is_uni_alpha(pTHX_ UV c) |
a0ed51b3 | 901 | { |
ad391ad9 | 902 | U8 tmpbuf[UTF8_MAXLEN+1]; |
230880c1 | 903 | uvchr_to_utf8(tmpbuf, c); |
a0ed51b3 LW |
904 | return is_utf8_alpha(tmpbuf); |
905 | } | |
906 | ||
907 | bool | |
84afefe6 | 908 | Perl_is_uni_ascii(pTHX_ UV c) |
4d61ec05 | 909 | { |
ad391ad9 | 910 | U8 tmpbuf[UTF8_MAXLEN+1]; |
230880c1 | 911 | uvchr_to_utf8(tmpbuf, c); |
4d61ec05 GS |
912 | return is_utf8_ascii(tmpbuf); |
913 | } | |
914 | ||
915 | bool | |
84afefe6 | 916 | Perl_is_uni_space(pTHX_ UV c) |
a0ed51b3 | 917 | { |
ad391ad9 | 918 | U8 tmpbuf[UTF8_MAXLEN+1]; |
230880c1 | 919 | uvchr_to_utf8(tmpbuf, c); |
a0ed51b3 LW |
920 | return is_utf8_space(tmpbuf); |
921 | } | |
922 | ||
923 | bool | |
84afefe6 | 924 | Perl_is_uni_digit(pTHX_ UV c) |
a0ed51b3 | 925 | { |
ad391ad9 | 926 | U8 tmpbuf[UTF8_MAXLEN+1]; |
230880c1 | 927 | uvchr_to_utf8(tmpbuf, c); |
a0ed51b3 LW |
928 | return is_utf8_digit(tmpbuf); |
929 | } | |
930 | ||
931 | bool | |
84afefe6 | 932 | Perl_is_uni_upper(pTHX_ UV c) |
a0ed51b3 | 933 | { |
ad391ad9 | 934 | U8 tmpbuf[UTF8_MAXLEN+1]; |
230880c1 | 935 | uvchr_to_utf8(tmpbuf, c); |
a0ed51b3 LW |
936 | return is_utf8_upper(tmpbuf); |
937 | } | |
938 | ||
939 | bool | |
84afefe6 | 940 | Perl_is_uni_lower(pTHX_ UV c) |
a0ed51b3 | 941 | { |
ad391ad9 | 942 | U8 tmpbuf[UTF8_MAXLEN+1]; |
230880c1 | 943 | uvchr_to_utf8(tmpbuf, c); |
a0ed51b3 LW |
944 | return is_utf8_lower(tmpbuf); |
945 | } | |
946 | ||
947 | bool | |
84afefe6 | 948 | Perl_is_uni_cntrl(pTHX_ UV c) |
b8c5462f | 949 | { |
ad391ad9 | 950 | U8 tmpbuf[UTF8_MAXLEN+1]; |
230880c1 | 951 | uvchr_to_utf8(tmpbuf, c); |
b8c5462f JH |
952 | return is_utf8_cntrl(tmpbuf); |
953 | } | |
954 | ||
955 | bool | |
84afefe6 | 956 | Perl_is_uni_graph(pTHX_ UV c) |
b8c5462f | 957 | { |
ad391ad9 | 958 | U8 tmpbuf[UTF8_MAXLEN+1]; |
230880c1 | 959 | uvchr_to_utf8(tmpbuf, c); |
b8c5462f JH |
960 | return is_utf8_graph(tmpbuf); |
961 | } | |
962 | ||
963 | bool | |
84afefe6 | 964 | Perl_is_uni_print(pTHX_ UV c) |
a0ed51b3 | 965 | { |
ad391ad9 | 966 | U8 tmpbuf[UTF8_MAXLEN+1]; |
230880c1 | 967 | uvchr_to_utf8(tmpbuf, c); |
a0ed51b3 LW |
968 | return is_utf8_print(tmpbuf); |
969 | } | |
970 | ||
b8c5462f | 971 | bool |
84afefe6 | 972 | Perl_is_uni_punct(pTHX_ UV c) |
b8c5462f | 973 | { |
ad391ad9 | 974 | U8 tmpbuf[UTF8_MAXLEN+1]; |
230880c1 | 975 | uvchr_to_utf8(tmpbuf, c); |
b8c5462f JH |
976 | return is_utf8_punct(tmpbuf); |
977 | } | |
978 | ||
4d61ec05 | 979 | bool |
84afefe6 | 980 | Perl_is_uni_xdigit(pTHX_ UV c) |
4d61ec05 | 981 | { |
e7ae6809 | 982 | U8 tmpbuf[UTF8_MAXLEN_UCLC+1]; |
230880c1 | 983 | uvchr_to_utf8(tmpbuf, c); |
4d61ec05 GS |
984 | return is_utf8_xdigit(tmpbuf); |
985 | } | |
986 | ||
84afefe6 JH |
987 | UV |
988 | Perl_to_uni_upper(pTHX_ UV c, U8* p, STRLEN *lenp) | |
a0ed51b3 | 989 | { |
0ebc6274 JH |
990 | uvchr_to_utf8(p, c); |
991 | return to_utf8_upper(p, p, lenp); | |
a0ed51b3 LW |
992 | } |
993 | ||
84afefe6 JH |
994 | UV |
995 | Perl_to_uni_title(pTHX_ UV c, U8* p, STRLEN *lenp) | |
a0ed51b3 | 996 | { |
0ebc6274 JH |
997 | uvchr_to_utf8(p, c); |
998 | return to_utf8_title(p, p, lenp); | |
a0ed51b3 LW |
999 | } |
1000 | ||
84afefe6 JH |
1001 | UV |
1002 | Perl_to_uni_lower(pTHX_ UV c, U8* p, STRLEN *lenp) | |
a0ed51b3 | 1003 | { |
0ebc6274 JH |
1004 | uvchr_to_utf8(p, c); |
1005 | return to_utf8_lower(p, p, lenp); | |
a0ed51b3 LW |
1006 | } |
1007 | ||
84afefe6 JH |
1008 | UV |
1009 | Perl_to_uni_fold(pTHX_ UV c, U8* p, STRLEN *lenp) | |
1010 | { | |
0ebc6274 JH |
1011 | uvchr_to_utf8(p, c); |
1012 | return to_utf8_fold(p, p, lenp); | |
84afefe6 JH |
1013 | } |
1014 | ||
a0ed51b3 LW |
1015 | /* for now these all assume no locale info available for Unicode > 255 */ |
1016 | ||
1017 | bool | |
84afefe6 | 1018 | Perl_is_uni_alnum_lc(pTHX_ UV c) |
a0ed51b3 LW |
1019 | { |
1020 | return is_uni_alnum(c); /* XXX no locale support yet */ | |
1021 | } | |
1022 | ||
1023 | bool | |
84afefe6 | 1024 | Perl_is_uni_alnumc_lc(pTHX_ UV c) |
b8c5462f JH |
1025 | { |
1026 | return is_uni_alnumc(c); /* XXX no locale support yet */ | |
1027 | } | |
1028 | ||
1029 | bool | |
84afefe6 | 1030 | Perl_is_uni_idfirst_lc(pTHX_ UV c) |
a0ed51b3 LW |
1031 | { |
1032 | return is_uni_idfirst(c); /* XXX no locale support yet */ | |
1033 | } | |
1034 | ||
1035 | bool | |
84afefe6 | 1036 | Perl_is_uni_alpha_lc(pTHX_ UV c) |
a0ed51b3 LW |
1037 | { |
1038 | return is_uni_alpha(c); /* XXX no locale support yet */ | |
1039 | } | |
1040 | ||
1041 | bool | |
84afefe6 | 1042 | Perl_is_uni_ascii_lc(pTHX_ UV c) |
4d61ec05 GS |
1043 | { |
1044 | return is_uni_ascii(c); /* XXX no locale support yet */ | |
1045 | } | |
1046 | ||
1047 | bool | |
84afefe6 | 1048 | Perl_is_uni_space_lc(pTHX_ UV c) |
a0ed51b3 LW |
1049 | { |
1050 | return is_uni_space(c); /* XXX no locale support yet */ | |
1051 | } | |
1052 | ||
1053 | bool | |
84afefe6 | 1054 | Perl_is_uni_digit_lc(pTHX_ UV c) |
a0ed51b3 LW |
1055 | { |
1056 | return is_uni_digit(c); /* XXX no locale support yet */ | |
1057 | } | |
1058 | ||
1059 | bool | |
84afefe6 | 1060 | Perl_is_uni_upper_lc(pTHX_ UV c) |
a0ed51b3 LW |
1061 | { |
1062 | return is_uni_upper(c); /* XXX no locale support yet */ | |
