Commit | Line | Data |
a0ed51b3 |
1 | /* utf8.c |
2 | * |
af3babe4 |
3 | * Copyright (C) 2000, 2001, 2002, 2003, 2004, 2005 by Larry Wall and |
4 | * others |
a0ed51b3 |
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 |
26 | #include "perl.h" |
27 | |
27da23d5 |
28 | static const char unees[] = |
29 | "Malformed UTF-8 character (unexpected end of string)"; |
901b21bf |
30 | |
ccfc67b7 |
31 | /* |
32 | =head1 Unicode Support |
a0ed51b3 |
33 | |
166f8a29 |
34 | This file contains various utility functions for manipulating UTF8-encoded |
35 | strings. For the uninitiated, this is a method of representing arbitrary |
61296642 |
36 | Unicode characters as a variable number of bytes, in such a way that |
56da48f7 |
37 | characters in the ASCII range are unmodified, and a zero byte never appears |
38 | within non-zero characters. |
166f8a29 |
39 | |
b851fbc1 |
40 | =for apidoc A|U8 *|uvuni_to_utf8_flags|U8 *d|UV uv|UV flags |
eebe1485 |
41 | |
1e54db1a |
42 | Adds the UTF-8 representation of the Unicode codepoint C<uv> to the end |
89ebb4a3 |
43 | of the string C<d>; C<d> should be have at least C<UTF8_MAXBYTES+1> free |
eebe1485 |
44 | bytes available. The return value is the pointer to the byte after the |
9041c2e3 |
45 | end of the new character. In other words, |
eebe1485 |
46 | |
b851fbc1 |
47 | d = uvuni_to_utf8_flags(d, uv, flags); |
48 | |
49 | or, in most cases, |
50 | |
9041c2e3 |
51 | d = uvuni_to_utf8(d, uv); |
eebe1485 |
52 | |
b851fbc1 |
53 | (which is equivalent to) |
54 | |
55 | d = uvuni_to_utf8_flags(d, uv, 0); |
56 | |
eebe1485 |
57 | is the recommended Unicode-aware way of saying |
58 | |
59 | *(d++) = uv; |
60 | |
61 | =cut |
62 | */ |
63 | |
dfe13c55 |
64 | U8 * |
b851fbc1 |
65 | Perl_uvuni_to_utf8_flags(pTHX_ U8 *d, UV uv, UV flags) |
a0ed51b3 |
66 | { |
62961d2e |
67 | if (ckWARN(WARN_UTF8)) { |
b851fbc1 |
68 | if (UNICODE_IS_SURROGATE(uv) && |
69 | !(flags & UNICODE_ALLOW_SURROGATE)) |
9014280d |
70 | Perl_warner(aTHX_ packWARN(WARN_UTF8), "UTF-16 surrogate 0x%04"UVxf, uv); |
b851fbc1 |
71 | else if ( |
72 | ((uv >= 0xFDD0 && uv <= 0xFDEF && |
73 | !(flags & UNICODE_ALLOW_FDD0)) |
74 | || |
c867b360 |
75 | ((uv & 0xFFFE) == 0xFFFE && /* Either FFFE or FFFF. */ |
b851fbc1 |
76 | !(flags & UNICODE_ALLOW_FFFF))) && |
77 | /* UNICODE_ALLOW_SUPER includes |
2a20b9da |
78 | * FFFEs and FFFFs beyond 0x10FFFF. */ |
b851fbc1 |
79 | ((uv <= PERL_UNICODE_MAX) || |
80 | !(flags & UNICODE_ALLOW_SUPER)) |
81 | ) |
9014280d |
82 | Perl_warner(aTHX_ packWARN(WARN_UTF8), |
507b9800 |
83 | "Unicode character 0x%04"UVxf" is illegal", uv); |
84 | } |
c4d5f83a |
85 | if (UNI_IS_INVARIANT(uv)) { |
eb160463 |
86 | *d++ = (U8)UTF_TO_NATIVE(uv); |
a0ed51b3 |
87 | return d; |
88 | } |
2d331972 |
89 | #if defined(EBCDIC) |
1d72bdf6 |
90 | else { |
91 | STRLEN len = UNISKIP(uv); |
92 | U8 *p = d+len-1; |
93 | while (p > d) { |
eb160463 |
94 | *p-- = (U8)UTF_TO_NATIVE((uv & UTF_CONTINUATION_MASK) | UTF_CONTINUATION_MARK); |
1d72bdf6 |
95 | uv >>= UTF_ACCUMULATION_SHIFT; |
96 | } |
eb160463 |
97 | *p = (U8)UTF_TO_NATIVE((uv & UTF_START_MASK(len)) | UTF_START_MARK(len)); |
1d72bdf6 |
98 | return d+len; |
99 | } |
100 | #else /* Non loop style */ |
a0ed51b3 |
101 | if (uv < 0x800) { |
eb160463 |
102 | *d++ = (U8)(( uv >> 6) | 0xc0); |
103 | *d++ = (U8)(( uv & 0x3f) | 0x80); |
a0ed51b3 |
104 | return d; |
105 | } |
106 | if (uv < 0x10000) { |
eb160463 |
107 | *d++ = (U8)(( uv >> 12) | 0xe0); |
108 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); |
109 | *d++ = (U8)(( uv & 0x3f) | 0x80); |
a0ed51b3 |
110 | return d; |
111 | } |
112 | if (uv < 0x200000) { |
eb160463 |
113 | *d++ = (U8)(( uv >> 18) | 0xf0); |
114 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); |
115 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); |
116 | *d++ = (U8)(( uv & 0x3f) | 0x80); |
a0ed51b3 |
117 | return d; |
118 | } |
119 | if (uv < 0x4000000) { |
eb160463 |
120 | *d++ = (U8)(( uv >> 24) | 0xf8); |
121 | *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80); |
122 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); |
123 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); |
124 | *d++ = (U8)(( uv & 0x3f) | 0x80); |
a0ed51b3 |
125 | return d; |
126 | } |
127 | if (uv < 0x80000000) { |
eb160463 |
128 | *d++ = (U8)(( uv >> 30) | 0xfc); |
129 | *d++ = (U8)(((uv >> 24) & 0x3f) | 0x80); |
130 | *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80); |
131 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); |
132 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); |
133 | *d++ = (U8)(( uv & 0x3f) | 0x80); |
a0ed51b3 |
134 | return d; |
135 | } |
6b8eaf93 |
136 | #ifdef HAS_QUAD |
d7578b48 |
137 | if (uv < UTF8_QUAD_MAX) |
a0ed51b3 |
138 | #endif |
139 | { |
eb160463 |
140 | *d++ = 0xfe; /* Can't match U+FEFF! */ |
141 | *d++ = (U8)(((uv >> 30) & 0x3f) | 0x80); |
142 | *d++ = (U8)(((uv >> 24) & 0x3f) | 0x80); |
143 | *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80); |
144 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); |
145 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); |
146 | *d++ = (U8)(( uv & 0x3f) | 0x80); |
a0ed51b3 |
147 | return d; |
148 | } |
6b8eaf93 |
149 | #ifdef HAS_QUAD |
a0ed51b3 |
150 | { |
eb160463 |
151 | *d++ = 0xff; /* Can't match U+FFFE! */ |
152 | *d++ = 0x80; /* 6 Reserved bits */ |
153 | *d++ = (U8)(((uv >> 60) & 0x0f) | 0x80); /* 2 Reserved bits */ |
154 | *d++ = (U8)(((uv >> 54) & 0x3f) | 0x80); |
155 | *d++ = (U8)(((uv >> 48) & 0x3f) | 0x80); |
156 | *d++ = (U8)(((uv >> 42) & 0x3f) | 0x80); |
157 | *d++ = (U8)(((uv >> 36) & 0x3f) | 0x80); |
158 | *d++ = (U8)(((uv >> 30) & 0x3f) | 0x80); |
159 | *d++ = (U8)(((uv >> 24) & 0x3f) | 0x80); |
160 | *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80); |
161 | *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80); |
162 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); |
163 | *d++ = (U8)(( uv & 0x3f) | 0x80); |
a0ed51b3 |
164 | return d; |
165 | } |
166 | #endif |
1d72bdf6 |
167 | #endif /* Loop style */ |
a0ed51b3 |
168 | } |
b851fbc1 |
169 | |
170 | U8 * |
171 | Perl_uvuni_to_utf8(pTHX_ U8 *d, UV uv) |
172 | { |
173 | return Perl_uvuni_to_utf8_flags(aTHX_ d, uv, 0); |
174 | } |
9041c2e3 |
175 | |
176 | |
177 | /* |
7fc63493 |
178 | =for apidoc A|STRLEN|is_utf8_char|const U8 *s |
eebe1485 |
179 | |
5da9da9e |
180 | Tests if some arbitrary number of bytes begins in a valid UTF-8 |
82686b01 |
181 | character. Note that an INVARIANT (i.e. ASCII) character is a valid |
182 | UTF-8 character. The actual number of bytes in the UTF-8 character |
183 | will be returned if it is valid, otherwise 0. |
9041c2e3 |
184 | |
82686b01 |
185 | =cut */ |
067a85ef |
186 | STRLEN |
7fc63493 |
187 | Perl_is_utf8_char(pTHX_ const U8 *s) |
386d01d6 |
188 | { |
189 | U8 u = *s; |
067a85ef |
190 | STRLEN slen, len; |
191 | UV uv, ouv; |
386d01d6 |
192 | |
1d72bdf6 |
193 | if (UTF8_IS_INVARIANT(u)) |
386d01d6 |
194 | return 1; |
195 | |
60006e79 |
196 | if (!UTF8_IS_START(u)) |
386d01d6 |
197 | return 0; |
198 | |
9f07fdcd |
199 | len = UTF8SKIP(s); |
386d01d6 |
200 | |
60006e79 |
201 | if (len < 2 || !UTF8_IS_CONTINUATION(s[1])) |
067a85ef |
202 | return 0; |
203 | |
386d01d6 |
204 | slen = len - 1; |
205 | s++; |
6eb6869e |
206 | u &= UTF_START_MASK(len); |
1d72bdf6 |
207 | uv = u; |
067a85ef |
208 | ouv = uv; |
386d01d6 |
209 | while (slen--) { |
