This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
op.c:ck_glob: Don’t do gv_fetchpv("CORE::GLOBAL::glob")
[perl5.git] / utf8.c
CommitLineData
a0ed51b3
LW
1/* utf8.c
2 *
1129b882 3 * Copyright (C) 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008
b94e2f88 4 * by Larry Wall and others
a0ed51b3
LW
5 *
6 * You may distribute under the terms of either the GNU General Public
7 * License or the Artistic License, as specified in the README file.
8 *
9 */
10
11/*
4ac71550
TC
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 *
cdad3b53 16 * [p.603 of _The Lord of the Rings_, IV/I: "The Taming of Sméagol"]
a0ed51b3
LW
17 *
18 * 'Well do I understand your speech,' he answered in the same language;
19 * 'yet few strangers do so. Why then do you not speak in the Common Tongue,
4ac71550 20 * as is the custom in the West, if you wish to be answered?'
cdad3b53 21 * --Gandalf, addressing Théoden's door wardens
4ac71550
TC
22 *
23 * [p.508 of _The Lord of the Rings_, III/vi: "The King of the Golden Hall"]
a0ed51b3
LW
24 *
25 * ...the travellers perceived that the floor was paved with stones of many
26 * hues; branching runes and strange devices intertwined beneath their feet.
4ac71550
TC
27 *
28 * [p.512 of _The Lord of the Rings_, III/vi: "The King of the Golden Hall"]
a0ed51b3
LW
29 */
30
31#include "EXTERN.h"
864dbfa3 32#define PERL_IN_UTF8_C
a0ed51b3
LW
33#include "perl.h"
34
a0c21aa1 35#ifndef EBCDIC
970ea3cb 36/* Separate prototypes needed because in ASCII systems these are
a0c21aa1
JH
37 * usually macros but they still are compiled as code, too. */
38PERL_CALLCONV UV Perl_utf8n_to_uvchr(pTHX_ const U8 *s, STRLEN curlen, STRLEN *retlen, U32 flags);
39PERL_CALLCONV U8* Perl_uvchr_to_utf8(pTHX_ U8 *d, UV uv);
40#endif
41
27da23d5
JH
42static const char unees[] =
43 "Malformed UTF-8 character (unexpected end of string)";
901b21bf 44
48ef279e 45/*
ccfc67b7 46=head1 Unicode Support
a0ed51b3 47
166f8a29
DM
48This file contains various utility functions for manipulating UTF8-encoded
49strings. For the uninitiated, this is a method of representing arbitrary
61296642 50Unicode characters as a variable number of bytes, in such a way that
56da48f7
DM
51characters in the ASCII range are unmodified, and a zero byte never appears
52within non-zero characters.
166f8a29 53
eaf7a4d2
CS
54=cut
55*/
56
57/*
58=for apidoc is_ascii_string
59
a1433954 60Returns true if the first C<len> bytes of the string C<s> are the same whether
970ea3cb
KW
61or not the string is encoded in UTF-8 (or UTF-EBCDIC on EBCDIC machines). That
62is, if they are invariant. On ASCII-ish machines, only ASCII characters
63fit this definition, hence the function's name.
eaf7a4d2 64
9f7e3d64
MH
65If C<len> is 0, it will be calculated using C<strlen(s)>.
66
a1433954 67See also L</is_utf8_string>(), L</is_utf8_string_loclen>(), and L</is_utf8_string_loc>().
eaf7a4d2
CS
68
69=cut
70*/
71
72bool
668b6d8d 73Perl_is_ascii_string(const U8 *s, STRLEN len)
eaf7a4d2
CS
74{
75 const U8* const send = s + (len ? len : strlen((const char *)s));
76 const U8* x = s;
77
78 PERL_ARGS_ASSERT_IS_ASCII_STRING;
eaf7a4d2
CS
79
80 for (; x < send; ++x) {
81 if (!UTF8_IS_INVARIANT(*x))
82 break;
83 }
84
85 return x == send;
86}
87
88/*
87cea99e 89=for apidoc uvuni_to_utf8_flags
eebe1485 90
6ee84de2
KW
91Adds the UTF-8 representation of the code point C<uv> to the end
92of the string C<d>; C<d> should have at least C<UTF8_MAXBYTES+1> free
eebe1485 93bytes available. The return value is the pointer to the byte after the
9041c2e3 94end of the new character. In other words,
eebe1485 95
b851fbc1
JH
96 d = uvuni_to_utf8_flags(d, uv, flags);
97
98or, in most cases,
99
9041c2e3 100 d = uvuni_to_utf8(d, uv);
eebe1485 101
b851fbc1
JH
102(which is equivalent to)
103
104 d = uvuni_to_utf8_flags(d, uv, 0);
105
949cf498 106This is the recommended Unicode-aware way of saying
eebe1485
SC
107
108 *(d++) = uv;
109
949cf498
KW
110This function will convert to UTF-8 (and not warn) even code points that aren't
111legal Unicode or are problematic, unless C<flags> contains one or more of the
a1433954
KW
112following flags:
113
949cf498
KW
114If C<uv> is a Unicode surrogate code point and UNICODE_WARN_SURROGATE is set,
115the function will raise a warning, provided UTF8 warnings are enabled. If instead
116UNICODE_DISALLOW_SURROGATE is set, the function will fail and return NULL.