1063 | } | |
1064 | ||
1065 | bool | |
84afefe6 | 1066 | Perl_is_uni_lower_lc(pTHX_ UV c) |
a0ed51b3 LW |
1067 | { |
1068 | return is_uni_lower(c); /* XXX no locale support yet */ | |
1069 | } | |
1070 | ||
1071 | bool | |
84afefe6 | 1072 | Perl_is_uni_cntrl_lc(pTHX_ UV c) |
b8c5462f JH |
1073 | { |
1074 | return is_uni_cntrl(c); /* XXX no locale support yet */ | |
1075 | } | |
1076 | ||
1077 | bool | |
84afefe6 | 1078 | Perl_is_uni_graph_lc(pTHX_ UV c) |
b8c5462f JH |
1079 | { |
1080 | return is_uni_graph(c); /* XXX no locale support yet */ | |
1081 | } | |
1082 | ||
1083 | bool | |
84afefe6 | 1084 | Perl_is_uni_print_lc(pTHX_ UV c) |
a0ed51b3 LW |
1085 | { |
1086 | return is_uni_print(c); /* XXX no locale support yet */ | |
1087 | } | |
1088 | ||
b8c5462f | 1089 | bool |
84afefe6 | 1090 | Perl_is_uni_punct_lc(pTHX_ UV c) |
b8c5462f JH |
1091 | { |
1092 | return is_uni_punct(c); /* XXX no locale support yet */ | |
1093 | } | |
1094 | ||
4d61ec05 | 1095 | bool |
84afefe6 | 1096 | Perl_is_uni_xdigit_lc(pTHX_ UV c) |
4d61ec05 GS |
1097 | { |
1098 | return is_uni_xdigit(c); /* XXX no locale support yet */ | |
1099 | } | |
1100 | ||
b7ac61fa JH |
1101 | U32 |
1102 | Perl_to_uni_upper_lc(pTHX_ U32 c) | |
1103 | { | |
ee099d14 JH |
1104 | /* XXX returns only the first character -- do not use XXX */ |
1105 | /* XXX no locale support yet */ | |
1106 | STRLEN len; | |
1107 | U8 tmpbuf[UTF8_MAXLEN_UCLC+1]; | |
1108 | return (U32)to_uni_upper(c, tmpbuf, &len); | |
b7ac61fa JH |
1109 | } |
1110 | ||
1111 | U32 | |
1112 | Perl_to_uni_title_lc(pTHX_ U32 c) | |
1113 | { | |
ee099d14 JH |
1114 | /* XXX returns only the first character XXX -- do not use XXX */ |
1115 | /* XXX no locale support yet */ | |
1116 | STRLEN len; | |
1117 | U8 tmpbuf[UTF8_MAXLEN_UCLC+1]; | |
1118 | return (U32)to_uni_title(c, tmpbuf, &len); | |
b7ac61fa JH |
1119 | } |
1120 | ||
1121 | U32 | |
1122 | Perl_to_uni_lower_lc(pTHX_ U32 c) | |
1123 | { | |
ee099d14 JH |
1124 | /* XXX returns only the first character -- do not use XXX */ |
1125 | /* XXX no locale support yet */ | |
1126 | STRLEN len; | |
1127 | U8 tmpbuf[UTF8_MAXLEN_UCLC+1]; | |
1128 | return (U32)to_uni_lower(c, tmpbuf, &len); | |
b7ac61fa JH |
1129 | } |
1130 | ||
a0ed51b3 | 1131 | bool |
864dbfa3 | 1132 | Perl_is_utf8_alnum(pTHX_ U8 *p) |
a0ed51b3 | 1133 | { |
386d01d6 GS |
1134 | if (!is_utf8_char(p)) |
1135 | return FALSE; | |
a0ed51b3 | 1136 | if (!PL_utf8_alnum) |
289d4f09 ML |
1137 | /* NOTE: "IsWord", not "IsAlnum", since Alnum is a true |
1138 | * descendant of isalnum(3), in other words, it doesn't | |
1139 | * contain the '_'. --jhi */ | |
1140 | PL_utf8_alnum = swash_init("utf8", "IsWord", &PL_sv_undef, 0, 0); | |
eb160463 | 1141 | return swash_fetch(PL_utf8_alnum, p, TRUE) != 0; |
a0ed51b3 LW |
1142 | /* return *p == '_' || is_utf8_alpha(p) || is_utf8_digit(p); */ |
1143 | #ifdef SURPRISINGLY_SLOWER /* probably because alpha is usually true */ | |
1144 | if (!PL_utf8_alnum) | |
1145 | PL_utf8_alnum = swash_init("utf8", "", | |
1146 | sv_2mortal(newSVpv("+utf8::IsAlpha\n+utf8::IsDigit\n005F\n",0)), 0, 0); | |
eb160463 | 1147 | return swash_fetch(PL_utf8_alnum, p, TRUE) != 0; |
a0ed51b3 LW |
1148 | #endif |
1149 | } | |
1150 | ||
1151 | bool | |
b8c5462f JH |
1152 | Perl_is_utf8_alnumc(pTHX_ U8 *p) |
1153 | { | |
386d01d6 GS |
1154 | if (!is_utf8_char(p)) |
1155 | return FALSE; | |
b8c5462f JH |
1156 | if (!PL_utf8_alnum) |
1157 | PL_utf8_alnum = swash_init("utf8", "IsAlnumC", &PL_sv_undef, 0, 0); | |
eb160463 | 1158 | return swash_fetch(PL_utf8_alnum, p, TRUE) != 0; |
b8c5462f JH |
1159 | /* return is_utf8_alpha(p) || is_utf8_digit(p); */ |
1160 | #ifdef SURPRISINGLY_SLOWER /* probably because alpha is usually true */ | |
1161 | if (!PL_utf8_alnum) | |
1162 | PL_utf8_alnum = swash_init("utf8", "", | |
1163 | sv_2mortal(newSVpv("+utf8::IsAlpha\n+utf8::IsDigit\n005F\n",0)), 0, 0); | |
eb160463 | 1164 | return swash_fetch(PL_utf8_alnum, p, TRUE) != 0; |
b8c5462f JH |
1165 | #endif |
1166 | } | |
1167 | ||
1168 | bool | |
82686b01 | 1169 | Perl_is_utf8_idfirst(pTHX_ U8 *p) /* The naming is historical. */ |
a0ed51b3 | 1170 | { |
82686b01 JH |
1171 | if (*p == '_') |
1172 | return TRUE; | |
1173 | if (!is_utf8_char(p)) | |
1174 | return FALSE; | |
1175 | if (!PL_utf8_idstart) /* is_utf8_idstart would be more logical. */ | |
1176 | PL_utf8_idstart = swash_init("utf8", "IdStart", &PL_sv_undef, 0, 0); | |
eb160463 | 1177 | return swash_fetch(PL_utf8_idstart, p, TRUE) != 0; |
82686b01 JH |
1178 | } |
1179 | ||
1180 | bool | |
1181 | Perl_is_utf8_idcont(pTHX_ U8 *p) | |
1182 | { | |
1183 | if (*p == '_') | |
1184 | return TRUE; | |
1185 | if (!is_utf8_char(p)) | |
1186 | return FALSE; | |
1187 | if (!PL_utf8_idcont) | |
1188 | PL_utf8_idcont = swash_init("utf8", "IdContinue", &PL_sv_undef, 0, 0); | |
eb160463 | 1189 | return swash_fetch(PL_utf8_idcont, p, TRUE) != 0; |
a0ed51b3 LW |
1190 | } |
1191 | ||
1192 | bool | |
864dbfa3 | 1193 | Perl_is_utf8_alpha(pTHX_ U8 *p) |
a0ed51b3 | 1194 | { |
386d01d6 GS |
1195 | if (!is_utf8_char(p)) |
1196 | return FALSE; | |
a0ed51b3 | 1197 | if (!PL_utf8_alpha) |