60006e79 |
210 | if (!UTF8_IS_CONTINUATION(*s)) |
386d01d6 |
211 | return 0; |
8850bf83 |
212 | uv = UTF8_ACCUMULATE(uv, *s); |
209a9bc1 |
213 | if (uv < ouv) |
067a85ef |
214 | return 0; |
215 | ouv = uv; |
386d01d6 |
216 | s++; |
217 | } |
067a85ef |
218 | |
eb160463 |
219 | if ((STRLEN)UNISKIP(uv) < len) |
067a85ef |
220 | return 0; |
221 | |
386d01d6 |
222 | return len; |
223 | } |
224 | |
6662521e |
225 | /* |
7fc63493 |
226 | =for apidoc A|bool|is_utf8_string|const U8 *s|STRLEN len |
6662521e |
227 | |
c9ada85f |
228 | Returns true if first C<len> bytes of the given string form a valid |
1e54db1a |
229 | UTF-8 string, false otherwise. Note that 'a valid UTF-8 string' does |
230 | not mean 'a string that contains code points above 0x7F encoded in UTF-8' |
231 | because a valid ASCII string is a valid UTF-8 string. |
6662521e |
232 | |
233 | =cut |
234 | */ |
235 | |
8e84507e |
236 | bool |
7fc63493 |
237 | Perl_is_utf8_string(pTHX_ const U8 *s, STRLEN len) |
6662521e |
238 | { |
7fc63493 |
239 | const U8* x = s; |
240 | const U8* send; |
067a85ef |
241 | STRLEN c; |
242 | |
61468b03 |
243 | if (!len && s) |
e1ec3a88 |
244 | len = strlen((const char *)s); |
1aa99e6b |
245 | send = s + len; |
246 | |
6662521e |
247 | while (x < send) { |
1acdb0da |
248 | /* Inline the easy bits of is_utf8_char() here for speed... */ |
249 | if (UTF8_IS_INVARIANT(*x)) |
250 | c = 1; |
251 | else if (!UTF8_IS_START(*x)) |
252 | return FALSE; |
253 | else { |
254 | /* ... and call is_utf8_char() only if really needed. */ |
255 | c = is_utf8_char(x); |
256 | if (!c) |
257 | return FALSE; |
258 | } |
6662521e |
259 | x += c; |
6662521e |
260 | } |
60006e79 |
261 | if (x != send) |
262 | return FALSE; |
067a85ef |
263 | |
264 | return TRUE; |
6662521e |
265 | } |
266 | |
67e989fb |
267 | /* |
7fc63493 |
268 | =for apidoc A|bool|is_utf8_string_loc|const U8 *s|STRLEN len|const U8 **p |
81cd54e3 |
269 | |
270 | Like is_ut8_string but store the location of the failure in |
271 | the last argument. |
272 | |
273 | =cut |
274 | */ |
275 | |
276 | bool |
7fc63493 |
277 | Perl_is_utf8_string_loc(pTHX_ const U8 *s, STRLEN len, const U8 **p) |
81cd54e3 |
278 | { |
7fc63493 |
279 | const U8* x = s; |
280 | const U8* send; |
81cd54e3 |
281 | STRLEN c; |
282 | |
61468b03 |
283 | if (!len && s) |
e1ec3a88 |
284 | len = strlen((const char *)s); |
81cd54e3 |
285 | send = s + len; |
286 | |
287 | while (x < send) { |
288 | /* Inline the easy bits of is_utf8_char() here for speed... */ |
289 | if (UTF8_IS_INVARIANT(*x)) |
290 | c = 1; |
291 | else if (!UTF8_IS_START(*x)) { |
292 | if (p) |
293 | *p = x; |
294 | return FALSE; |
295 | } |
296 | else { |
297 | /* ... and call is_utf8_char() only if really needed. */ |
298 | c = is_utf8_char(x); |
299 | if (!c) { |
300 | if (p) |
301 | *p = x; |
302 | return FALSE; |
303 | } |
304 | } |
305 | x += c; |
306 | } |
307 | if (x != send) { |
308 | if (p) |
309 | *p = x; |
310 | return FALSE; |
311 | } |
312 | |
313 | return TRUE; |
314 | } |
315 | |
316 | /* |
7fc63493 |
317 | =for apidoc A|UV|utf8n_to_uvuni|const U8 *s|STRLEN curlen|STRLEN *retlen|U32 flags |
67e989fb |
318 | |
9041c2e3 |
319 | Bottom level UTF-8 decode routine. |
320 | Returns the unicode code point value of the first character in the string C<s> |
1e54db1a |
321 | which is assumed to be in UTF-8 encoding and no longer than C<curlen>; |
7df053ec |
322 | C<retlen> will be set to the length, in bytes, of that character. |
67e989fb |
323 | |
1e54db1a |
324 | If C<s> does not point to a well-formed UTF-8 character, the behaviour |
dcad2880 |
325 | is dependent on the value of C<flags>: if it contains UTF8_CHECK_ONLY, |
326 | it is assumed that the caller will raise a warning, and this function |
28d3d195 |
327 | will silently just set C<retlen> to C<-1> and return zero. If the |
328 | C<flags> does not contain UTF8_CHECK_ONLY, warnings about |
329 | malformations will be given, C<retlen> will be set to the expected |
330 | length of the UTF-8 character in bytes, and zero will be returned. |
331 | |
332 | The C<flags> can also contain various flags to allow deviations from |
333 | the strict UTF-8 encoding (see F<utf8.h>). |
67e989fb |
334 | |
9041c2e3 |
335 | Most code should use utf8_to_uvchr() rather than call this directly. |
336 | |
37607a96 |
337 | =cut |
338 | */ |
67e989fb |
339 | |
a0ed51b3 |
340 | UV |
7fc63493 |
341 | Perl_utf8n_to_uvuni(pTHX_ const U8 *s, STRLEN curlen, STRLEN *retlen, U32 flags) |
a0ed51b3 |
342 | { |
7fc63493 |
343 | const U8 *s0 = s; |
9c5ffd7c |
344 | UV uv = *s, ouv = 0; |
ba210ebe |
345 | STRLEN len = 1; |
7fc63493 |
346 | const bool dowarn = ckWARN_d(WARN_UTF8); |
347 | const UV startbyte = *s; |
ba210ebe |
348 | STRLEN expectlen = 0; |
a0dbb045 |
349 | U32 warning = 0; |
350 | |
351 | /* This list is a superset of the UTF8_ALLOW_XXX. */ |
352 | |
353 | #define UTF8_WARN_EMPTY 1 |
354 | #define UTF8_WARN_CONTINUATION 2 |
355 | #define UTF8_WARN_NON_CONTINUATION 3 |
356 | #define UTF8_WARN_FE_FF 4 |
357 | #define UTF8_WARN_SHORT 5 |
358 | #define UTF8_WARN_OVERFLOW 6 |
359 | #define UTF8_WARN_SURROGATE 7 |
c867b360 |
360 | #define UTF8_WARN_LONG 8 |
361 | #define UTF8_WARN_FFFF 9 /* Also FFFE. */ |
a0dbb045 |
362 | |
363 | if (curlen == 0 && |
364 | !(flags & UTF8_ALLOW_EMPTY)) { |
365 | warning = UTF8_WARN_EMPTY; |
0c443dc2 |
366 | goto malformed; |
367 | } |
368 | |
1d72bdf6 |
369 | if (UTF8_IS_INVARIANT(uv)) { |
a0ed51b3 |
370 | if (retlen) |
371 | *retlen = 1; |
c4d5f83a |
372 | return (UV) (NATIVE_TO_UTF(*s)); |
a0ed51b3 |
373 | } |
67e989fb |
374 | |
421a8bf2 |
375 | if (UTF8_IS_CONTINUATION(uv) && |
fcc8fcf6 |
376 | !(flags & UTF8_ALLOW_CONTINUATION)) { |
a0dbb045 |
377 | warning = UTF8_WARN_CONTINUATION; |
ba210ebe |
378 | goto malformed; |
379 | } |
380 | |
421a8bf2 |
381 | if (UTF8_IS_START(uv) && curlen > 1 && !UTF8_IS_CONTINUATION(s[1]) && |
fcc8fcf6 |
382 | !(flags & UTF8_ALLOW_NON_CONTINUATION)) { |
a0dbb045 |
383 | warning = UTF8_WARN_NON_CONTINUATION; |
ba210ebe |
384 | goto malformed; |
385 | } |
9041c2e3 |
386 | |
1d72bdf6 |
387 | #ifdef EBCDIC |
75383841 |
388 | uv = NATIVE_TO_UTF(uv); |
1d72bdf6 |
389 | #else |
fcc8fcf6 |
390 | if ((uv == 0xfe || uv == 0xff) && |
391 | !(flags & UTF8_ALLOW_FE_FF)) { |
a0dbb045 |
392 | warning = UTF8_WARN_FE_FF; |
ba210ebe |
393 | goto malformed; |
a0ed51b3 |
394 | } |
1d72bdf6 |
395 | #endif |
396 | |
ba210ebe |
397 | if (!(uv & 0x20)) { len = 2; uv &= 0x1f; } |
398 | else if (!(uv & 0x10)) { len = 3; uv &= 0x0f; } |
399 | else if (!(uv & 0x08)) { len = 4; uv &= 0x07; } |
400 | else if (!(uv & 0x04)) { len = 5; uv &= 0x03; } |
1d72bdf6 |
401 | #ifdef EBCDIC |
402 | else if (!(uv & 0x02)) { len = 6; uv &= 0x01; } |
403 | else { len = 7; uv &= 0x01; } |
404 | #else |
ba210ebe |
405 | else if (!(uv & 0x02)) { len = 6; uv &= 0x01; } |
406 | else if (!(uv & 0x01)) { len = 7; uv = 0; } |
1d72bdf6 |
407 | else { len = 13; uv = 0; } /* whoa! */ |
408 | #endif |
409 | |
a0ed51b3 |
410 | if (retlen) |
411 | *retlen = len; |
9041c2e3 |
412 | |
ba210ebe |
413 | expectlen = len; |
414 | |
fcc8fcf6 |
415 | if ((curlen < expectlen) && |
416 | !(flags & UTF8_ALLOW_SHORT)) { |
a0dbb045 |
417 | warning = UTF8_WARN_SHORT; |
ba210ebe |
418 | goto malformed; |
419 | } |
420 | |
421 | len--; |
a0ed51b3 |
422 | s++; |
ba210ebe |
423 | ouv = uv; |
424 | |
a0ed51b3 |
425 | while (len--) { |
421a8bf2 |