117If both flags are set, the function will both warn and return NULL.
118
119The UNICODE_WARN_NONCHAR and UNICODE_DISALLOW_NONCHAR flags correspondingly
120affect how the function handles a Unicode non-character. And, likewise for the
121UNICODE_WARN_SUPER and UNICODE_DISALLOW_SUPER flags, and code points that are
122above the Unicode maximum of 0x10FFFF. Code points above 0x7FFF_FFFF (which are
123even less portable) can be warned and/or disallowed even if other above-Unicode
124code points are accepted by the UNICODE_WARN_FE_FF and UNICODE_DISALLOW_FE_FF
125flags.
126
127And finally, the flag UNICODE_WARN_ILLEGAL_INTERCHANGE selects all four of the
128above WARN flags; and UNICODE_DISALLOW_ILLEGAL_INTERCHANGE selects all four
129DISALLOW flags.
130
131
eebe1485
SC
132=cut
133*/
134
dfe13c55 135U8 *
b851fbc1 136Perl_uvuni_to_utf8_flags(pTHX_ U8 *d, UV uv, UV flags)
a0ed51b3 137{
7918f24d
NC
138 PERL_ARGS_ASSERT_UVUNI_TO_UTF8_FLAGS;
139
a099aed4 140 if (ckWARN4_d(WARN_UTF8, WARN_SURROGATE, WARN_NON_UNICODE, WARN_NONCHAR)) {
949cf498
KW
141 if (UNICODE_IS_SURROGATE(uv)) {
142 if (flags & UNICODE_WARN_SURROGATE) {
8457b38f 143 Perl_ck_warner_d(aTHX_ packWARN(WARN_SURROGATE),
949cf498
KW
144 "UTF-16 surrogate U+%04"UVXf, uv);
145 }
146 if (flags & UNICODE_DISALLOW_SURROGATE) {
147 return NULL;
148 }
149 }
150 else if (UNICODE_IS_SUPER(uv)) {
151 if (flags & UNICODE_WARN_SUPER
152 || (UNICODE_IS_FE_FF(uv) && (flags & UNICODE_WARN_FE_FF)))
153 {
8457b38f 154 Perl_ck_warner_d(aTHX_ packWARN(WARN_NON_UNICODE),
949cf498
KW
155 "Code point 0x%04"UVXf" is not Unicode, may not be portable", uv);
156 }
157 if (flags & UNICODE_DISALLOW_SUPER
158 || (UNICODE_IS_FE_FF(uv) && (flags & UNICODE_DISALLOW_FE_FF)))
159 {
160 return NULL;
161 }
162 }
163 else if (UNICODE_IS_NONCHAR(uv)) {
164 if (flags & UNICODE_WARN_NONCHAR) {
8457b38f 165 Perl_ck_warner_d(aTHX_ packWARN(WARN_NONCHAR),
949cf498
KW
166 "Unicode non-character U+%04"UVXf" is illegal for open interchange",
167 uv);
168 }
169 if (flags & UNICODE_DISALLOW_NONCHAR) {
170 return NULL;
171 }
172 }
507b9800 173 }
c4d5f83a 174 if (UNI_IS_INVARIANT(uv)) {
eb160463 175 *d++ = (U8)UTF_TO_NATIVE(uv);
a0ed51b3
LW
176 return d;
177 }
2d331972 178#if defined(EBCDIC)
1d72bdf6
NIS
179 else {
180 STRLEN len = UNISKIP(uv);
181 U8 *p = d+len-1;
182 while (p > d) {
eb160463 183 *p-- = (U8)UTF_TO_NATIVE((uv & UTF_CONTINUATION_MASK) | UTF_CONTINUATION_MARK);
1d72bdf6
NIS
184 uv >>= UTF_ACCUMULATION_SHIFT;
185 }
eb160463 186 *p = (U8)UTF_TO_NATIVE((uv & UTF_START_MASK(len)) | UTF_START_MARK(len));
1d72bdf6
NIS
187 return d+len;
188 }
189#else /* Non loop style */
a0ed51b3 190 if (uv < 0x800) {
eb160463
GS
191 *d++ = (U8)(( uv >> 6) | 0xc0);
192 *d++ = (U8)(( uv & 0x3f) | 0x80);
a0ed51b3
LW
193 return d;
194 }