e24b16f9 | 1198 | PL_utf8_alpha = swash_init("utf8", "IsAlpha", &PL_sv_undef, 0, 0); |
eb160463 | 1199 | return swash_fetch(PL_utf8_alpha, p, TRUE) != 0; |
a0ed51b3 LW |
1200 | } |
1201 | ||
1202 | bool | |
b8c5462f JH |
1203 | Perl_is_utf8_ascii(pTHX_ U8 *p) |
1204 | { | |
386d01d6 GS |
1205 | if (!is_utf8_char(p)) |
1206 | return FALSE; | |
b8c5462f JH |
1207 | if (!PL_utf8_ascii) |
1208 | PL_utf8_ascii = swash_init("utf8", "IsAscii", &PL_sv_undef, 0, 0); | |
eb160463 | 1209 | return swash_fetch(PL_utf8_ascii, p, TRUE) != 0; |
b8c5462f JH |
1210 | } |
1211 | ||
1212 | bool | |
864dbfa3 | 1213 | Perl_is_utf8_space(pTHX_ U8 *p) |
a0ed51b3 | 1214 | { |
386d01d6 GS |
1215 | if (!is_utf8_char(p)) |
1216 | return FALSE; | |
a0ed51b3 | 1217 | if (!PL_utf8_space) |
3bec3564 | 1218 | PL_utf8_space = swash_init("utf8", "IsSpacePerl", &PL_sv_undef, 0, 0); |
eb160463 | 1219 | return swash_fetch(PL_utf8_space, p, TRUE) != 0; |
a0ed51b3 LW |
1220 | } |
1221 | ||
1222 | bool | |
864dbfa3 | 1223 | Perl_is_utf8_digit(pTHX_ U8 *p) |
a0ed51b3 | 1224 | { |
386d01d6 GS |
1225 | if (!is_utf8_char(p)) |
1226 | return FALSE; | |
a0ed51b3 | 1227 | if (!PL_utf8_digit) |
e24b16f9 | 1228 | PL_utf8_digit = swash_init("utf8", "IsDigit", &PL_sv_undef, 0, 0); |
eb160463 | 1229 | return swash_fetch(PL_utf8_digit, p, TRUE) != 0; |
a0ed51b3 LW |
1230 | } |
1231 | ||
1232 | bool | |
864dbfa3 | 1233 | Perl_is_utf8_upper(pTHX_ U8 *p) |
a0ed51b3 | 1234 | { |
386d01d6 GS |
1235 | if (!is_utf8_char(p)) |
1236 | return FALSE; | |
a0ed51b3 | 1237 | if (!PL_utf8_upper) |
c65e4d19 | 1238 | PL_utf8_upper = swash_init("utf8", "IsUppercase", &PL_sv_undef, 0, 0); |
eb160463 | 1239 | return swash_fetch(PL_utf8_upper, p, TRUE) != 0; |
a0ed51b3 LW |
1240 | } |
1241 | ||
1242 | bool | |
864dbfa3 | 1243 | Perl_is_utf8_lower(pTHX_ U8 *p) |
a0ed51b3 | 1244 | { |
386d01d6 GS |
1245 | if (!is_utf8_char(p)) |
1246 | return FALSE; | |
a0ed51b3 | 1247 | if (!PL_utf8_lower) |
c65e4d19 | 1248 | PL_utf8_lower = swash_init("utf8", "IsLowercase", &PL_sv_undef, 0, 0); |
eb160463 | 1249 | return swash_fetch(PL_utf8_lower, p, TRUE) != 0; |
a0ed51b3 LW |
1250 | } |
1251 | ||
1252 | bool | |
b8c5462f JH |
1253 | Perl_is_utf8_cntrl(pTHX_ U8 *p) |
1254 | { | |
386d01d6 GS |
1255 | if (!is_utf8_char(p)) |
1256 | return FALSE; | |
b8c5462f JH |
1257 | if (!PL_utf8_cntrl) |
1258 | PL_utf8_cntrl = swash_init("utf8", "IsCntrl", &PL_sv_undef, 0, 0); | |
eb160463 | 1259 | return swash_fetch(PL_utf8_cntrl, p, TRUE) != 0; |
b8c5462f JH |
1260 | } |
1261 | ||
1262 | bool | |
1263 | Perl_is_utf8_graph(pTHX_ U8 *p) | |
1264 | { | |
386d01d6 GS |
1265 | if (!is_utf8_char(p)) |
1266 | return FALSE; | |
b8c5462f JH |
1267 | if (!PL_utf8_graph) |
1268 | PL_utf8_graph = swash_init("utf8", "IsGraph", &PL_sv_undef, 0, 0); | |
eb160463 | 1269 | return swash_fetch(PL_utf8_graph, p, TRUE) != 0; |
b8c5462f JH |
1270 | } |
1271 | ||
1272 | bool | |
864dbfa3 | 1273 | Perl_is_utf8_print(pTHX_ U8 *p) |
a0ed51b3 | 1274 | { |
386d01d6 GS |
1275 | if (!is_utf8_char(p)) |
1276 | return FALSE; | |
a0ed51b3 | 1277 | if (!PL_utf8_print) |
e24b16f9 | 1278 | PL_utf8_print = swash_init("utf8", "IsPrint", &PL_sv_undef, 0, 0); |
eb160463 | 1279 | return swash_fetch(PL_utf8_print, p, TRUE) != 0; |
a0ed51b3 LW |
1280 | } |
1281 | ||
1282 | bool | |
b8c5462f JH |
1283 | Perl_is_utf8_punct(pTHX_ U8 *p) |
1284 | { | |
386d01d6 GS |
1285 | if (!is_utf8_char(p)) |
1286 | return FALSE; | |
b8c5462f JH |
1287 | if (!PL_utf8_punct) |
1288 | PL_utf8_punct = swash_init("utf8", "IsPunct", &PL_sv_undef, 0, 0); | |
eb160463 | 1289 | return swash_fetch(PL_utf8_punct, p, TRUE) != 0; |
b8c5462f JH |
1290 | } |
1291 | ||
1292 | bool | |
1293 | Perl_is_utf8_xdigit(pTHX_ U8 *p) | |
1294 | { | |
386d01d6 GS |
1295 | if (!is_utf8_char(p)) |
1296 | return FALSE; | |
b8c5462f JH |
1297 | if (!PL_utf8_xdigit) |
1298 | PL_utf8_xdigit = swash_init("utf8", "IsXDigit", &PL_sv_undef, 0, 0); | |
eb160463 | 1299 | return swash_fetch(PL_utf8_xdigit, p, TRUE) != 0; |
b8c5462f JH |
1300 | } |
1301 | ||
1302 | bool | |
864dbfa3 | 1303 | Perl_is_utf8_mark(pTHX_ U8 *p) |
a0ed51b3 | 1304 | { |
386d01d6 GS |
1305 | if (!is_utf8_char(p)) |
1306 | return FALSE; | |
a0ed51b3 | 1307 | if (!PL_utf8_mark) |
e24b16f9 | 1308 | PL_utf8_mark = swash_init("utf8", "IsM", &PL_sv_undef, 0, 0); |
eb160463 | 1309 | return swash_fetch(PL_utf8_mark, p, TRUE) != 0; |
a0ed51b3 LW |
1310 | } |
1311 | ||
6b5c0936 JH |
1312 | /* |
1313 | =for apidoc A|UV|to_utf8_case|U8 *p|U8* ustrp|STRLEN *lenp|SV **swash|char *normal|char *special | |
1314 | ||
1315 | The "p" contains the pointer to the UTF-8 string encoding | |
1316 | the character that is being converted. | |
1317 | ||
1318 | The "ustrp" is a pointer to the character buffer to put the | |
1319 | conversion result to. The "lenp" is a pointer to the length | |
1320 | of the result. | |
1321 | ||
0134edef | 1322 | The "swashp" is a pointer to the swash to use. |
6b5c0936 | 1323 | |
0134edef JH |
1324 | Both the special and normal mappings are stored lib/unicore/To/Foo.pl, |
1325 | and loaded by SWASHGET, using lib/utf8_heavy.pl. The special (usually, | |
1326 | but not always, a multicharacter mapping), is tried first. | |