426 | if (!UTF8_IS_CONTINUATION(*s) && |
427 | !(flags & UTF8_ALLOW_NON_CONTINUATION)) { |
a0dbb045 |
428 | s--; |
429 | warning = UTF8_WARN_NON_CONTINUATION; |
ba210ebe |
430 | goto malformed; |
a0ed51b3 |
431 | } |
432 | else |
8850bf83 |
433 | uv = UTF8_ACCUMULATE(uv, *s); |
a0dbb045 |
434 | if (!(uv > ouv)) { |
435 | /* These cannot be allowed. */ |
436 | if (uv == ouv) { |
75dbc644 |
437 | if (expectlen != 13 && !(flags & UTF8_ALLOW_LONG)) { |
a0dbb045 |
438 | warning = UTF8_WARN_LONG; |
439 | goto malformed; |
440 | } |
441 | } |
442 | else { /* uv < ouv */ |
443 | /* This cannot be allowed. */ |
444 | warning = UTF8_WARN_OVERFLOW; |
445 | goto malformed; |
446 | } |
ba210ebe |
447 | } |
448 | s++; |
449 | ouv = uv; |
450 | } |
451 | |
421a8bf2 |
452 | if (UNICODE_IS_SURROGATE(uv) && |
fcc8fcf6 |
453 | !(flags & UTF8_ALLOW_SURROGATE)) { |
a0dbb045 |
454 | warning = UTF8_WARN_SURROGATE; |
ba210ebe |
455 | goto malformed; |
eb160463 |
456 | } else if ((expectlen > (STRLEN)UNISKIP(uv)) && |
fcc8fcf6 |
457 | !(flags & UTF8_ALLOW_LONG)) { |
a0dbb045 |
458 | warning = UTF8_WARN_LONG; |
ba210ebe |
459 | goto malformed; |
421a8bf2 |
460 | } else if (UNICODE_IS_ILLEGAL(uv) && |
a9917092 |
461 | !(flags & UTF8_ALLOW_FFFF)) { |
a0dbb045 |
462 | warning = UTF8_WARN_FFFF; |
a9917092 |
463 | goto malformed; |
a0ed51b3 |
464 | } |
ba210ebe |
465 | |
a0ed51b3 |
466 | return uv; |
ba210ebe |
467 | |
468 | malformed: |
469 | |
fcc8fcf6 |
470 | if (flags & UTF8_CHECK_ONLY) { |
ba210ebe |
471 | if (retlen) |
cc366d4b |
472 | *retlen = -1; |
ba210ebe |
473 | return 0; |
474 | } |
475 | |
a0dbb045 |
476 | if (dowarn) { |
477 | SV* sv = sv_2mortal(newSVpv("Malformed UTF-8 character ", 0)); |
478 | |
479 | switch (warning) { |
480 | case 0: /* Intentionally empty. */ break; |
481 | case UTF8_WARN_EMPTY: |
54667de8 |
482 | Perl_sv_catpv(aTHX_ sv, "(empty string)"); |
a0dbb045 |
483 | break; |
484 | case UTF8_WARN_CONTINUATION: |
097fb8e2 |
485 | Perl_sv_catpvf(aTHX_ sv, "(unexpected continuation byte 0x%02"UVxf", with no preceding start byte)", uv); |
a0dbb045 |
486 | break; |
487 | case UTF8_WARN_NON_CONTINUATION: |
097fb8e2 |
488 | if (s == s0) |
489 | Perl_sv_catpvf(aTHX_ sv, "(unexpected non-continuation byte 0x%02"UVxf", immediately after start byte 0x%02"UVxf")", |
490 | (UV)s[1], startbyte); |
491 | else |
492 | Perl_sv_catpvf(aTHX_ sv, "(unexpected non-continuation byte 0x%02"UVxf", %d byte%s after start byte 0x%02"UVxf", expected %d bytes)", |
493 | (UV)s[1], s - s0, s - s0 > 1 ? "s" : "", startbyte, expectlen); |
494 | |
a0dbb045 |
495 | break; |
496 | case UTF8_WARN_FE_FF: |
497 | Perl_sv_catpvf(aTHX_ sv, "(byte 0x%02"UVxf")", uv); |
498 | break; |
499 | case UTF8_WARN_SHORT: |
097fb8e2 |
500 | Perl_sv_catpvf(aTHX_ sv, "(%d byte%s, need %d, after start byte 0x%02"UVxf")", |
501 | curlen, curlen == 1 ? "" : "s", expectlen, startbyte); |
b31f83c2 |
502 | expectlen = curlen; /* distance for caller to skip */ |
a0dbb045 |
503 | break; |
504 | case UTF8_WARN_OVERFLOW: |
097fb8e2 |
505 | Perl_sv_catpvf(aTHX_ sv, "(overflow at 0x%"UVxf", byte 0x%02x, after start byte 0x%02"UVxf")", |
506 | ouv, *s, startbyte); |
a0dbb045 |
507 | break; |
508 | case UTF8_WARN_SURROGATE: |
509 | Perl_sv_catpvf(aTHX_ sv, "(UTF-16 surrogate 0x%04"UVxf")", uv); |
510 | break; |
a0dbb045 |
511 | case UTF8_WARN_LONG: |
097fb8e2 |
512 | Perl_sv_catpvf(aTHX_ sv, "(%d byte%s, need %d, after start byte 0x%02"UVxf")", |
513 | expectlen, expectlen == 1 ? "": "s", UNISKIP(uv), startbyte); |
a0dbb045 |
514 | break; |
515 | case UTF8_WARN_FFFF: |
516 | Perl_sv_catpvf(aTHX_ sv, "(character 0x%04"UVxf")", uv); |
517 | break; |
518 | default: |
54667de8 |
519 | Perl_sv_catpv(aTHX_ sv, "(unknown reason)"); |
a0dbb045 |
520 | break; |
521 | } |
522 | |
523 | if (warning) { |
524 | char *s = SvPVX(sv); |
525 | |
526 | if (PL_op) |
9014280d |
527 | Perl_warner(aTHX_ packWARN(WARN_UTF8), |
53e06cf0 |
528 | "%s in %s", s, OP_DESC(PL_op)); |
a0dbb045 |
529 | else |
9014280d |
530 | Perl_warner(aTHX_ packWARN(WARN_UTF8), "%s", s); |
a0dbb045 |
531 | } |
532 | } |
533 | |
ba210ebe |
534 | if (retlen) |
28d3d195 |
535 | *retlen = expectlen ? expectlen : len; |
ba210ebe |
536 | |
28d3d195 |
537 | return 0; |
a0ed51b3 |
538 | } |
539 | |
8e84507e |
540 | /* |
7fc63493 |
541 | =for apidoc A|UV|utf8_to_uvchr|const U8 *s|STRLEN *retlen |
9041c2e3 |
542 | |
543 | Returns the native character value of the first character in the string C<s> |
1e54db1a |
544 | which is assumed to be in UTF-8 encoding; C<retlen> will be set to the |
9041c2e3 |
545 | length, in bytes, of that character. |
546 | |
1e54db1a |
547 | If C<s> does not point to a well-formed UTF-8 character, zero is |
9041c2e3 |
548 | returned and retlen is set, if possible, to -1. |
549 | |
550 | =cut |
551 | */ |
552 | |
553 | UV |
7fc63493 |
554 | Perl_utf8_to_uvchr(pTHX_ const U8 *s, STRLEN *retlen) |
9041c2e3 |
555 | { |
89ebb4a3 |
556 | return Perl_utf8n_to_uvchr(aTHX_ s, UTF8_MAXBYTES, retlen, |
872c91ae |
557 | ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY); |
9041c2e3 |
558 | } |
559 | |
560 | /* |
7fc63493 |
561 | =for apidoc A|UV|utf8_to_uvuni|const U8 *s|STRLEN *retlen |
9041c2e3 |
562 | |
563 | Returns the Unicode code point of the first character in the string C<s> |
1e54db1a |
564 | which is assumed to be in UTF-8 encoding; C<retlen> will be set to the |
9041c2e3 |
565 | length, in bytes, of that character. |
566 | |
567 | This function should only be used when returned UV is considered |
568 | an index into the Unicode semantic tables (e.g. swashes). |
569 | |
1e54db1a |
570 | If C<s> does not point to a well-formed UTF-8 character, zero is |
ba210ebe |
571 | returned and retlen is set, if possible, to -1. |
8e84507e |
572 | |
573 | =cut |
574 | */ |
575 | |
576 | UV |
7fc63493 |
577 | Perl_utf8_to_uvuni(pTHX_ const U8 *s, STRLEN *retlen) |
8e84507e |
578 | { |
9041c2e3 |
579 | /* Call the low level routine asking for checks */ |
89ebb4a3 |
580 | return Perl_utf8n_to_uvuni(aTHX_ s, UTF8_MAXBYTES, retlen, |
872c91ae |
581 | ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY); |
8e84507e |
582 | } |
583 | |
b76347f2 |
584 | /* |
35a4481c |
585 | =for apidoc A|STRLEN|utf8_length|const U8 *s|const U8 *e |
b76347f2 |
586 | |
587 | Return the length of the UTF-8 char encoded string C<s> in characters. |
02eb7b47 |
588 | Stops at C<e> (inclusive). If C<e E<lt> s> or if the scan would end |
589 | up past C<e>, croaks. |
b76347f2 |
590 | |
591 | =cut |
592 | */ |
593 | |
594 | STRLEN |
35a4481c |
595 | Perl_utf8_length(pTHX_ const U8 *s, const U8 *e) |
b76347f2 |
596 | { |
597 | STRLEN len = 0; |
598 | |
8850bf83 |
599 | /* Note: cannot use UTF8_IS_...() too eagerly here since e.g. |
600 | * the bitops (especially ~) can create illegal UTF-8. |
601 | * In other words: in Perl UTF-8 is not just for Unicode. */ |
602 | |
901b21bf |
603 | if (e < s) { |
604 | if (ckWARN_d(WARN_UTF8)) { |
605 | if (PL_op) |
606 | Perl_warner(aTHX_ packWARN(WARN_UTF8), |
607 | "%s in %s", unees, OP_DESC(PL_op)); |
608 | else |
609 | Perl_warner(aTHX_ packWARN(WARN_UTF8), unees); |
610 | } |
611 | return 0; |
612 | } |
b76347f2 |
613 | while (s < e) { |
02eb7b47 |
614 | U8 t = UTF8SKIP(s); |
b76347f2 |
615 | |
901b21bf |
616 | if (e - s < t) { |
617 | if (ckWARN_d(WARN_UTF8)) { |
618 | if (PL_op) |
619 | Perl_warner(aTHX_ packWARN(WARN_UTF8), |