195 if (uv < 0x10000) {
eb160463
GS
196 *d++ = (U8)(( uv >> 12) | 0xe0);
197 *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
198 *d++ = (U8)(( uv & 0x3f) | 0x80);
a0ed51b3
LW
199 return d;
200 }
201 if (uv < 0x200000) {
eb160463
GS
202 *d++ = (U8)(( uv >> 18) | 0xf0);
203 *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80);
204 *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
205 *d++ = (U8)(( uv & 0x3f) | 0x80);
a0ed51b3
LW
206 return d;
207 }
208 if (uv < 0x4000000) {
eb160463
GS
209 *d++ = (U8)(( uv >> 24) | 0xf8);
210 *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80);
211 *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80);
212 *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
213 *d++ = (U8)(( uv & 0x3f) | 0x80);
a0ed51b3
LW
214 return d;
215 }
216 if (uv < 0x80000000) {
eb160463
GS
217 *d++ = (U8)(( uv >> 30) | 0xfc);
218 *d++ = (U8)(((uv >> 24) & 0x3f) | 0x80);
219 *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80);
220 *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80);
221 *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
222 *d++ = (U8)(( uv & 0x3f) | 0x80);
a0ed51b3
LW
223 return d;
224 }
6b8eaf93 225#ifdef HAS_QUAD
d7578b48 226 if (uv < UTF8_QUAD_MAX)
a0ed51b3
LW
227#endif
228 {
eb160463
GS
229 *d++ = 0xfe; /* Can't match U+FEFF! */
230 *d++ = (U8)(((uv >> 30) & 0x3f) | 0x80);
231 *d++ = (U8)(((uv >> 24) & 0x3f) | 0x80);
232 *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80);
233 *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80);
234 *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
235 *d++ = (U8)(( uv & 0x3f) | 0x80);
a0ed51b3
LW
236 return d;
237 }
6b8eaf93 238#ifdef HAS_QUAD
a0ed51b3 239 {
eb160463
GS
240 *d++ = 0xff; /* Can't match U+FFFE! */
241 *d++ = 0x80; /* 6 Reserved bits */
242 *d++ = (U8)(((uv >> 60) & 0x0f) | 0x80); /* 2 Reserved bits */
243 *d++ = (U8)(((uv >> 54) & 0x3f) | 0x80);
244 *d++ = (U8)(((uv >> 48) & 0x3f) | 0x80);
245 *d++ = (U8)(((uv >> 42) & 0x3f) | 0x80);
246 *d++ = (U8)(((uv >> 36) & 0x3f) | 0x80);
247 *d++ = (U8)(((uv >> 30) & 0x3f) | 0x80);
248 *d++ = (U8)(((uv >> 24) & 0x3f) | 0x80);
249 *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80);
250 *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80);
251 *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
252 *d++ = (U8)(( uv & 0x3f) | 0x80);
a0ed51b3
LW
253 return d;
254 }
255#endif
1d72bdf6 256#endif /* Loop style */
a0ed51b3 257}
9041c2e3 258
646ca15d
JH
259/*
260
f7d739d1 261Tests if the first C<len> bytes of string C<s> form a valid UTF-8
646ca15d 262character. Note that an INVARIANT (i.e. ASCII) character is a valid
f7d739d1 263UTF-8 character. The number of bytes in the UTF-8 character
646ca15d
JH
264will be returned if it is valid, otherwise 0.