6b5c0936 | 1327 | |
0134edef JH |
1328 | The "special" is a string like "utf8::ToSpecLower", which means the |
1329 | hash %utf8::ToSpecLower. The access to the hash is through | |
1330 | Perl_to_utf8_case(). | |
6b5c0936 | 1331 | |
0134edef JH |
1332 | The "normal" is a string like "ToLower" which means the swash |
1333 | %utf8::ToLower. | |
1334 | ||
1335 | =cut */ | |
6b5c0936 | 1336 | |
2104c8d9 | 1337 | UV |
a6872d42 | 1338 | Perl_to_utf8_case(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp, SV **swashp, char *normal, char *special) |
a0ed51b3 | 1339 | { |
0134edef | 1340 | UV uv0, uv1; |
2f9475ad | 1341 | U8 tmpbuf[UTF8_MAXLEN_FOLD+1]; |
0134edef | 1342 | STRLEN len = 0; |
a0ed51b3 | 1343 | |
1feea2c7 JH |
1344 | uv0 = utf8_to_uvchr(p, 0); |
1345 | /* The NATIVE_TO_UNI() and UNI_TO_NATIVE() mappings | |
1346 | * are necessary in EBCDIC, they are redundant no-ops | |
1347 | * in ASCII-ish platforms, and hopefully optimized away. */ | |
1348 | uv1 = NATIVE_TO_UNI(uv0); | |
1349 | uvuni_to_utf8(tmpbuf, uv1); | |
0134edef JH |
1350 | |
1351 | if (!*swashp) /* load on-demand */ | |
1352 | *swashp = swash_init("utf8", normal, &PL_sv_undef, 4, 0); | |
1353 | ||
1354 | if (special) { | |
1355 | /* It might be "special" (sometimes, but not always, | |
2a37f04d | 1356 | * a multicharacter mapping) */ |
983ffd37 JH |
1357 | HV *hv; |
1358 | SV *keysv; | |
1359 | HE *he; | |
2a37f04d | 1360 | SV *val; |
2f9475ad | 1361 | |
983ffd37 | 1362 | if ((hv = get_hv(special, FALSE)) && |
1feea2c7 | 1363 | (keysv = sv_2mortal(Perl_newSVpvf(aTHX_ "%04"UVXf, uv1))) && |
2a37f04d JH |
1364 | (he = hv_fetch_ent(hv, keysv, FALSE, 0)) && |
1365 | (val = HeVAL(he))) { | |
47654450 | 1366 | char *s; |
47654450 | 1367 | |
2a37f04d | 1368 | s = SvPV(val, len); |
47654450 JH |
1369 | if (len == 1) |
1370 | len = uvuni_to_utf8(ustrp, NATIVE_TO_UNI(*(U8*)s)) - ustrp; | |
2a37f04d | 1371 | else { |
2f9475ad JH |
1372 | #ifdef EBCDIC |
1373 | /* If we have EBCDIC we need to remap the characters | |
1374 | * since any characters in the low 256 are Unicode | |
1375 | * code points, not EBCDIC. */ | |
7cda7a3d | 1376 | U8 *t = (U8*)s, *tend = t + len, *d; |
2f9475ad JH |
1377 | |
1378 | d = tmpbuf; | |
1379 | if (SvUTF8(val)) { | |
1380 | STRLEN tlen = 0; | |
1381 | ||
1382 | while (t < tend) { | |
1383 | UV c = utf8_to_uvchr(t, &tlen); | |
1384 | if (tlen > 0) { | |
1385 | d = uvchr_to_utf8(d, UNI_TO_NATIVE(c)); | |
1386 | t += tlen; | |
1387 | } | |
1388 | else | |
1389 | break; | |
1390 | } | |
1391 | } | |
1392 | else { | |
36fec512 JH |
1393 | while (t < tend) { |
1394 | d = uvchr_to_utf8(d, UNI_TO_NATIVE(*t)); | |
1395 | t++; | |
1396 | } | |
2f9475ad JH |
1397 | } |
1398 | len = d - tmpbuf; | |
1399 | Copy(tmpbuf, ustrp, len, U8); | |
1400 | #else | |
d2dcd0fb | 1401 | Copy(s, ustrp, len, U8); |
2f9475ad | 1402 | #endif |
29e98929 | 1403 | } |
983ffd37 | 1404 | } |
0134edef JH |
1405 | } |
1406 | ||
1407 | if (!len && *swashp) { | |
1408 | UV uv2 = swash_fetch(*swashp, tmpbuf, TRUE); | |
1409 | ||
1410 | if (uv2) { | |
1411 | /* It was "normal" (a single character mapping). */ | |
1412 | UV uv3 = UNI_TO_NATIVE(uv2); | |
1413 | ||
e9101d72 | 1414 | len = uvchr_to_utf8(ustrp, uv3) - ustrp; |
2a37f04d JH |
1415 | } |
1416 | } | |
1feea2c7 | 1417 | |
0134edef JH |
1418 | if (!len) /* Neither: just copy. */ |
1419 | len = uvchr_to_utf8(ustrp, uv0) - ustrp; | |
1420 | ||
2a37f04d JH |
1421 | if (lenp) |
1422 | *lenp = len; | |
1423 | ||
0134edef | 1424 | return len ? utf8_to_uvchr(ustrp, 0) : 0; |
a0ed51b3 LW |
1425 | } |
1426 | ||
d3e79532 JH |
1427 | /* |
1428 | =for apidoc A|UV|to_utf8_upper|U8 *p|U8 *ustrp|STRLEN *lenp | |
1429 | ||
1430 | Convert the UTF-8 encoded character at p to its uppercase version and | |
1431 | store that in UTF-8 in ustrp and its length in bytes in lenp. Note | |
1432 | that the ustrp needs to be at least UTF8_MAXLEN_UCLC+1 bytes since the | |
1433 | uppercase version may be longer than the original character (up to two | |
1434 | characters). | |
1435 | ||
1436 | The first character of the uppercased version is returned | |
1437 | (but note, as explained above, that there may be more.) | |
1438 | ||
1439 | =cut */ | |
1440 | ||
2104c8d9 | 1441 | UV |
983ffd37 | 1442 | Perl_to_utf8_upper(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp) |
a0ed51b3 | 1443 | { |
983ffd37 | 1444 | return Perl_to_utf8_case(aTHX_ p, ustrp, lenp, |
b4e400f9 | 1445 | &PL_utf8_toupper, "ToUpper", "utf8::ToSpecUpper"); |
983ffd37 | 1446 | } |
a0ed51b3 | 1447 | |
d3e79532 JH |
1448 | /* |
1449 | =for apidoc A|UV|to_utf8_title|U8 *p|U8 *ustrp|STRLEN *lenp | |
1450 | ||
1451 | Convert the UTF-8 encoded character at p to its titlecase version and | |
1452 | store that in UTF-8 in ustrp and its length in bytes in lenp. Note | |
1453 | that the ustrp needs to be at least UTF8_MAXLEN_UCLC+1 bytes since the | |
1454 | titlecase version may be longer than the original character (up to two | |
1455 | characters). | |
1456 | ||
1457 | The first character of the titlecased version is returned | |
1458 | (but note, as explained above, that there may be more.) | |
1459 | ||
1460 | =cut */ | |
1461 | ||