620 | unees, OP_DESC(PL_op)); |
621 | else |
622 | Perl_warner(aTHX_ packWARN(WARN_UTF8), unees); |
623 | } |
624 | return len; |
625 | } |
b76347f2 |
626 | s += t; |
627 | len++; |
628 | } |
629 | |
630 | return len; |
631 | } |
632 | |
b06226ff |
633 | /* |
35a4481c |
634 | =for apidoc A|IV|utf8_distance|const U8 *a|const U8 *b |
b06226ff |
635 | |
1e54db1a |
636 | Returns the number of UTF-8 characters between the UTF-8 pointers C<a> |
b06226ff |
637 | and C<b>. |
638 | |
639 | WARNING: use only if you *know* that the pointers point inside the |
640 | same UTF-8 buffer. |
641 | |
37607a96 |
642 | =cut |
643 | */ |
a0ed51b3 |
644 | |
02eb7b47 |
645 | IV |
35a4481c |
646 | Perl_utf8_distance(pTHX_ const U8 *a, const U8 *b) |
a0ed51b3 |
647 | { |
02eb7b47 |
648 | IV off = 0; |
649 | |
8850bf83 |
650 | /* Note: cannot use UTF8_IS_...() too eagerly here since e.g. |
651 | * the bitops (especially ~) can create illegal UTF-8. |
652 | * In other words: in Perl UTF-8 is not just for Unicode. */ |
653 | |
a0ed51b3 |
654 | if (a < b) { |
655 | while (a < b) { |
35a4481c |
656 | const U8 c = UTF8SKIP(a); |
02eb7b47 |
657 | |
901b21bf |
658 | if (b - a < c) { |
659 | if (ckWARN_d(WARN_UTF8)) { |
660 | if (PL_op) |
661 | Perl_warner(aTHX_ packWARN(WARN_UTF8), |
662 | "%s in %s", unees, OP_DESC(PL_op)); |
663 | else |
664 | Perl_warner(aTHX_ packWARN(WARN_UTF8), unees); |
665 | } |
666 | return off; |
667 | } |
02eb7b47 |
668 | a += c; |
a0ed51b3 |
669 | off--; |
670 | } |
671 | } |
672 | else { |
673 | while (b < a) { |
02eb7b47 |
674 | U8 c = UTF8SKIP(b); |
675 | |
901b21bf |
676 | if (a - b < c) { |
677 | if (ckWARN_d(WARN_UTF8)) { |
678 | if (PL_op) |
679 | Perl_warner(aTHX_ packWARN(WARN_UTF8), |
680 | "%s in %s", unees, OP_DESC(PL_op)); |
681 | else |
682 | Perl_warner(aTHX_ packWARN(WARN_UTF8), unees); |
683 | } |
684 | return off; |
685 | } |
02eb7b47 |
686 | b += c; |
a0ed51b3 |
687 | off++; |
688 | } |
689 | } |
02eb7b47 |
690 | |
a0ed51b3 |
691 | return off; |
692 | } |
693 | |
b06226ff |
694 | /* |
37607a96 |
695 | =for apidoc A|U8 *|utf8_hop|U8 *s|I32 off |
b06226ff |
696 | |
8850bf83 |
697 | Return the UTF-8 pointer C<s> displaced by C<off> characters, either |
698 | forward or backward. |
b06226ff |
699 | |
700 | WARNING: do not use the following unless you *know* C<off> is within |
8850bf83 |
701 | the UTF-8 data pointed to by C<s> *and* that on entry C<s> is aligned |
702 | on the first byte of character or just after the last byte of a character. |
b06226ff |
703 | |
37607a96 |
704 | =cut |
705 | */ |
a0ed51b3 |
706 | |
707 | U8 * |
864dbfa3 |
708 | Perl_utf8_hop(pTHX_ U8 *s, I32 off) |
a0ed51b3 |
709 | { |
8850bf83 |
710 | /* Note: cannot use UTF8_IS_...() too eagerly here since e.g |
711 | * the bitops (especially ~) can create illegal UTF-8. |
712 | * In other words: in Perl UTF-8 is not just for Unicode. */ |
713 | |
a0ed51b3 |
714 | if (off >= 0) { |
715 | while (off--) |
716 | s += UTF8SKIP(s); |
717 | } |
718 | else { |
719 | while (off++) { |
720 | s--; |
8850bf83 |
721 | while (UTF8_IS_CONTINUATION(*s)) |
722 | s--; |
a0ed51b3 |
723 | } |
724 | } |
725 | return s; |
726 | } |
727 | |
6940069f |
728 | /* |
eebe1485 |
729 | =for apidoc A|U8 *|utf8_to_bytes|U8 *s|STRLEN *len |
6940069f |
730 | |
1e54db1a |
731 | Converts a string C<s> of length C<len> from UTF-8 into byte encoding. |
246fae53 |
732 | Unlike C<bytes_to_utf8>, this over-writes the original string, and |
733 | updates len to contain the new length. |
67e989fb |
734 | Returns zero on failure, setting C<len> to -1. |
6940069f |
735 | |
736 | =cut |
737 | */ |
738 | |
739 | U8 * |
37607a96 |
740 | Perl_utf8_to_bytes(pTHX_ U8 *s, STRLEN *len) |
6940069f |
741 | { |
6940069f |
742 | U8 *send; |
743 | U8 *d; |
dcad2880 |
744 | U8 *save = s; |
246fae53 |
745 | |
1e54db1a |
746 | /* ensure valid UTF-8 and chars < 256 before updating string */ |
dcad2880 |
747 | for (send = s + *len; s < send; ) { |
748 | U8 c = *s++; |
749 | |
1d72bdf6 |
750 | if (!UTF8_IS_INVARIANT(c) && |
751 | (!UTF8_IS_DOWNGRADEABLE_START(c) || (s >= send) |
752 | || !(c = *s++) || !UTF8_IS_CONTINUATION(c))) { |
dcad2880 |
753 | *len = -1; |
754 | return 0; |
755 | } |
246fae53 |
756 | } |
dcad2880 |
757 | |
758 | d = s = save; |
6940069f |
759 | while (s < send) { |
ed646e6e |
760 | STRLEN ulen; |
9041c2e3 |
761 | *d++ = (U8)utf8_to_uvchr(s, &ulen); |
ed646e6e |
762 | s += ulen; |
6940069f |
763 | } |
764 | *d = '\0'; |
246fae53 |
765 | *len = d - save; |
6940069f |
766 | return save; |
767 | } |
768 | |
769 | /* |
e1ec3a88 |
770 | =for apidoc A|U8 *|bytes_from_utf8|const U8 *s|STRLEN *len|bool *is_utf8 |
f9a63242 |
771 | |
1e54db1a |
772 | Converts a string C<s> of length C<len> from UTF-8 into byte encoding. |
35a4481c |
773 | Unlike C<utf8_to_bytes> but like C<bytes_to_utf8>, returns a pointer to |
ef9edfd0 |
774 | the newly-created string, and updates C<len> to contain the new |
775 | length. Returns the original string if no conversion occurs, C<len> |
776 | is unchanged. Do nothing if C<is_utf8> points to 0. Sets C<is_utf8> to |
777 | 0 if C<s> is converted or contains all 7bit characters. |
f9a63242 |
778 | |
37607a96 |
779 | =cut |
780 | */ |
f9a63242 |
781 | |
782 | U8 * |
e1ec3a88 |
783 | Perl_bytes_from_utf8(pTHX_ const U8 *s, STRLEN *len, bool *is_utf8) |
f9a63242 |
784 | { |
f9a63242 |
785 | U8 *d; |
e1ec3a88 |
786 | const U8 *start = s; |
787 | const U8 *send; |
f9a63242 |
788 | I32 count = 0; |
789 | |
790 | if (!*is_utf8) |
73d840c0 |
791 | return (U8 *)start; |
f9a63242 |
792 | |
1e54db1a |
793 | /* ensure valid UTF-8 and chars < 256 before converting string */ |
f9a63242 |
794 | for (send = s + *len; s < send;) { |
e1ec3a88 |
795 | U8 c = *s++; |
1d72bdf6 |
796 | if (!UTF8_IS_INVARIANT(c)) { |
db42d148 |
797 | if (UTF8_IS_DOWNGRADEABLE_START(c) && s < send && |
798 | (c = *s++) && UTF8_IS_CONTINUATION(c)) |
799 | count++; |
800 | else |
73d840c0 |
801 | return (U8 *)start; |
db42d148 |
802 | } |
f9a63242 |
803 | } |
804 | |
805 | *is_utf8 = 0; |
806 | |
f9a63242 |
807 | Newz(801, d, (*len) - count + 1, U8); |
ef9edfd0 |
808 | s = start; start = d; |
f9a63242 |
809 | while (s < send) { |
810 | U8 c = *s++; |
c4d5f83a |
811 | if (!UTF8_IS_INVARIANT(c)) { |
812 | /* Then it is two-byte encoded */ |
813 | c = UTF8_ACCUMULATE(NATIVE_TO_UTF(c), *s++); |
814 | c = ASCII_TO_NATIVE(c); |
815 | } |
816 | *d++ = c; |
f9a63242 |
817 | } |
818 | *d = '\0'; |
819 | *len = d - start; |
73d840c0 |
820 | return (U8 *)start; |
f9a63242 |
821 | } |
822 | |
823 | /* |
35a4481c |
824 | =for apidoc A|U8 *|bytes_to_utf8|const U8 *s|STRLEN *len |
6940069f |
825 | |
1e54db1a |
826 | Converts a string C<s> of length C<len> from ASCII into UTF-8 encoding. |
6662521e |
827 | Returns a pointer to the newly-created string, and sets C<len> to |
828 | reflect the new length. |
6940069f |
829 | |
1e54db1a |
830 | If you want to convert to UTF-8 from other encodings than ASCII, |
c9ada85f |
831 | see sv_recode_to_utf8(). |
832 | |
497711e7 |
833 | =cut |
6940069f |
834 | */ |
835 | |
836 | U8* |
35a4481c |
837 | Perl_bytes_to_utf8(pTHX_ const U8 *s, STRLEN *len) |
6940069f |
838 | { |
35a4481c |
839 | const U8 * const send = s + (*len); |
6940069f |
840 | U8 *d; |
841 | U8 *dst; |
6940069f |
842 | |
6662521e |
843 | Newz(801, d, (*len) * 2 + 1, U8); |
6940069f |
844 | dst = d; |
845 | |
846 | while (s < send) { |