265
266This is the "slow" version as opposed to the "fast" version which is
267the "unrolled" IS_UTF8_CHAR(). E.g. for t/uni/class.t the speed
268difference is a factor of 2 to 3. For lengths (UTF8SKIP(s)) of four
269or less you should use the IS_UTF8_CHAR(), for lengths of five or more
270you should use the _slow(). In practice this means that the _slow()
271will be used very rarely, since the maximum Unicode code point (as of
272Unicode 4.1) is U+10FFFF, which encodes in UTF-8 to four bytes. Only
273the "Perl extended UTF-8" (the infamous 'v-strings') will encode into
274five bytes or more.
275
276=cut */
c053b435 277STATIC STRLEN
5f66b61c 278S_is_utf8_char_slow(const U8 *s, const STRLEN len)
646ca15d 279{
cd7e6c88 280 dTHX; /* The function called below requires thread context */
646ca15d 281
cd7e6c88 282 STRLEN actual_len;
646ca15d 283
cd7e6c88 284 PERL_ARGS_ASSERT_IS_UTF8_CHAR_SLOW;
646ca15d 285
cd7e6c88 286 utf8n_to_uvuni(s, len, &actual_len, UTF8_CHECK_ONLY);
646ca15d 287
cd7e6c88 288 return (actual_len == (STRLEN) -1) ? 0 : actual_len;
646ca15d 289}
9041c2e3
NIS
290
291/*
492a624f
KW
292=for apidoc is_utf8_char_buf
293
294Returns the number of bytes that comprise the first UTF-8 encoded character in
295buffer C<buf>. C<buf_end> should point to one position beyond the end of the
296buffer. 0 is returned if C<buf> does not point to a complete, valid UTF-8
297encoded character.
298
299Note that an INVARIANT character (i.e. ASCII on non-EBCDIC
300machines) is a valid UTF-8 character.
301
302=cut */
303
304STRLEN
305Perl_is_utf8_char_buf(const U8 *buf, const U8* buf_end)
306{
307
308 STRLEN len;
309
310 PERL_ARGS_ASSERT_IS_UTF8_CHAR_BUF;
311
312 if (buf_end <= buf) {
313 return 0;
314 }
315
316 len = buf_end - buf;
317 if (len > UTF8SKIP(buf)) {
318 len = UTF8SKIP(buf);
319 }
320
321#ifdef IS_UTF8_CHAR
322 if (IS_UTF8_CHAR_FAST(len))
323 return IS_UTF8_CHAR(buf, len) ? len : 0;
324#endif /* #ifdef IS_UTF8_CHAR */
325 return is_utf8_char_slow(buf, len);
326}
327
328/*
87cea99e 329=for apidoc is_utf8_char
eebe1485 330
76848387
KW
331DEPRECATED!
332
5da9da9e 333Tests if some arbitrary number of bytes begins in a valid UTF-8
2bbc8d55
SP
334character. Note that an INVARIANT (i.e. ASCII on non-EBCDIC machines)
335character is a valid UTF-8 character. The actual number of bytes in the UTF-8
336character will be returned if it is valid, otherwise 0.
9041c2e3 337
76848387 338This function is deprecated due to the possibility that malformed input could
a1433954 339cause reading beyond the end of the input buffer. Use L</is_utf8_char_buf>
76848387 340instead.
e0328548 341
82686b01 342=cut */
76848387 343
067a85ef 344STRLEN
668b6d8d 345Perl_is_utf8_char(const U8 *s)
386d01d6 346{
7918f24d 347 PERL_ARGS_ASSERT_IS_UTF8_CHAR;
492a624f 348
76848387 349 /* Assumes we have enough space, which is why this is deprecated */
492a624f 350 return is_utf8_char_buf(s, s + UTF8SKIP(s));
386d01d6
GS
351}
352
eaf7a4d2 353
6662521e 354/*
87cea99e 355=for apidoc is_utf8_string
6662521e 356
a1433954 357Returns true if the first C<len> bytes of string C<s> form a valid
9f7e3d64 358UTF-8 string, false otherwise. If C<len> is 0, it will be calculated
e0328548
KW
359using C<strlen(s)> (which means if you use this option, that C<s> has to have a
360terminating NUL byte). Note that all characters being ASCII constitute 'a
361valid UTF-8 string'.
6662521e 362
a1433954 363See also L</is_ascii_string>(), L</is_utf8_string_loclen>(), and L</is_utf8_string_loc>().