983ffd37 JH |
1462 | UV |
1463 | Perl_to_utf8_title(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp) | |
1464 | { | |
1465 | return Perl_to_utf8_case(aTHX_ p, ustrp, lenp, | |
b4e400f9 | 1466 | &PL_utf8_totitle, "ToTitle", "utf8::ToSpecTitle"); |
a0ed51b3 LW |
1467 | } |
1468 | ||
d3e79532 JH |
1469 | /* |
1470 | =for apidoc A|UV|to_utf8_lower|U8 *p|U8 *ustrp|STRLEN *lenp | |
1471 | ||
1472 | Convert the UTF-8 encoded character at p to its lowercase version and | |
1473 | store that in UTF-8 in ustrp and its length in bytes in lenp. Note | |
1474 | that the ustrp needs to be at least UTF8_MAXLEN_UCLC+1 bytes since the | |
1475 | lowercase version may be longer than the original character (up to two | |
1476 | characters). | |
1477 | ||
1478 | The first character of the lowercased version is returned | |
1479 | (but note, as explained above, that there may be more.) | |
1480 | ||
1481 | =cut */ | |
1482 | ||
2104c8d9 | 1483 | UV |
a2a2844f | 1484 | Perl_to_utf8_lower(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp) |
a0ed51b3 | 1485 | { |
983ffd37 | 1486 | return Perl_to_utf8_case(aTHX_ p, ustrp, lenp, |
b4e400f9 JH |
1487 | &PL_utf8_tolower, "ToLower", "utf8::ToSpecLower"); |
1488 | } | |
1489 | ||
d3e79532 JH |
1490 | /* |
1491 | =for apidoc A|UV|to_utf8_fold|U8 *p|U8 *ustrp|STRLEN *lenp | |
1492 | ||
1493 | Convert the UTF-8 encoded character at p to its foldcase version and | |
1494 | store that in UTF-8 in ustrp and its length in bytes in lenp. Note | |
1495 | that the ustrp needs to be at least UTF8_MAXLEN_FOLD+1 bytes since the | |
1496 | foldcase version may be longer than the original character (up to | |
1497 | three characters). | |
1498 | ||
1499 | The first character of the foldcased version is returned | |
1500 | (but note, as explained above, that there may be more.) | |
1501 | ||
1502 | =cut */ | |
1503 | ||
b4e400f9 JH |
1504 | UV |
1505 | Perl_to_utf8_fold(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp) | |
1506 | { | |
1507 | return Perl_to_utf8_case(aTHX_ p, ustrp, lenp, | |
1508 | &PL_utf8_tofold, "ToFold", "utf8::ToSpecFold"); | |
a0ed51b3 LW |
1509 | } |
1510 | ||
1511 | /* a "swash" is a swatch hash */ | |
1512 | ||
1513 | SV* | |
864dbfa3 | 1514 | Perl_swash_init(pTHX_ char* pkg, char* name, SV *listsv, I32 minbits, I32 none) |
a0ed51b3 LW |
1515 | { |
1516 | SV* retval; | |
bf1fed83 | 1517 | SV* tokenbufsv = sv_2mortal(NEWSV(0,0)); |
8e84507e | 1518 | dSP; |
1b026014 | 1519 | HV *stash = gv_stashpvn(pkg, strlen(pkg), FALSE); |
f8be5cf0 | 1520 | SV* errsv_save; |
ce3b816e | 1521 | |
1b026014 | 1522 | if (!gv_fetchmeth(stash, "SWASHNEW", 8, -1)) { /* demand load utf8 */ |
ce3b816e | 1523 | ENTER; |
f8be5cf0 | 1524 | errsv_save = newSVsv(ERRSV); |
ce3b816e | 1525 | Perl_load_module(aTHX_ PERL_LOADMOD_NOIMPORT, newSVpv(pkg,0), Nullsv); |
f8be5cf0 JH |
1526 | if (!SvTRUE(ERRSV)) |
1527 | sv_setsv(ERRSV, errsv_save); | |
1528 | SvREFCNT_dec(errsv_save); | |
ce3b816e GS |
1529 | LEAVE; |
1530 | } | |
1531 | SPAGAIN; | |
a0ed51b3 LW |
1532 | PUSHSTACKi(PERLSI_MAGIC); |
1533 | PUSHMARK(SP); | |
1534 | EXTEND(SP,5); | |
1535 | PUSHs(sv_2mortal(newSVpvn(pkg, strlen(pkg)))); | |
1536 | PUSHs(sv_2mortal(newSVpvn(name, strlen(name)))); | |
1537 | PUSHs(listsv); | |
1538 | PUSHs(sv_2mortal(newSViv(minbits))); | |
1539 | PUSHs(sv_2mortal(newSViv(none))); | |
1540 | PUTBACK; | |
1541 | ENTER; | |
1542 | SAVEI32(PL_hints); | |
1543 | PL_hints = 0; | |
1544 | save_re_context(); | |
82686b01 | 1545 | if (PL_curcop == &PL_compiling) { |
bf1fed83 | 1546 | /* XXX ought to be handled by lex_start */ |
82686b01 | 1547 | SAVEI32(PL_in_my); |
2b4bd638 | 1548 | PL_in_my = 0; |
bf1fed83 | 1549 | sv_setpv(tokenbufsv, PL_tokenbuf); |
82686b01 | 1550 | } |
f8be5cf0 | 1551 | errsv_save = newSVsv(ERRSV); |
864dbfa3 | 1552 | if (call_method("SWASHNEW", G_SCALAR)) |
8e84507e | 1553 | retval = newSVsv(*PL_stack_sp--); |
a0ed51b3 | 1554 | else |
e24b16f9 | 1555 | retval = &PL_sv_undef; |
f8be5cf0 JH |
1556 | if (!SvTRUE(ERRSV)) |
1557 | sv_setsv(ERRSV, errsv_save); | |
1558 | SvREFCNT_dec(errsv_save); | |
a0ed51b3 LW |
1559 | LEAVE; |
1560 | POPSTACK; | |
e24b16f9 | 1561 | if (PL_curcop == &PL_compiling) { |
bf1fed83 JH |
1562 | STRLEN len; |
1563 | char* pv = SvPV(tokenbufsv, len); | |
1564 | ||
1565 | Copy(pv, PL_tokenbuf, len+1, char); | |
eb160463 | 1566 | PL_curcop->op_private = (U8)(PL_hints & HINT_PRIVATE_MASK); |
a0ed51b3 | 1567 | } |
bc45ce41 JH |
1568 | if (!SvROK(retval) || SvTYPE(SvRV(retval)) != SVt_PVHV) { |
1569 | if (SvPOK(retval)) | |
35c1215d NC |
1570 | Perl_croak(aTHX_ "Can't find Unicode property definition \"%"SVf"\"", |
1571 | retval); | |
cea2e8a9 | 1572 | Perl_croak(aTHX_ "SWASHNEW didn't return an HV ref"); |
bc45ce41 | 1573 | } |
a0ed51b3 LW |
1574 | return retval; |
1575 | } | |
1576 | ||
035d37be JH |
1577 | |
1578 | /* This API is wrong for special case conversions since we may need to | |
1579 | * return several Unicode characters for a single Unicode character | |
1580 | * (see lib/unicore/SpecCase.txt) The SWASHGET in lib/utf8_heavy.pl is | |
1581 | * the lower-level routine, and it is similarly broken for returning | |