35a4481c |
847 | const 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 |
851 | *d++ = (U8)UTF8_EIGHT_BIT_HI(uv); |
852 | *d++ = (U8)UTF8_EIGHT_BIT_LO(uv); |
6940069f |
853 | } |
854 | } |
855 | *d = '\0'; |
6662521e |
856 | *len = d-dst; |
6940069f |
857 | return dst; |
858 | } |
859 | |
a0ed51b3 |
860 | /* |
dea0fc0b |
861 | * Convert native (big-endian) or reversed (little-endian) UTF-16 to UTF-8. |
a0ed51b3 |
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 |
869 | U8* pend; |
870 | U8* dstart = d; |
871 | |
1de9afcd |
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 |
880 | |
881 | pend = p + bytelen; |
882 | |
a0ed51b3 |
883 | while (p < pend) { |
dea0fc0b |
884 | UV uv = (p[0] << 8) + p[1]; /* UTF-16BE */ |
885 | p += 2; |
a0ed51b3 |
886 | if (uv < 0x80) { |
eb160463 |
887 | *d++ = (U8)uv; |
a0ed51b3 |
888 | continue; |
889 | } |
890 | if (uv < 0x800) { |
eb160463 |
891 | *d++ = (U8)(( uv >> 6) | 0xc0); |
892 | *d++ = (U8)(( uv & 0x3f) | 0x80); |
a0ed51b3 |
893 | continue; |
894 | } |
895 | if (uv >= 0xd800 && uv < 0xdbff) { /* surrogates */ |
30f84f9e |
896 | UV low = (p[0] << 8) + p[1]; |
897 | p += 2; |
dea0fc0b |
898 | if (low < 0xdc00 || low >= 0xdfff) |
899 | Perl_croak(aTHX_ "Malformed UTF-16 surrogate"); |
a0ed51b3 |
900 | uv = ((uv - 0xd800) << 10) + (low - 0xdc00) + 0x10000; |
901 | } |
902 | if (uv < 0x10000) { |
eb160463 |
903 | *d++ = (U8)(( uv >> 12) | 0xe0); |
904 | *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80); |
905 | *d++ = (U8)(( uv & 0x3f) | 0x80); |
a0ed51b3 |
906 | continue; |
907 | } |
908 | else { |
eb160463 |
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 |
913 | continue; |
914 | } |
915 | } |
dea0fc0b |
916 | *newlen = d - dstart; |
a0ed51b3 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
1047 | return is_utf8_xdigit(tmpbuf); |
1048 | } |
1049 | |
84afefe6 |
1050 | UV |
1051 | Perl_to_uni_upper(pTHX_ UV c, U8* p, STRLEN *lenp) |
a0ed51b3 |
1052 | { |
0ebc6274 |
1053 | uvchr_to_utf8(p, c); |
1054 | return to_utf8_upper(p, p, lenp); |
a0ed51b3 |
1055 | } |
1056 | |
84afefe6 |
1057 | UV |
1058 | Perl_to_uni_title(pTHX_ UV c, U8* p, STRLEN *lenp) |
a0ed51b3 |
1059 | { |
0ebc6274 |
1060 | uvchr_to_utf8(p, c); |
1061 | return to_utf8_title(p, p, lenp); |
a0ed51b3 |
1062 | } |
1063 | |
84afefe6 |
1064 | UV |
1065 | Perl_to_uni_lower(pTHX_ UV c, U8* p, STRLEN *lenp) |
a0ed51b3 |
1066 | { |
0ebc6274 |
1067 | uvchr_to_utf8(p, c); |
1068 | return to_utf8_lower(p, p, lenp); |
a0ed51b3 |
1069 | } |
1070 | |
84afefe6 |
1071 | UV |
1072 | Perl_to_uni_fold(pTHX_ UV c, U8* p, STRLEN *lenp) |
1073 | { |
0ebc6274 |
1074 | uvchr_to_utf8(p, c); |
1075 | return to_utf8_fold(p, p, lenp); |
84afefe6 |
1076 | } |
1077 | |
a0ed51b3 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
1160 | { |
1161 | return is_uni_xdigit(c); /* XXX no locale support yet */ |
1162 | } |
1163 | |
b7ac61fa |
1164 | U32 |
1165 | Perl_to_uni_upper_lc(pTHX_ U32 c) |
1166 | { |
ee099d14 |
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 |
1172 | } |
1173 | |
1174 | U32 |
1175 | Perl_to_uni_title_lc(pTHX_ U32 c) |
1176 | { |
ee099d14 |
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 |
1182 | } |
1183 | |
1184 | U32 |
1185 | Perl_to_uni_lower_lc(pTHX_ U32 c) |
1186 | { |
ee099d14 |
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 |
1192 | } |
1193 | |
a0ed51b3 |
1194 | bool |
7fc63493 |
1195 | Perl_is_utf8_alnum(pTHX_ const U8 *p) |
a0ed51b3 |
1196 | { |
386d01d6 |
1197 | if (!is_utf8_char(p)) |
1198 | return FALSE; |
a0ed51b3 |
1199 | if (!PL_utf8_alnum) |
289d4f09 |
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 |
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 |
1211 | #endif |
1212 | } |
1213 | |
1214 | bool |
7fc63493 |
1215 | Perl_is_utf8_alnumc(pTHX_ const U8 *p) |
b8c5462f |
1216 | { |
386d01d6 |
1217 | if (!is_utf8_char(p)) |
1218 | return FALSE; |
b8c5462f |
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 |
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 |
1228 | #endif |
1229 | } |
1230 | |
1231 | bool |
7fc63493 |
1232 | Perl_is_utf8_idfirst(pTHX_ const U8 *p) /* The naming is historical. */ |
a0ed51b3 |
1233 | { |
82686b01 |
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 |
1241 | } |
1242 | |
1243 | bool |
7fc63493 |
1244 | Perl_is_utf8_idcont(pTHX_ const U8 *p) |
82686b01 |
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 |
1253 | } |
1254 | |
1255 | bool |
7fc63493 |
1256 | Perl_is_utf8_alpha(pTHX_ const U8 *p) |
a0ed51b3 |
1257 | { |
386d01d6 |
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 |
1263 | } |
1264 | |
1265 | bool |
7fc63493 |
1266 | Perl_is_utf8_ascii(pTHX_ const U8 *p) |
b8c5462f |
1267 | { |
386d01d6 |
1268 | if (!is_utf8_char(p)) |
1269 | return FALSE; |
b8c5462f |
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 |
1273 | } |
1274 | |
1275 | bool |
7fc63493 |
1276 | Perl_is_utf8_space(pTHX_ const U8 *p) |
a0ed51b3 |
1277 | { |
386d01d6 |
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 |
1283 | } |
1284 | |
1285 | bool |
7fc63493 |
1286 | Perl_is_utf8_digit(pTHX_ const U8 *p) |
a0ed51b3 |
1287 | { |
386d01d6 |
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 |
1293 | } |
1294 | |
1295 | bool |
7fc63493 |
1296 | Perl_is_utf8_upper(pTHX_ const U8 *p) |
a0ed51b3 |
1297 | { |
386d01d6 |
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 |
1303 | } |
1304 | |
1305 | bool |
7fc63493 |
1306 | Perl_is_utf8_lower(pTHX_ const U8 *p) |
a0ed51b3 |
1307 | { |
386d01d6 |
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 |
1313 | } |
1314 | |
1315 | bool |
7fc63493 |
1316 | Perl_is_utf8_cntrl(pTHX_ const U8 *p) |
b8c5462f |
1317 | { |
386d01d6 |
1318 | if (!is_utf8_char(p)) |
1319 | return FALSE; |
b8c5462f |
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 |
1323 | } |
1324 | |
1325 | bool |
7fc63493 |
1326 | Perl_is_utf8_graph(pTHX_ const U8 *p) |
b8c5462f |
1327 | { |
386d01d6 |
1328 | if (!is_utf8_char(p)) |
1329 | return FALSE; |
b8c5462f |
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 |
1333 | } |
1334 | |
1335 | bool |
7fc63493 |
1336 | Perl_is_utf8_print(pTHX_ const U8 *p) |
a0ed51b3 |
1337 | { |
386d01d6 |
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 |
1343 | } |
1344 | |
1345 | bool |
7fc63493 |
1346 | Perl_is_utf8_punct(pTHX_ const U8 *p) |
b8c5462f |
1347 | { |
386d01d6 |
1348 | if (!is_utf8_char(p)) |
1349 | return FALSE; |
b8c5462f |
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 |
1353 | } |
1354 | |
1355 | bool |
7fc63493 |
1356 | Perl_is_utf8_xdigit(pTHX_ const U8 *p) |
b8c5462f |
1357 | { |
386d01d6 |
1358 | if (!is_utf8_char(p)) |
1359 | return FALSE; |
b8c5462f |
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 |
1363 | } |
1364 | |
1365 | bool |
7fc63493 |
1366 | Perl_is_utf8_mark(pTHX_ const U8 *p) |
a0ed51b3 |
1367 | { |
386d01d6 |
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 |
1373 | } |
1374 | |
6b5c0936 |
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 |
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 |
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 |
1395 | The "normal" is a string like "ToLower" which means the swash |
1396 | %utf8::ToLower. |
1397 | |
1398 | =cut */ |
6b5c0936 |
1399 | |
2104c8d9 |
1400 | UV |
e1ec3a88 |
1401 | Perl_to_utf8_case(pTHX_ const U8 *p, U8* ustrp, STRLEN *lenp, SV **swashp, const char *normal, const char *special) |