768c67ee 364
6662521e
GS
365=cut
366*/
367
8e84507e 368bool
668b6d8d 369Perl_is_utf8_string(const U8 *s, STRLEN len)
6662521e 370{
35da51f7 371 const U8* const send = s + (len ? len : strlen((const char *)s));
7fc63493 372 const U8* x = s;
067a85ef 373
7918f24d 374 PERL_ARGS_ASSERT_IS_UTF8_STRING;
1aa99e6b 375
6662521e 376 while (x < send) {
1acdb0da 377 /* Inline the easy bits of is_utf8_char() here for speed... */
e0328548
KW
378 if (UTF8_IS_INVARIANT(*x)) {
379 x++;
380 }
1acdb0da 381 else if (!UTF8_IS_START(*x))
e0328548 382 return FALSE;
1acdb0da
JH
383 else {
384 /* ... and call is_utf8_char() only if really needed. */
e0328548
KW
385 const STRLEN c = UTF8SKIP(x);
386 const U8* const next_char_ptr = x + c;
387
388 if (next_char_ptr > send) {
389 return FALSE;
390 }
391
768c67ee
JH
392 if (IS_UTF8_CHAR_FAST(c)) {
393 if (!IS_UTF8_CHAR(x, c))
e0328548 394 return FALSE;
3c614e38 395 }
e0328548
KW
396 else if (! is_utf8_char_slow(x, c)) {
397 return FALSE;
398 }
399 x = next_char_ptr;
1acdb0da 400 }
6662521e 401 }
768c67ee 402
067a85ef 403 return TRUE;
6662521e
GS
404}
405
67e989fb 406/*
814fafa7
NC
407Implemented as a macro in utf8.h
408
87cea99e 409=for apidoc is_utf8_string_loc
814fafa7 410
a1433954
KW
411Like L</is_utf8_string> but stores the location of the failure (in the
412case of "utf8ness failure") or the location C<s>+C<len> (in the case of
814fafa7
NC
413"utf8ness success") in the C<ep>.
414
a1433954 415See also L</is_utf8_string_loclen>() and L</is_utf8_string>().
814fafa7 416
87cea99e 417=for apidoc is_utf8_string_loclen
81cd54e3 418
a1433954
KW
419Like L</is_utf8_string>() but stores the location of the failure (in the
420case of "utf8ness failure") or the location C<s>+C<len> (in the case of
768c67ee
JH
421"utf8ness success") in the C<ep>, and the number of UTF-8
422encoded characters in the C<el>.
423
a1433954 424See also L</is_utf8_string_loc>() and L</is_utf8_string>().
81cd54e3
JH
425
426=cut
427*/
428
429bool
668b6d8d 430Perl_is_utf8_string_loclen(const U8 *s, STRLEN len, const U8 **ep, STRLEN *el)
81cd54e3 431{
35da51f7 432 const U8* const send = s + (len ? len : strlen((const char *)s));
7fc63493 433 const U8* x = s;
81cd54e3 434 STRLEN c;
3ebfea28 435 STRLEN outlen = 0;
7918f24d
NC
436
437 PERL_ARGS_ASSERT_IS_UTF8_STRING_LOCLEN;
81cd54e3 438
81cd54e3 439 while (x < send) {
e0328548
KW
440 const U8* next_char_ptr;
441
81cd54e3
JH
442 /* Inline the easy bits of is_utf8_char() here for speed... */
443 if (UTF8_IS_INVARIANT(*x))
e0328548 444 next_char_ptr = x + 1;
768c67ee
JH
445 else if (!UTF8_IS_START(*x))
446 goto out;
81cd54e3 447 else {
768c67ee 448 /* ... and call is_utf8_char() only if really needed. */
768c67ee 449 c = UTF8SKIP(x);
e0328548
KW
450 next_char_ptr = c + x;
451 if (next_char_ptr > send) {
452 goto out;
453 }
768c67ee
JH
454 if (IS_UTF8_CHAR_FAST(c)) {
455 if (!IS_UTF8_CHAR(x, c))
456 c = 0;
457 } else
458 c = is_utf8_char_slow(x, c);
768c67ee
JH
459 if (!c)
460 goto out;
81cd54e3 461 }
e0328548 462 x = next_char_ptr;
3ebfea28 463 outlen++;
81cd54e3 464 }
768c67ee
JH
465
466 out:
3ebfea28
AL
467 if (el)
468 *el = outlen;
469
768c67ee
JH
470 if (ep)
471 *ep = x;
3ebfea28 472 return (x == send);
81cd54e3
JH
473}
474
475/*
768c67ee 476
87cea99e 477=for apidoc utf8n_to_uvuni
67e989fb 478
9041c2e3 479Bottom level UTF-8 decode routine.