1582 | * multiple values. --jhi */ | |
a0ed51b3 | 1583 | UV |
3568d838 | 1584 | Perl_swash_fetch(pTHX_ SV *sv, U8 *ptr, bool do_utf8) |
a0ed51b3 LW |
1585 | { |
1586 | HV* hv = (HV*)SvRV(sv); | |
3568d838 JH |
1587 | U32 klen; |
1588 | U32 off; | |
a0ed51b3 | 1589 | STRLEN slen; |
7d85a32c | 1590 | STRLEN needents; |
4ea42e7f | 1591 | U8 *tmps = NULL; |
a0ed51b3 LW |
1592 | U32 bit; |
1593 | SV *retval; | |
3568d838 JH |
1594 | U8 tmputf8[2]; |
1595 | UV c = NATIVE_TO_ASCII(*ptr); | |
1596 | ||
1597 | if (!do_utf8 && !UNI_IS_INVARIANT(c)) { | |
eb160463 GS |
1598 | tmputf8[0] = (U8)UTF8_EIGHT_BIT_HI(c); |
1599 | tmputf8[1] = (U8)UTF8_EIGHT_BIT_LO(c); | |
3568d838 JH |
1600 | ptr = tmputf8; |
1601 | } | |
1602 | /* Given a UTF-X encoded char 0xAA..0xYY,0xZZ | |
1603 | * then the "swatch" is a vec() for al the chars which start | |
1604 | * with 0xAA..0xYY | |
1605 | * So the key in the hash (klen) is length of encoded char -1 | |
1606 | */ | |
1607 | klen = UTF8SKIP(ptr) - 1; | |
1608 | off = ptr[klen]; | |
a0ed51b3 | 1609 | |
7d85a32c JH |
1610 | if (klen == 0) |
1611 | { | |
1612 | /* If char in invariant then swatch is for all the invariant chars | |
1613 | * In both UTF-8 and UTF8-MOD that happens to be UTF_CONTINUATION_MARK | |
1614 | */ | |
1615 | needents = UTF_CONTINUATION_MARK; | |
1616 | off = NATIVE_TO_UTF(ptr[klen]); | |
1617 | } | |
1618 | else | |
1619 | { | |
1620 | /* If char is encoded then swatch is for the prefix */ | |
1621 | needents = (1 << UTF_ACCUMULATION_SHIFT); | |
1622 | off = NATIVE_TO_UTF(ptr[klen]) & UTF_CONTINUATION_MASK; | |
1623 | } | |
1624 | ||
a0ed51b3 LW |
1625 | /* |
1626 | * This single-entry cache saves about 1/3 of the utf8 overhead in test | |
1627 | * suite. (That is, only 7-8% overall over just a hash cache. Still, | |
1628 | * it's nothing to sniff at.) Pity we usually come through at least | |
1629 | * two function calls to get here... | |
1630 | * | |
1631 | * NB: this code assumes that swatches are never modified, once generated! | |
1632 | */ | |
1633 | ||
3568d838 | 1634 | if (hv == PL_last_swash_hv && |
a0ed51b3 | 1635 | klen == PL_last_swash_klen && |
3568d838 | 1636 | (!klen || memEQ((char *)ptr, (char *)PL_last_swash_key, klen)) ) |
a0ed51b3 LW |
1637 | { |
1638 | tmps = PL_last_swash_tmps; | |
1639 | slen = PL_last_swash_slen; | |
1640 | } | |
1641 | else { | |
1642 | /* Try our second-level swatch cache, kept in a hash. */ | |
dfe13c55 | 1643 | SV** svp = hv_fetch(hv, (char*)ptr, klen, FALSE); |
a0ed51b3 LW |
1644 | |
1645 | /* If not cached, generate it via utf8::SWASHGET */ | |
dfe13c55 | 1646 | if (!svp || !SvPOK(*svp) || !(tmps = (U8*)SvPV(*svp, slen))) { |
a0ed51b3 | 1647 | dSP; |
2b9d42f0 NIS |
1648 | /* We use utf8n_to_uvuni() as we want an index into |
1649 | Unicode tables, not a native character number. | |
1650 | */ | |
872c91ae JH |
1651 | UV code_point = utf8n_to_uvuni(ptr, UTF8_MAXLEN, 0, |
1652 | ckWARN(WARN_UTF8) ? | |
1653 | 0 : UTF8_ALLOW_ANY); | |
f8be5cf0 | 1654 | SV *errsv_save; |
a0ed51b3 LW |
1655 | ENTER; |
1656 | SAVETMPS; | |
1657 | save_re_context(); | |
1658 | PUSHSTACKi(PERLSI_MAGIC); | |
1659 | PUSHMARK(SP); | |
1660 | EXTEND(SP,3); | |
1661 | PUSHs((SV*)sv); | |
ffbc6a93 | 1662 | /* On EBCDIC & ~(0xA0-1) isn't a useful thing to do */ |
3568d838 JH |
1663 | PUSHs(sv_2mortal(newSViv((klen) ? |
1664 | (code_point & ~(needents - 1)) : 0))); | |
a0ed51b3 LW |
1665 | PUSHs(sv_2mortal(newSViv(needents))); |
1666 | PUTBACK; | |
f8be5cf0 | 1667 | errsv_save = newSVsv(ERRSV); |
864dbfa3 | 1668 | if (call_method("SWASHGET", G_SCALAR)) |
8e84507e | 1669 | retval = newSVsv(*PL_stack_sp--); |
a0ed51b3 | 1670 | else |
e24b16f9 | 1671 | retval = &PL_sv_undef; |
f8be5cf0 JH |
1672 | if (!SvTRUE(ERRSV)) |
1673 | sv_setsv(ERRSV, errsv_save); | |
1674 | SvREFCNT_dec(errsv_save); | |
a0ed51b3 LW |
1675 | POPSTACK; |
1676 | FREETMPS; | |
1677 | LEAVE; | |
e24b16f9 | 1678 | if (PL_curcop == &PL_compiling) |
eb160463 | 1679 | PL_curcop->op_private = (U8)(PL_hints & HINT_PRIVATE_MASK); |
a0ed51b3 | 1680 | |
dfe13c55 | 1681 | svp = hv_store(hv, (char*)ptr, klen, retval, 0); |
a0ed51b3 | 1682 | |
7d85a32c | 1683 | if (!svp || !(tmps = (U8*)SvPV(*svp, slen)) || (slen << 3) < needents) |
cea2e8a9 | 1684 | Perl_croak(aTHX_ "SWASHGET didn't return result of proper length"); |
a0ed51b3 LW |
1685 | } |
1686 | ||
1687 | PL_last_swash_hv = hv; | |
1688 | PL_last_swash_klen = klen; | |
1689 | PL_last_swash_tmps = tmps; | |
1690 | PL_last_swash_slen = slen; | |
1691 | if (klen) | |
1692 | Copy(ptr, PL_last_swash_key, klen, U8); | |
1693 | } | |
1694 | ||
9faf8d75 | 1695 | switch ((int)((slen << 3) / needents)) { |
a0ed51b3 LW |
1696 | case 1: |
1697 | bit = 1 << (off & 7); | |
1698 | off >>= 3; | |
1699 | return (tmps[off] & bit) != 0; | |
1700 | case 8: | |
1701 | return tmps[off]; | |
1702 | case 16: | |
1703 | off <<= 1; | |
1704 | return (tmps[off] << 8) + tmps[off + 1] ; | |
1705 | case 32: | |
1706 | off <<= 2; | |
1707 | return (tmps[off] << 24) + (tmps[off+1] << 16) + (tmps[off+2] << 8) + tmps[off + 3] ; | |
1708 | } | |
cea2e8a9 | 1709 | Perl_croak(aTHX_ "panic: swash_fetch"); |
a0ed51b3 LW |
1710 | return 0; |
1711 | } | |