a0ed51b3 |
1402 | { |
7fc63493 |
1403 | UV uv1; |
89ebb4a3 |
1404 | U8 tmpbuf[UTF8_MAXBYTES_CASE+1]; |
0134edef |
1405 | STRLEN len = 0; |
a0ed51b3 |
1406 | |
7fc63493 |
1407 | const UV uv0 = utf8_to_uvchr(p, 0); |
1feea2c7 |
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 |
1413 | |
1414 | if (!*swashp) /* load on-demand */ |
1415 | *swashp = swash_init("utf8", normal, &PL_sv_undef, 4, 0); |
1416 | |
b08cf34e |
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 |
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 |
1430 | if (len == 1) |
1431 | len = uvuni_to_utf8(ustrp, NATIVE_TO_UNI(*(U8*)s)) - ustrp; |
2a37f04d |
1432 | else { |
2f9475ad |
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 |
1438 | |
1439 | d = tmpbuf; |
b08cf34e |
1440 | if (SvUTF8(*svp)) { |
2f9475ad |
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 |
1454 | while (t < tend) { |
1455 | d = uvchr_to_utf8(d, UNI_TO_NATIVE(*t)); |
1456 | t++; |
1457 | } |
2f9475ad |
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 |
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 |
1476 | } |
1477 | } |
1feea2c7 |
1478 | |
0134edef |
1479 | if (!len) /* Neither: just copy. */ |
1480 | len = uvchr_to_utf8(ustrp, uv0) - ustrp; |
1481 | |
2a37f04d |
1482 | if (lenp) |
1483 | *lenp = len; |
1484 | |
0134edef |
1485 | return len ? utf8_to_uvchr(ustrp, 0) : 0; |
a0ed51b3 |
1486 | } |
1487 | |
d3e79532 |
1488 | /* |
7fc63493 |
1489 | =for apidoc A|UV|to_utf8_upper|const U8 *p|U8 *ustrp|STRLEN *lenp |
d3e79532 |
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 |
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 |
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 |
7fc63493 |
1502 | Perl_to_utf8_upper(pTHX_ const 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 |
1508 | /* |
7fc63493 |
1509 | =for apidoc A|UV|to_utf8_title|const U8 *p|U8 *ustrp|STRLEN *lenp |
d3e79532 |
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 |
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 |
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 |
1521 | UV |
7fc63493 |
1522 | Perl_to_utf8_title(pTHX_ const U8 *p, U8* ustrp, STRLEN *lenp) |
983ffd37 |
1523 | { |
1524 | return Perl_to_utf8_case(aTHX_ p, ustrp, lenp, |
b4e400f9 |
1525 | &PL_utf8_totitle, "ToTitle", "utf8::ToSpecTitle"); |
a0ed51b3 |
1526 | } |
1527 | |
d3e79532 |
1528 | /* |
7fc63493 |
1529 | =for apidoc A|UV|to_utf8_lower|const U8 *p|U8 *ustrp|STRLEN *lenp |
d3e79532 |
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 |
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 |
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 |
7fc63493 |
1542 | Perl_to_utf8_lower(pTHX_ const U8 *p, U8* ustrp, STRLEN *lenp) |
a0ed51b3 |
1543 | { |
983ffd37 |
1544 | return Perl_to_utf8_case(aTHX_ p, ustrp, lenp, |
b4e400f9 |
1545 | &PL_utf8_tolower, "ToLower", "utf8::ToSpecLower"); |
1546 | } |
1547 | |
d3e79532 |
1548 | /* |
7fc63493 |
1549 | =for apidoc A|UV|to_utf8_fold|const U8 *p|U8 *ustrp|STRLEN *lenp |
d3e79532 |
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 |
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 |
1562 | UV |
7fc63493 |
1563 | Perl_to_utf8_fold(pTHX_ const U8 *p, U8* ustrp, STRLEN *lenp) |
b4e400f9 |
1564 | { |
1565 | return Perl_to_utf8_case(aTHX_ p, ustrp, lenp, |
1566 | &PL_utf8_tofold, "ToFold", "utf8::ToSpecFold"); |
a0ed51b3 |
1567 | } |
1568 | |
1569 | /* a "swash" is a swatch hash */ |
1570 | |
1571 | SV* |
7fc63493 |
1572 | Perl_swash_init(pTHX_ const char* pkg, const char* name, SV *listsv, I32 minbits, I32 none) |
a0ed51b3 |
1573 | { |
27da23d5 |
1574 | dVAR; |
a0ed51b3 |
1575 | SV* retval; |
243b1711 |
1576 | SV* tokenbufsv = sv_newmortal(); |
8e84507e |
1577 | dSP; |
7fc63493 |
1578 | const size_t pkg_len = strlen(pkg); |
1579 | const size_t name_len = strlen(name); |
71bed85a |
1580 | HV *stash = gv_stashpvn(pkg, pkg_len, FALSE); |
f8be5cf0 |
1581 | SV* errsv_save; |
ce3b816e |
1582 | |
96ca9f55 |
1583 | PUSHSTACKi(PERLSI_MAGIC); |
1584 | ENTER; |
1585 | SAVEI32(PL_hints); |
1586 | PL_hints = 0; |
1587 | save_re_context(); |
1b026014 |
1588 | if (!gv_fetchmeth(stash, "SWASHNEW", 8, -1)) { /* demand load utf8 */ |
ce3b816e |
1589 | ENTER; |
f8be5cf0 |
1590 | errsv_save = newSVsv(ERRSV); |
71bed85a |
1591 | Perl_load_module(aTHX_ PERL_LOADMOD_NOIMPORT, newSVpvn(pkg,pkg_len), |
1592 | Nullsv); |
f8be5cf0 |
1593 | if (!SvTRUE(ERRSV)) |
1594 | sv_setsv(ERRSV, errsv_save); |
1595 | SvREFCNT_dec(errsv_save); |
ce3b816e |
1596 | LEAVE; |
1597 | } |
1598 | SPAGAIN; |
a0ed51b3 |
1599 | PUSHMARK(SP); |
1600 | EXTEND(SP,5); |
71bed85a |
1601 | PUSHs(sv_2mortal(newSVpvn(pkg, pkg_len))); |
1602 | PUSHs(sv_2mortal(newSVpvn(name, name_len))); |
a0ed51b3 |
1603 | PUSHs(listsv); |
1604 | PUSHs(sv_2mortal(newSViv(minbits))); |
1605 | PUSHs(sv_2mortal(newSViv(none))); |
1606 | PUTBACK; |
923e4eb5 |
1607 | if (IN_PERL_COMPILETIME) { |
bf1fed83 |
1608 | /* XXX ought to be handled by lex_start */ |
82686b01 |
1609 | SAVEI32(PL_in_my); |
2b4bd638 |
1610 | PL_in_my = 0; |
bf1fed83 |
1611 | sv_setpv(tokenbufsv, PL_tokenbuf); |
82686b01 |
1612 | } |
f8be5cf0 |
1613 | errsv_save = newSVsv(ERRSV); |
864dbfa3 |
1614 | if (call_method("SWASHNEW", G_SCALAR)) |
8e84507e |
1615 | retval = newSVsv(*PL_stack_sp--); |
a0ed51b3 |
1616 | else |
e24b16f9 |
1617 | retval = &PL_sv_undef; |
f8be5cf0 |
1618 | if (!SvTRUE(ERRSV)) |
1619 | sv_setsv(ERRSV, errsv_save); |
1620 | SvREFCNT_dec(errsv_save); |
a0ed51b3 |
1621 | LEAVE; |
1622 | POPSTACK; |
923e4eb5 |
1623 | if (IN_PERL_COMPILETIME) { |
bf1fed83 |
1624 | STRLEN len; |
73d840c0 |
1625 | const char* pv = SvPV(tokenbufsv, len); |
bf1fed83 |
1626 | |
1627 | Copy(pv, PL_tokenbuf, len+1, char); |
eb160463 |
1628 | PL_curcop->op_private = (U8)(PL_hints & HINT_PRIVATE_MASK); |
a0ed51b3 |
1629 | } |
bc45ce41 |
1630 | if (!SvROK(retval) || SvTYPE(SvRV(retval)) != SVt_PVHV) { |
1631 | if (SvPOK(retval)) |
35c1215d |
1632 | Perl_croak(aTHX_ "Can't find Unicode property definition \"%"SVf"\"", |
1633 | retval); |
cea2e8a9 |
1634 | Perl_croak(aTHX_ "SWASHNEW didn't return an HV ref"); |
bc45ce41 |
1635 | } |
a0ed51b3 |
1636 | return retval; |
1637 | } |
1638 | |
035d37be |
1639 | |
1640 | /* This API is wrong for special case conversions since we may need to |
1641 | * return several Unicode characters for a single Unicode character |
1642 | * (see lib/unicore/SpecCase.txt) The SWASHGET in lib/utf8_heavy.pl is |
1643 | * the lower-level routine, and it is similarly broken for returning |
1644 | * multiple values. --jhi */ |
a0ed51b3 |
1645 | UV |
7fc63493 |
1646 | Perl_swash_fetch(pTHX_ SV *sv, const U8 *ptr, bool do_utf8) |
a0ed51b3 |
1647 | { |
27da23d5 |
1648 | dVAR; |
a0ed51b3 |
1649 | HV* hv = (HV*)SvRV(sv); |
3568d838 |
1650 | U32 klen; |
1651 | U32 off; |
a0ed51b3 |
1652 | STRLEN slen; |
7d85a32c |
1653 | STRLEN needents; |
4ea42e7f |
1654 | U8 *tmps = NULL; |
a0ed51b3 |
1655 | U32 bit; |
1656 | SV *retval; |
3568d838 |
1657 | U8 tmputf8[2]; |
1658 | UV c = NATIVE_TO_ASCII(*ptr); |
1659 | |
1660 | if (!do_utf8 && !UNI_IS_INVARIANT(c)) { |
eb160463 |
1661 | tmputf8[0] = (U8)UTF8_EIGHT_BIT_HI(c); |
1662 | tmputf8[1] = (U8)UTF8_EIGHT_BIT_LO(c); |
3568d838 |
1663 | ptr = tmputf8; |
1664 | } |
1665 | /* Given a UTF-X encoded char 0xAA..0xYY,0xZZ |
1666 | * then the "swatch" is a vec() for al the chars which start |
1667 | * with 0xAA..0xYY |