746afd53
KW
480Returns the code point value of the first character in the string C<s>,
481which is assumed to be in UTF-8 (or UTF-EBCDIC) encoding, and no longer than
482C<curlen> bytes; C<*retlen> (if C<retlen> isn't NULL) will be set to
483the length, in bytes, of that character.
949cf498
KW
484
485The value of C<flags> determines the behavior when C<s> does not point to a
486well-formed UTF-8 character. If C<flags> is 0, when a malformation is found,
524080c4
KW
487zero is returned and C<*retlen> is set so that (S<C<s> + C<*retlen>>) is the
488next possible position in C<s> that could begin a non-malformed character.
489Also, if UTF-8 warnings haven't been lexically disabled, a warning is raised.
949cf498
KW
490
491Various ALLOW flags can be set in C<flags> to allow (and not warn on)
492individual types of malformations, such as the sequence being overlong (that
493is, when there is a shorter sequence that can express the same code point;
494overlong sequences are expressly forbidden in the UTF-8 standard due to
495potential security issues). Another malformation example is the first byte of
496a character not being a legal first byte. See F<utf8.h> for the list of such
524080c4
KW
497flags. For allowed 0 length strings, this function returns 0; for allowed
498overlong sequences, the computed code point is returned; for all other allowed
499malformations, the Unicode REPLACEMENT CHARACTER is returned, as these have no
500determinable reasonable value.
949cf498
KW
501
502The UTF8_CHECK_ONLY flag overrides the behavior when a non-allowed (by other
503flags) malformation is found. If this flag is set, the routine assumes that
504the caller will raise a warning, and this function will silently just set
505C<retlen> to C<-1> and return zero.
506
507Certain code points are considered problematic. These are Unicode surrogates,
746afd53 508Unicode non-characters, and code points above the Unicode maximum of 0x10FFFF.
949cf498 509By default these are considered regular code points, but certain situations
5eafe189 510warrant special handling for them. If C<flags> contains
949cf498
KW
511UTF8_DISALLOW_ILLEGAL_INTERCHANGE, all three classes are treated as
512malformations and handled as such. The flags UTF8_DISALLOW_SURROGATE,
513UTF8_DISALLOW_NONCHAR, and UTF8_DISALLOW_SUPER (meaning above the legal Unicode
514maximum) can be set to disallow these categories individually.
515
516The flags UTF8_WARN_ILLEGAL_INTERCHANGE, UTF8_WARN_SURROGATE,
517UTF8_WARN_NONCHAR, and UTF8_WARN_SUPER will cause warning messages to be raised
518for their respective categories, but otherwise the code points are considered
519valid (not malformations). To get a category to both be treated as a
520malformation and raise a warning, specify both the WARN and DISALLOW flags.
521(But note that warnings are not raised if lexically disabled nor if
522UTF8_CHECK_ONLY is also specified.)
523
524Very large code points (above 0x7FFF_FFFF) are considered more problematic than
525the others that are above the Unicode legal maximum. There are several
eb83ed87
KW
526reasons: they requre at least 32 bits to represent them on ASCII platforms, are
527not representable at all on EBCDIC platforms, and the original UTF-8
528specification never went above this number (the current 0x10FFFF limit was
529imposed later). (The smaller ones, those that fit into 32 bits, are
530representable by a UV on ASCII platforms, but not by an IV, which means that
531the number of operations that can be performed on them is quite restricted.)
532The UTF-8 encoding on ASCII platforms for these large code points begins with a
533byte containing 0xFE or 0xFF. The UTF8_DISALLOW_FE_FF flag will cause them to
534be treated as malformations, while allowing smaller above-Unicode code points.
535(Of course UTF8_DISALLOW_SUPER will treat all above-Unicode code points,
536including these, as malformations.) Similarly, UTF8_WARN_FE_FF acts just like
537the other WARN flags, but applies just to these code points.
949cf498
KW
538
539All other code points corresponding to Unicode characters, including private
540use and those yet to be assigned, are never considered malformed and never
541warn.