2b9d42f0 NIS |
1712 | |
1713 | ||
1714 | /* | |
37607a96 | 1715 | =for apidoc A|U8 *|uvchr_to_utf8|U8 *d|UV uv |
2b9d42f0 NIS |
1716 | |
1717 | Adds the UTF8 representation of the Native codepoint C<uv> to the end | |
1718 | of the string C<d>; C<d> should be have at least C<UTF8_MAXLEN+1> free | |
1719 | bytes available. The return value is the pointer to the byte after the | |
1720 | end of the new character. In other words, | |
1721 | ||
1722 | d = uvchr_to_utf8(d, uv); | |
1723 | ||
1724 | is the recommended wide native character-aware way of saying | |
1725 | ||
1726 | *(d++) = uv; | |
1727 | ||
1728 | =cut | |
1729 | */ | |
1730 | ||
1731 | /* On ASCII machines this is normally a macro but we want a | |
1732 | real function in case XS code wants it | |
1733 | */ | |
1734 | #undef Perl_uvchr_to_utf8 | |
1735 | U8 * | |
1736 | Perl_uvchr_to_utf8(pTHX_ U8 *d, UV uv) | |
1737 | { | |
b851fbc1 | 1738 | return Perl_uvuni_to_utf8_flags(aTHX_ d, NATIVE_TO_UNI(uv), 0); |
2b9d42f0 NIS |
1739 | } |
1740 | ||
b851fbc1 JH |
1741 | U8 * |
1742 | Perl_uvchr_to_utf8_flags(pTHX_ U8 *d, UV uv, UV flags) | |
1743 | { | |
1744 | return Perl_uvuni_to_utf8_flags(aTHX_ d, NATIVE_TO_UNI(uv), flags); | |
1745 | } | |
2b9d42f0 NIS |
1746 | |
1747 | /* | |
37607a96 | 1748 | =for apidoc A|UV|utf8n_to_uvchr|U8 *s|STRLEN curlen|STRLEN *retlen|U32 flags |
2b9d42f0 NIS |
1749 | |
1750 | Returns the native character value of the first character in the string C<s> | |
1751 | which is assumed to be in UTF8 encoding; C<retlen> will be set to the | |
1752 | length, in bytes, of that character. | |
1753 | ||
1754 | Allows length and flags to be passed to low level routine. | |
1755 | ||
1756 | =cut | |
1757 | */ | |
0a2ef054 JH |
1758 | /* On ASCII machines this is normally a macro but we want |
1759 | a real function in case XS code wants it | |
2b9d42f0 NIS |
1760 | */ |
1761 | #undef Perl_utf8n_to_uvchr | |
1762 | UV | |
37607a96 | 1763 | Perl_utf8n_to_uvchr(pTHX_ U8 *s, STRLEN curlen, STRLEN *retlen, U32 flags) |
2b9d42f0 NIS |
1764 | { |
1765 | UV uv = Perl_utf8n_to_uvuni(aTHX_ s, curlen, retlen, flags); | |
1766 | return UNI_TO_NATIVE(uv); | |
1767 | } | |
1768 | ||
d2cc3551 JH |
1769 | /* |
1770 | =for apidoc A|char *|pv_uni_display|SV *dsv|U8 *spv|STRLEN len|STRLEN pvlim|UV flags | |
1771 | ||
1772 | Build to the scalar dsv a displayable version of the string spv, | |
1773 | length len, the displayable version being at most pvlim bytes long | |
1774 | (if longer, the rest is truncated and "..." will be appended). | |
0a2ef054 | 1775 | |
9e55ce06 | 1776 | The flags argument can have UNI_DISPLAY_ISPRINT set to display |
00e86452 | 1777 | isPRINT()able characters as themselves, UNI_DISPLAY_BACKSLASH |
0a2ef054 JH |
1778 | to display the \\[nrfta\\] as the backslashed versions (like '\n') |
1779 | (UNI_DISPLAY_BACKSLASH is preferred over UNI_DISPLAY_ISPRINT for \\). | |
1780 | UNI_DISPLAY_QQ (and its alias UNI_DISPLAY_REGEX) have both | |
1781 | UNI_DISPLAY_BACKSLASH and UNI_DISPLAY_ISPRINT turned on. | |
1782 | ||
d2cc3551 JH |
1783 | The pointer to the PV of the dsv is returned. |
1784 | ||
1785 | =cut */ | |
e6b2e755 JH |
1786 | char * |
1787 | Perl_pv_uni_display(pTHX_ SV *dsv, U8 *spv, STRLEN len, STRLEN pvlim, UV flags) | |
1788 | { | |
1789 | int truncated = 0; | |
1790 | char *s, *e; | |
1791 | ||
1792 | sv_setpvn(dsv, "", 0); | |
1793 | for (s = (char *)spv, e = s + len; s < e; s += UTF8SKIP(s)) { | |
1794 | UV u; | |
c728cb41 JH |
1795 | bool ok = FALSE; |
1796 | ||
e6b2e755 JH |
1797 | if (pvlim && SvCUR(dsv) >= pvlim) { |
1798 | truncated++; | |
1799 | break; | |
1800 | } | |
1801 | u = utf8_to_uvchr((U8*)s, 0); | |
c728cb41 | 1802 | if (u < 256) { |
c728cb41 JH |
1803 | if (!ok && (flags & UNI_DISPLAY_BACKSLASH)) { |
1804 | switch (u & 0xFF) { | |
1805 | case '\n': | |
1806 | Perl_sv_catpvf(aTHX_ dsv, "\\n"); ok = TRUE; break; | |
1807 | case '\r': | |
1808 | Perl_sv_catpvf(aTHX_ dsv, "\\r"); ok = TRUE; break; | |
1809 | case '\t': | |
1810 | Perl_sv_catpvf(aTHX_ dsv, "\\t"); ok = TRUE; break; | |
1811 | case '\f': | |
1812 | Perl_sv_catpvf(aTHX_ dsv, "\\f"); ok = TRUE; break; | |
1813 | case '\a': | |
1814 | Perl_sv_catpvf(aTHX_ dsv, "\\a"); ok = TRUE; break; | |
1815 | case '\\': | |
d79a7a3d | 1816 | Perl_sv_catpvf(aTHX_ dsv, "\\\\" ); ok = TRUE; break; |
c728cb41 JH |
1817 | default: break; |
1818 | } | |
1819 | } | |
00e86452 JH |
1820 | /* isPRINT() is the locale-blind version. */ |
1821 | if (!ok && (flags & UNI_DISPLAY_ISPRINT) && isPRINT(u & 0xFF)) { | |
2c4547fe | 1822 | Perl_sv_catpvf(aTHX_ dsv, "%c", (char)(u & 0xFF)); |
0a2ef054 JH |
1823 | ok = TRUE; |
1824 | } | |
c728cb41 JH |
1825 | } |
1826 | if (!ok) | |
9e55ce06 | 1827 | Perl_sv_catpvf(aTHX_ dsv, "\\x{%"UVxf"}", u); |
e6b2e755 JH |
1828 | } |
1829 | if (truncated) | |
1830 | sv_catpvn(dsv, "...", 3); | |
1831 | ||
1832 | return SvPVX(dsv); | |
1833 | } | |
2b9d42f0 | 1834 | |
d2cc3551 JH |
1835 | /* |
1836 | =for apidoc A|char *|sv_uni_display|SV *dsv|SV *ssv|STRLEN pvlim|UV flags | |
1837 | ||
1838 | Build to the scalar dsv a displayable version of the scalar sv, | |
0a2ef054 | 1839 | the displayable version being at most pvlim bytes long |