1668 | * So the key in the hash (klen) is length of encoded char -1 |
1669 | */ |
1670 | klen = UTF8SKIP(ptr) - 1; |
1671 | off = ptr[klen]; |
a0ed51b3 |
1672 | |
7d85a32c |
1673 | if (klen == 0) |
1674 | { |
1675 | /* If char in invariant then swatch is for all the invariant chars |
1e54db1a |
1676 | * In both UTF-8 and UTF-8-MOD that happens to be UTF_CONTINUATION_MARK |
7d85a32c |
1677 | */ |
1678 | needents = UTF_CONTINUATION_MARK; |
1679 | off = NATIVE_TO_UTF(ptr[klen]); |
1680 | } |
1681 | else |
1682 | { |
1683 | /* If char is encoded then swatch is for the prefix */ |
1684 | needents = (1 << UTF_ACCUMULATION_SHIFT); |
1685 | off = NATIVE_TO_UTF(ptr[klen]) & UTF_CONTINUATION_MASK; |
1686 | } |
1687 | |
a0ed51b3 |
1688 | /* |
1689 | * This single-entry cache saves about 1/3 of the utf8 overhead in test |
1690 | * suite. (That is, only 7-8% overall over just a hash cache. Still, |
1691 | * it's nothing to sniff at.) Pity we usually come through at least |
1692 | * two function calls to get here... |
1693 | * |
1694 | * NB: this code assumes that swatches are never modified, once generated! |
1695 | */ |
1696 | |
3568d838 |
1697 | if (hv == PL_last_swash_hv && |
a0ed51b3 |
1698 | klen == PL_last_swash_klen && |
27da23d5 |
1699 | (!klen || memEQ((char *)ptr, (char *)PL_last_swash_key, klen)) ) |
a0ed51b3 |
1700 | { |
1701 | tmps = PL_last_swash_tmps; |
1702 | slen = PL_last_swash_slen; |
1703 | } |
1704 | else { |
1705 | /* Try our second-level swatch cache, kept in a hash. */ |
e1ec3a88 |
1706 | SV** svp = hv_fetch(hv, (const char*)ptr, klen, FALSE); |
a0ed51b3 |
1707 | |
1708 | /* If not cached, generate it via utf8::SWASHGET */ |
dfe13c55 |
1709 | if (!svp || !SvPOK(*svp) || !(tmps = (U8*)SvPV(*svp, slen))) { |
a0ed51b3 |
1710 | dSP; |
2b9d42f0 |
1711 | /* We use utf8n_to_uvuni() as we want an index into |
1712 | Unicode tables, not a native character number. |
1713 | */ |
89ebb4a3 |
1714 | UV code_point = utf8n_to_uvuni(ptr, UTF8_MAXBYTES, 0, |
872c91ae |
1715 | ckWARN(WARN_UTF8) ? |
1716 | 0 : UTF8_ALLOW_ANY); |
f8be5cf0 |
1717 | SV *errsv_save; |
a0ed51b3 |
1718 | ENTER; |
1719 | SAVETMPS; |
1720 | save_re_context(); |
1721 | PUSHSTACKi(PERLSI_MAGIC); |
1722 | PUSHMARK(SP); |
1723 | EXTEND(SP,3); |
1724 | PUSHs((SV*)sv); |
ffbc6a93 |
1725 | /* On EBCDIC & ~(0xA0-1) isn't a useful thing to do */ |
3568d838 |
1726 | PUSHs(sv_2mortal(newSViv((klen) ? |
1727 | (code_point & ~(needents - 1)) : 0))); |
a0ed51b3 |
1728 | PUSHs(sv_2mortal(newSViv(needents))); |
1729 | PUTBACK; |
f8be5cf0 |
1730 | errsv_save = newSVsv(ERRSV); |
864dbfa3 |
1731 | if (call_method("SWASHGET", G_SCALAR)) |
8e84507e |
1732 | retval = newSVsv(*PL_stack_sp--); |
a0ed51b3 |
1733 | else |
e24b16f9 |
1734 | retval = &PL_sv_undef; |
f8be5cf0 |
1735 | if (!SvTRUE(ERRSV)) |
1736 | sv_setsv(ERRSV, errsv_save); |
1737 | SvREFCNT_dec(errsv_save); |
a0ed51b3 |
1738 | POPSTACK; |
1739 | FREETMPS; |
1740 | LEAVE; |
923e4eb5 |
1741 | if (IN_PERL_COMPILETIME) |
eb160463 |
1742 | PL_curcop->op_private = (U8)(PL_hints & HINT_PRIVATE_MASK); |
a0ed51b3 |
1743 | |
e1ec3a88 |
1744 | svp = hv_store(hv, (const char *)ptr, klen, retval, 0); |
a0ed51b3 |
1745 | |
7d85a32c |
1746 | if (!svp || !(tmps = (U8*)SvPV(*svp, slen)) || (slen << 3) < needents) |
cea2e8a9 |
1747 | Perl_croak(aTHX_ "SWASHGET didn't return result of proper length"); |
a0ed51b3 |
1748 | } |
1749 | |
1750 | PL_last_swash_hv = hv; |
1751 | PL_last_swash_klen = klen; |
1752 | PL_last_swash_tmps = tmps; |
1753 | PL_last_swash_slen = slen; |
1754 | if (klen) |
1755 | Copy(ptr, PL_last_swash_key, klen, U8); |
1756 | } |
1757 | |
9faf8d75 |
1758 | switch ((int)((slen << 3) / needents)) { |
a0ed51b3 |
1759 | case 1: |
1760 | bit = 1 << (off & 7); |
1761 | off >>= 3; |
1762 | return (tmps[off] & bit) != 0; |
1763 | case 8: |
1764 | return tmps[off]; |
1765 | case 16: |
1766 | off <<= 1; |
1767 | return (tmps[off] << 8) + tmps[off + 1] ; |
1768 | case 32: |
1769 | off <<= 2; |
1770 | return (tmps[off] << 24) + (tmps[off+1] << 16) + (tmps[off+2] << 8) + tmps[off + 3] ; |
1771 | } |
cea2e8a9 |
1772 | Perl_croak(aTHX_ "panic: swash_fetch"); |
a0ed51b3 |
1773 | return 0; |
1774 | } |
2b9d42f0 |
1775 | |
1776 | |
1777 | /* |
37607a96 |
1778 | =for apidoc A|U8 *|uvchr_to_utf8|U8 *d|UV uv |
2b9d42f0 |
1779 | |
1e54db1a |
1780 | Adds the UTF-8 representation of the Native codepoint C<uv> to the end |
89ebb4a3 |
1781 | of the string C<d>; C<d> should be have at least C<UTF8_MAXBYTES+1> free |
2b9d42f0 |
1782 | bytes available. The return value is the pointer to the byte after the |
1783 | end of the new character. In other words, |
1784 | |
1785 | d = uvchr_to_utf8(d, uv); |
1786 | |
1787 | is the recommended wide native character-aware way of saying |
1788 | |
1789 | *(d++) = uv; |
1790 | |
1791 | =cut |
1792 | */ |
1793 | |
1794 | /* On ASCII machines this is normally a macro but we want a |
1795 | real function in case XS code wants it |
1796 | */ |
1797 | #undef Perl_uvchr_to_utf8 |
1798 | U8 * |
1799 | Perl_uvchr_to_utf8(pTHX_ U8 *d, UV uv) |
1800 | { |
b851fbc1 |
1801 | return Perl_uvuni_to_utf8_flags(aTHX_ d, NATIVE_TO_UNI(uv), 0); |
2b9d42f0 |
1802 | } |
1803 | |
b851fbc1 |
1804 | U8 * |
1805 | Perl_uvchr_to_utf8_flags(pTHX_ U8 *d, UV uv, UV flags) |
1806 | { |
1807 | return Perl_uvuni_to_utf8_flags(aTHX_ d, NATIVE_TO_UNI(uv), flags); |
1808 | } |
2b9d42f0 |
1809 | |
1810 | /* |
37607a96 |
1811 | =for apidoc A|UV|utf8n_to_uvchr|U8 *s|STRLEN curlen|STRLEN *retlen|U32 flags |
2b9d42f0 |
1812 | |
1813 | Returns the native character value of the first character in the string C<s> |
1e54db1a |
1814 | which is assumed to be in UTF-8 encoding; C<retlen> will be set to the |
2b9d42f0 |
1815 | length, in bytes, of that character. |
1816 | |
1817 | Allows length and flags to be passed to low level routine. |
1818 | |
1819 | =cut |
1820 | */ |
0a2ef054 |
1821 | /* On ASCII machines this is normally a macro but we want |
1822 | a real function in case XS code wants it |
2b9d42f0 |
1823 | */ |
1824 | #undef Perl_utf8n_to_uvchr |
1825 | UV |
7fc63493 |
1826 | Perl_utf8n_to_uvchr(pTHX_ const U8 *s, STRLEN curlen, STRLEN *retlen, U32 flags) |
2b9d42f0 |
1827 | { |
1828 | UV uv = Perl_utf8n_to_uvuni(aTHX_ s, curlen, retlen, flags); |
1829 | return UNI_TO_NATIVE(uv); |
1830 | } |
1831 | |
d2cc3551 |
1832 | /* |
1833 | =for apidoc A|char *|pv_uni_display|SV *dsv|U8 *spv|STRLEN len|STRLEN pvlim|UV flags |
1834 | |
1835 | Build to the scalar dsv a displayable version of the string spv, |
1836 | length len, the displayable version being at most pvlim bytes long |
1837 | (if longer, the rest is truncated and "..." will be appended). |
0a2ef054 |
1838 | |
9e55ce06 |
1839 | The flags argument can have UNI_DISPLAY_ISPRINT set to display |
00e86452 |
1840 | isPRINT()able characters as themselves, UNI_DISPLAY_BACKSLASH |
0a2ef054 |
1841 | to display the \\[nrfta\\] as the backslashed versions (like '\n') |
1842 | (UNI_DISPLAY_BACKSLASH is preferred over UNI_DISPLAY_ISPRINT for \\). |
1843 | UNI_DISPLAY_QQ (and its alias UNI_DISPLAY_REGEX) have both |
1844 | UNI_DISPLAY_BACKSLASH and UNI_DISPLAY_ISPRINT turned on. |
1845 | |
d2cc3551 |
1846 | The pointer to the PV of the dsv is returned. |
1847 | |
1848 | =cut */ |
e6b2e755 |
1849 | char * |
e1ec3a88 |
1850 | Perl_pv_uni_display(pTHX_ SV *dsv, const U8 *spv, STRLEN len, STRLEN pvlim, UV flags) |