67e989fb 542
ec5f19d0 543Most code should use L</utf8_to_uvchr_buf>() rather than call this directly.
9041c2e3 544
37607a96
PK
545=cut
546*/
67e989fb 547
a0ed51b3 548UV
7fc63493 549Perl_utf8n_to_uvuni(pTHX_ const U8 *s, STRLEN curlen, STRLEN *retlen, U32 flags)
a0ed51b3 550{
97aff369 551 dVAR;
d4c19fe8 552 const U8 * const s0 = s;
eb83ed87 553 U8 overflow_byte = '\0'; /* Save byte in case of overflow */
0b8d30e8 554 U8 * send;
eb83ed87
KW
555 UV uv = *s;
556 STRLEN expectlen;
949cf498 557 SV* sv = NULL;
eb83ed87
KW
558 UV outlier_ret = 0; /* return value when input is in error or problematic
559 */
560 UV pack_warn = 0; /* Save result of packWARN() for later */
561 bool unexpected_non_continuation = FALSE;
562 bool overflowed = FALSE;
2f8f112e 563 bool do_overlong_test = TRUE; /* May have to skip this test */
a0dbb045 564
eb83ed87 565 const char* const malformed_text = "Malformed UTF-8 character";
7918f24d 566
eb83ed87 567 PERL_ARGS_ASSERT_UTF8N_TO_UVUNI;
a0dbb045 568
eb83ed87
KW
569 /* The order of malformation tests here is important. We should consume as
570 * few bytes as possible in order to not skip any valid character. This is
571 * required by the Unicode Standard (section 3.9 of Unicode 6.0); see also
572 * http://unicode.org/reports/tr36 for more discussion as to why. For
573 * example, once we've done a UTF8SKIP, we can tell the expected number of
574 * bytes, and could fail right off the bat if the input parameters indicate
575 * that there are too few available. But it could be that just that first
576 * byte is garbled, and the intended character occupies fewer bytes. If we
577 * blindly assumed that the first byte is correct, and skipped based on
578 * that number, we could skip over a valid input character. So instead, we
579 * always examine the sequence byte-by-byte.
580 *
581 * We also should not consume too few bytes, otherwise someone could inject
582 * things. For example, an input could be deliberately designed to
583 * overflow, and if this code bailed out immediately upon discovering that,
584 * returning to the caller *retlen pointing to the very next byte (one
585 * which is actually part of of the overflowing sequence), that could look
586 * legitimate to the caller, which could discard the initial partial
587 * sequence and process the rest, inappropriately */
588
589 /* Zero length strings, if allowed, of necessity are zero */
590 if (curlen == 0) {
591 if (retlen) {
592 *retlen = 0;
593 }
a0dbb045 594
eb83ed87
KW
595 if (flags & UTF8_ALLOW_EMPTY) {
596 return 0;
597 }
598 if (! (flags & UTF8_CHECK_ONLY)) {
599 sv = sv_2mortal(Perl_newSVpvf(aTHX_ "%s (empty string)", malformed_text));
600 }
0c443dc2
JH
601 goto malformed;
602 }
603
eb83ed87
KW
604 expectlen = UTF8SKIP(s);
605
606 /* A well-formed UTF-8 character, as the vast majority of calls to this
607 * function will be for, has this expected length. For efficiency, set
608 * things up here to return it. It will be overriden only in those rare
609 * cases where a malformation is found */
610 if (retlen) {
611 *retlen = expectlen;
612 }
613
614 /* An invariant is trivially well-formed */
1d72bdf6 615 if (UTF8_IS_INVARIANT(uv)) {
c4d5f83a 616 return (UV) (NATIVE_TO_UTF(*s));
a0ed51b3 617 }
67e989fb 618
eb83ed87
KW
619 /* A continuation character can't start a valid sequence */
620 if (UTF8_IS_CONTINUATION(uv)) {
621 if (flags & UTF8_ALLOW_CONTINUATION) {
622 if (retlen) {
623 *retlen = 1;
624 }
625 return UNICODE_REPLACEMENT;
626 }
ba210ebe 627
eb83ed87
KW
628 if (! (flags & UTF8_CHECK_ONLY)) {
629 sv = sv_2mortal(Perl_newSVpvf(aTHX_ "%s (unexpected continuation byte 0x%02x, with no preceding start byte)", malformed_text, *s0));
630 }
631 curlen = 1;
ba210ebe
JH
632 goto malformed;
633 }
9041c2e3 634
1d72bdf6 635#ifdef EBCDIC
75383841 636 uv = NATIVE_TO_UTF(uv);
1d72bdf6
NIS
637#endif
638
eb83ed87
KW