d2cc3551 | 1840 | (if longer, the rest is truncated and "..." will be appended). |
0a2ef054 JH |
1841 | |
1842 | The flags argument is as in pv_uni_display(). | |
1843 | ||
d2cc3551 JH |
1844 | The pointer to the PV of the dsv is returned. |
1845 | ||
1846 | =cut */ | |
e6b2e755 JH |
1847 | char * |
1848 | Perl_sv_uni_display(pTHX_ SV *dsv, SV *ssv, STRLEN pvlim, UV flags) | |
1849 | { | |
701a277b JH |
1850 | return Perl_pv_uni_display(aTHX_ dsv, (U8*)SvPVX(ssv), SvCUR(ssv), |
1851 | pvlim, flags); | |
1852 | } | |
1853 | ||
d2cc3551 | 1854 | /* |
d07ddd77 | 1855 | =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 |
1856 | |
1857 | Return true if the strings s1 and s2 differ case-insensitively, false | |
1858 | if not (if they are equal case-insensitively). If u1 is true, the | |
1859 | string s1 is assumed to be in UTF-8-encoded Unicode. If u2 is true, | |
d07ddd77 JH |
1860 | the string s2 is assumed to be in UTF-8-encoded Unicode. If u1 or u2 |
1861 | are false, the respective string is assumed to be in native 8-bit | |
1862 | encoding. | |
1863 | ||
1864 | If the pe1 and pe2 are non-NULL, the scanning pointers will be copied | |
1865 | in there (they will point at the beginning of the I<next> character). | |
1866 | If the pointers behind pe1 or pe2 are non-NULL, they are the end | |
1867 | pointers beyond which scanning will not continue under any | |
1868 | circustances. If the byte lengths l1 and l2 are non-zero, s1+l1 and | |
1869 | s2+l2 will be used as goal end pointers that will also stop the scan, | |
1870 | and which qualify towards defining a successful match: all the scans | |
1871 | that define an explicit length must reach their goal pointers for | |
1872 | a match to succeed). | |
d2cc3551 JH |
1873 | |
1874 | For case-insensitiveness, the "casefolding" of Unicode is used | |
1875 | instead of upper/lowercasing both the characters, see | |
1876 | http://www.unicode.org/unicode/reports/tr21/ (Case Mappings). | |
1877 | ||
1878 | =cut */ | |
701a277b | 1879 | I32 |
d07ddd77 | 1880 | 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 | 1881 | { |
5469e704 JH |
1882 | register U8 *p1 = (U8*)s1; |
1883 | register U8 *p2 = (U8*)s2; | |
d07ddd77 JH |
1884 | register U8 *e1 = 0, *f1 = 0, *q1 = 0; |
1885 | register U8 *e2 = 0, *f2 = 0, *q2 = 0; | |
1886 | STRLEN n1 = 0, n2 = 0; | |
ffce6cc2 JH |
1887 | U8 foldbuf1[UTF8_MAXLEN_FOLD+1]; |
1888 | U8 foldbuf2[UTF8_MAXLEN_FOLD+1]; | |
d7f013c8 JH |
1889 | U8 natbuf[1+1]; |
1890 | STRLEN foldlen1, foldlen2; | |
d07ddd77 | 1891 | bool match; |
332ddc25 | 1892 | |
d07ddd77 JH |
1893 | if (pe1) |
1894 | e1 = *(U8**)pe1; | |
eb160463 | 1895 | if (e1 == 0 || (l1 && l1 < (UV)(e1 - (U8*)s1))) |
d07ddd77 JH |
1896 | f1 = (U8*)s1 + l1; |
1897 | if (pe2) | |
1898 | e2 = *(U8**)pe2; | |
eb160463 | 1899 | if (e2 == 0 || (l2 && l2 < (UV)(e2 - (U8*)s2))) |
d07ddd77 JH |
1900 | f2 = (U8*)s2 + l2; |
1901 | ||
1902 | if ((e1 == 0 && f1 == 0) || (e2 == 0 && f2 == 0) || (f1 == 0 && f2 == 0)) | |
1903 | return 1; /* mismatch; possible infinite loop or false positive */ | |
1904 | ||
a6872d42 JH |
1905 | if (!u1 || !u2) |
1906 | natbuf[1] = 0; /* Need to terminate the buffer. */ | |
1907 | ||
d07ddd77 JH |
1908 | while ((e1 == 0 || p1 < e1) && |
1909 | (f1 == 0 || p1 < f1) && | |
1910 | (e2 == 0 || p2 < e2) && | |
1911 | (f2 == 0 || p2 < f2)) { | |
1912 | if (n1 == 0) { | |
d7f013c8 JH |
1913 | if (u1) |
1914 | to_utf8_fold(p1, foldbuf1, &foldlen1); | |
1915 | else { | |
f5cee151 | 1916 | natbuf[0] = *p1; |
d7f013c8 JH |
1917 | to_utf8_fold(natbuf, foldbuf1, &foldlen1); |
1918 | } | |
1919 | q1 = foldbuf1; | |
d07ddd77 | 1920 | n1 = foldlen1; |
332ddc25 | 1921 | } |
d07ddd77 | 1922 | if (n2 == 0) { |
d7f013c8 JH |
1923 | if (u2) |
1924 | to_utf8_fold(p2, foldbuf2, &foldlen2); | |
1925 | else { | |
f5cee151 | 1926 | natbuf[0] = *p2; |
d7f013c8 JH |
1927 | to_utf8_fold(natbuf, foldbuf2, &foldlen2); |
1928 | } | |
1929 | q2 = foldbuf2; | |
d07ddd77 | 1930 | n2 = foldlen2; |
332ddc25 | 1931 | } |
d07ddd77 JH |
1932 | while (n1 && n2) { |
1933 | if ( UTF8SKIP(q1) != UTF8SKIP(q2) || | |
1934 | (UTF8SKIP(q1) == 1 && *q1 != *q2) || | |
1935 | memNE((char*)q1, (char*)q2, UTF8SKIP(q1)) ) | |
d7f013c8 | 1936 | return 1; /* mismatch */ |
d07ddd77 | 1937 | n1 -= UTF8SKIP(q1); |
d7f013c8 | 1938 | q1 += UTF8SKIP(q1); |
d07ddd77 | 1939 | n2 -= UTF8SKIP(q2); |
d7f013c8 | 1940 | q2 += UTF8SKIP(q2); |
701a277b | 1941 | } |
d07ddd77 | 1942 | if (n1 == 0) |
d7f013c8 | 1943 | p1 += u1 ? UTF8SKIP(p1) : 1; |
d07ddd77 | 1944 | if (n2 == 0) |
d7f013c8 JH |
1945 | p2 += u2 ? UTF8SKIP(p2) : 1; |
1946 | ||
d2cc3551 | 1947 | } |
5469e704 | 1948 | |
d07ddd77 JH |
1949 | /* A match is defined by all the scans that specified |
1950 | * an explicit length reaching their final goals. */ | |
1951 | match = (f1 == 0 || p1 == f1) && (f2 == 0 || p2 == f2); | |
5469e704 JH |
1952 | |
1953 | if (match) { | |
d07ddd77 JH |
1954 | if (pe1) |
1955 | *pe1 = (char*)p1; | |
1956 | if (pe2) | |
1957 | *pe2 = (char*)p2; | |
5469e704 JH |
1958 | } |
1959 | ||
1960 | return match ? 0 : 1; /* 0 match, 1 mismatch */ | |
e6b2e755 | 1961 | } |
701a277b | 1962 |