e6b2e755 |
1851 | { |
1852 | int truncated = 0; |
e1ec3a88 |
1853 | const char *s, *e; |
e6b2e755 |
1854 | |
1855 | sv_setpvn(dsv, "", 0); |
e1ec3a88 |
1856 | for (s = (const char *)spv, e = s + len; s < e; s += UTF8SKIP(s)) { |
e6b2e755 |
1857 | UV u; |
a49f32c6 |
1858 | /* This serves double duty as a flag and a character to print after |
1859 | a \ when flags & UNI_DISPLAY_BACKSLASH is true. |
1860 | */ |
1861 | char ok = 0; |
c728cb41 |
1862 | |
e6b2e755 |
1863 | if (pvlim && SvCUR(dsv) >= pvlim) { |
1864 | truncated++; |
1865 | break; |
1866 | } |
1867 | u = utf8_to_uvchr((U8*)s, 0); |
c728cb41 |
1868 | if (u < 256) { |
fe2774ed |
1869 | unsigned char c = (unsigned char)u & 0xFF; |
c728cb41 |
1870 | if (!ok && (flags & UNI_DISPLAY_BACKSLASH)) { |
a49f32c6 |
1871 | switch (c) { |
c728cb41 |
1872 | case '\n': |
a49f32c6 |
1873 | ok = 'n'; break; |
c728cb41 |
1874 | case '\r': |
a49f32c6 |
1875 | ok = 'r'; break; |
c728cb41 |
1876 | case '\t': |
a49f32c6 |
1877 | ok = 't'; break; |
c728cb41 |
1878 | case '\f': |
a49f32c6 |
1879 | ok = 'f'; break; |
c728cb41 |
1880 | case '\a': |
a49f32c6 |
1881 | ok = 'a'; break; |
c728cb41 |
1882 | case '\\': |
a49f32c6 |
1883 | ok = '\\'; break; |
c728cb41 |
1884 | default: break; |
1885 | } |
a49f32c6 |
1886 | if (ok) { |
1887 | Perl_sv_catpvf(aTHX_ dsv, "\\%c", ok); |
1888 | } |
c728cb41 |
1889 | } |
00e86452 |
1890 | /* isPRINT() is the locale-blind version. */ |
a49f32c6 |
1891 | if (!ok && (flags & UNI_DISPLAY_ISPRINT) && isPRINT(c)) { |
1892 | Perl_sv_catpvf(aTHX_ dsv, "%c", c); |
1893 | ok = 1; |
0a2ef054 |
1894 | } |
c728cb41 |
1895 | } |
1896 | if (!ok) |
9e55ce06 |
1897 | Perl_sv_catpvf(aTHX_ dsv, "\\x{%"UVxf"}", u); |
e6b2e755 |
1898 | } |
1899 | if (truncated) |
1900 | sv_catpvn(dsv, "...", 3); |
1901 | |
1902 | return SvPVX(dsv); |
1903 | } |
2b9d42f0 |
1904 | |
d2cc3551 |
1905 | /* |
1906 | =for apidoc A|char *|sv_uni_display|SV *dsv|SV *ssv|STRLEN pvlim|UV flags |
1907 | |
1908 | Build to the scalar dsv a displayable version of the scalar sv, |
0a2ef054 |
1909 | the displayable version being at most pvlim bytes long |
d2cc3551 |
1910 | (if longer, the rest is truncated and "..." will be appended). |
0a2ef054 |
1911 | |
1912 | The flags argument is as in pv_uni_display(). |
1913 | |
d2cc3551 |
1914 | The pointer to the PV of the dsv is returned. |
1915 | |
1916 | =cut */ |
e6b2e755 |
1917 | char * |
1918 | Perl_sv_uni_display(pTHX_ SV *dsv, SV *ssv, STRLEN pvlim, UV flags) |
1919 | { |
701a277b |
1920 | return Perl_pv_uni_display(aTHX_ dsv, (U8*)SvPVX(ssv), SvCUR(ssv), |
1921 | pvlim, flags); |
1922 | } |
1923 | |
d2cc3551 |
1924 | /* |
d07ddd77 |
1925 | =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 |
1926 | |
1927 | Return true if the strings s1 and s2 differ case-insensitively, false |
1928 | if not (if they are equal case-insensitively). If u1 is true, the |
1929 | string s1 is assumed to be in UTF-8-encoded Unicode. If u2 is true, |
d07ddd77 |
1930 | the string s2 is assumed to be in UTF-8-encoded Unicode. If u1 or u2 |
1931 | are false, the respective string is assumed to be in native 8-bit |
1932 | encoding. |
1933 | |
1934 | If the pe1 and pe2 are non-NULL, the scanning pointers will be copied |
1935 | in there (they will point at the beginning of the I<next> character). |
1936 | If the pointers behind pe1 or pe2 are non-NULL, they are the end |
1937 | pointers beyond which scanning will not continue under any |
4cdaeff7 |
1938 | circumstances. If the byte lengths l1 and l2 are non-zero, s1+l1 and |
d07ddd77 |
1939 | s2+l2 will be used as goal end pointers that will also stop the scan, |
1940 | and which qualify towards defining a successful match: all the scans |
1941 | that define an explicit length must reach their goal pointers for |
1942 | a match to succeed). |
d2cc3551 |
1943 | |
1944 | For case-insensitiveness, the "casefolding" of Unicode is used |
1945 | instead of upper/lowercasing both the characters, see |
1946 | http://www.unicode.org/unicode/reports/tr21/ (Case Mappings). |
1947 | |
1948 | =cut */ |
701a277b |
1949 | I32 |
d07ddd77 |
1950 | 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 |
1951 | { |
e1ec3a88 |
1952 | register const U8 *p1 = (const U8*)s1; |
1953 | register const U8 *p2 = (const U8*)s2; |
1954 | register const U8 *f1 = 0, *f2 = 0; |
1955 | register U8 *e1 = 0, *q1 = 0; |
1956 | register U8 *e2 = 0, *q2 = 0; |
d07ddd77 |
1957 | STRLEN n1 = 0, n2 = 0; |
89ebb4a3 |
1958 | U8 foldbuf1[UTF8_MAXBYTES_CASE+1]; |
1959 | U8 foldbuf2[UTF8_MAXBYTES_CASE+1]; |
d7f013c8 |
1960 | U8 natbuf[1+1]; |
1961 | STRLEN foldlen1, foldlen2; |
d07ddd77 |
1962 | bool match; |
332ddc25 |
1963 | |
d07ddd77 |
1964 | if (pe1) |
1965 | e1 = *(U8**)pe1; |
e1ec3a88 |
1966 | if (e1 == 0 || (l1 && l1 < (UV)(e1 - (const U8*)s1))) |
1967 | f1 = (const U8*)s1 + l1; |
d07ddd77 |
1968 | if (pe2) |
1969 | e2 = *(U8**)pe2; |
e1ec3a88 |
1970 | if (e2 == 0 || (l2 && l2 < (UV)(e2 - (const U8*)s2))) |
1971 | f2 = (const U8*)s2 + l2; |
d07ddd77 |
1972 | |
1973 | if ((e1 == 0 && f1 == 0) || (e2 == 0 && f2 == 0) || (f1 == 0 && f2 == 0)) |
1974 | return 1; /* mismatch; possible infinite loop or false positive */ |
1975 | |
a6872d42 |
1976 | if (!u1 || !u2) |
1977 | natbuf[1] = 0; /* Need to terminate the buffer. */ |
1978 | |
d07ddd77 |
1979 | while ((e1 == 0 || p1 < e1) && |
1980 | (f1 == 0 || p1 < f1) && |
1981 | (e2 == 0 || p2 < e2) && |
1982 | (f2 == 0 || p2 < f2)) { |
1983 | if (n1 == 0) { |
d7f013c8 |
1984 | if (u1) |
1985 | to_utf8_fold(p1, foldbuf1, &foldlen1); |
1986 | else { |
f5cee151 |
1987 | natbuf[0] = *p1; |
d7f013c8 |
1988 | to_utf8_fold(natbuf, foldbuf1, &foldlen1); |
1989 | } |
1990 | q1 = foldbuf1; |
d07ddd77 |
1991 | n1 = foldlen1; |
332ddc25 |
1992 | } |
d07ddd77 |
1993 | if (n2 == 0) { |
d7f013c8 |
1994 | if (u2) |
1995 | to_utf8_fold(p2, foldbuf2, &foldlen2); |
1996 | else { |
f5cee151 |
1997 | natbuf[0] = *p2; |
d7f013c8 |
1998 | to_utf8_fold(natbuf, foldbuf2, &foldlen2); |
1999 | } |
2000 | q2 = foldbuf2; |
d07ddd77 |
2001 | n2 = foldlen2; |
332ddc25 |
2002 | } |
d07ddd77 |
2003 | while (n1 && n2) { |
2004 | if ( UTF8SKIP(q1) != UTF8SKIP(q2) || |
2005 | (UTF8SKIP(q1) == 1 && *q1 != *q2) || |
2006 | memNE((char*)q1, (char*)q2, UTF8SKIP(q1)) ) |
d7f013c8 |
2007 | return 1; /* mismatch */ |
d07ddd77 |
2008 | n1 -= UTF8SKIP(q1); |
d7f013c8 |
2009 | q1 += UTF8SKIP(q1); |
d07ddd77 |
2010 | n2 -= UTF8SKIP(q2); |
d7f013c8 |
2011 | q2 += UTF8SKIP(q2); |
701a277b |
2012 | } |
d07ddd77 |
2013 | if (n1 == 0) |
d7f013c8 |
2014 | p1 += u1 ? UTF8SKIP(p1) : 1; |
d07ddd77 |
2015 | if (n2 == 0) |
d7f013c8 |
2016 | p2 += u2 ? UTF8SKIP(p2) : 1; |
2017 | |
d2cc3551 |
2018 | } |
5469e704 |
2019 | |
d07ddd77 |
2020 | /* A match is defined by all the scans that specified |
2021 | * an explicit length reaching their final goals. */ |
2022 | match = (f1 == 0 || p1 == f1) && (f2 == 0 || p2 == f2); |
5469e704 |
2023 | |
2024 | if (match) { |
d07ddd77 |
2025 | if (pe1) |
2026 | *pe1 = (char*)p1; |
2027 | if (pe2) |
2028 | *pe2 = (char*)p2; |
5469e704 |
2029 | } |
2030 | |
2031 | return match ? 0 : 1; /* 0 match, 1 mismatch */ |
e6b2e755 |
2032 | } |
701a277b |
2033 | |
a49f32c6 |
2034 | /* |
2035 | * Local variables: |
2036 | * c-indentation-style: bsd |
2037 | * c-basic-offset: 4 |
2038 | * indent-tabs-mode: t |
2039 | * End: |
2040 | * |
2041 | * vim: shiftwidth=4: |
2042 | */ |