639 /* Here is not a continuation byte, nor an invariant. The only thing left
640 * is a start byte (possibly for an overlong) */
ba210ebe 641
eb83ed87
KW
642 /* Remove the leading bits that indicate the number of bytes in the
643 * character's whole UTF-8 sequence, leaving just the bits that are part of
644 * the value */
645 uv &= UTF_START_MASK(expectlen);
ba210ebe 646
eb83ed87
KW
647 /* Now, loop through the remaining bytes in the character's sequence,
648 * accumulating each into the working value as we go. Be sure to not look
649 * past the end of the input string */
0b8d30e8
KW
650 send = (U8*) s0 + ((expectlen <= curlen) ? expectlen : curlen);
651
eb83ed87
KW
652 for (s = s0 + 1; s < send; s++) {
653 if (UTF8_IS_CONTINUATION(*s)) {
654#ifndef EBCDIC /* Can't overflow in EBCDIC */
655 if (uv & UTF_ACCUMULATION_OVERFLOW_MASK) {
656
657 /* The original implementors viewed this malformation as more
658 * serious than the others (though I, khw, don't understand
659 * why, since other malformations also give very very wrong
660 * results), so there is no way to turn off checking for it.
661 * Set a flag, but keep going in the loop, so that we absorb
662 * the rest of the bytes that comprise the character. */
663 overflowed = TRUE;
664 overflow_byte = *s; /* Save for warning message's use */
665 }
666#endif
8850bf83 667 uv = UTF8_ACCUMULATE(uv, *s);
eb83ed87
KW
668 }
669 else {
670 /* Here, found a non-continuation before processing all expected
671 * bytes. This byte begins a new character, so quit, even if
672 * allowing this malformation. */
673 unexpected_non_continuation = TRUE;
674 break;
675 }
676 } /* End of loop through the character's bytes */
677
678 /* Save how many bytes were actually in the character */
679 curlen = s - s0;
680
681 /* The loop above finds two types of malformations: non-continuation and/or
682 * overflow. The non-continuation malformation is really a too-short
683 * malformation, as it means that the current character ended before it was
684 * expected to (being terminated prematurely by the beginning of the next
685 * character, whereas in the too-short malformation there just are too few
686 * bytes available to hold the character. In both cases, the check below
687 * that we have found the expected number of bytes would fail if executed.)
688 * Thus the non-continuation malformation is really unnecessary, being a
689 * subset of the too-short malformation. But there may be existing
690 * applications that are expecting the non-continuation type, so we retain
691 * it, and return it in preference to the too-short malformation. (If this
692 * code were being written from scratch, the two types might be collapsed
693 * into one.) I, khw, am also giving priority to returning the
694 * non-continuation and too-short malformations over overflow when multiple
695 * ones are present. I don't know of any real reason to prefer one over
696 * the other, except that it seems to me that multiple-byte errors trumps
697 * errors from a single byte */
698 if (unexpected_non_continuation) {
699 if (!(flags & UTF8_ALLOW_NON_CONTINUATION)) {
700 if (! (flags & UTF8_CHECK_ONLY)) {
701 if (curlen == 1) {
702 sv = sv_2mortal(Perl_newSVpvf(aTHX_ "%s (unexpected non-continuation byte 0x%02x, immediately after start byte 0x%02x)", malformed_text, *s, *s0));
703 }
704 else {
705 sv = sv_2mortal(Perl_newSVpvf(aTHX_ "%s (unexpected non-continuation byte 0x%02x, %d bytes after start byte 0x%02x, expected %d bytes)", malformed_text, *s, (int) curlen, *s0, (int)expectlen));
a0dbb045
JH
706 }
707 }
eb83ed87
KW
708 goto malformed;
709 }
710 uv = UNICODE_REPLACEMENT;
2f8f112e
KW
711
712 /* Skip testing for overlongs, as the REPLACEMENT may not be the same
713 * as what the original expectations were. */
714 do_overlong_test = FALSE;
eb83ed87
KW
715 if (retlen) {
716 *retlen = curlen;
717 }
718 }
719 else if (curlen < expectlen) {
720 if (! (flags & UTF8_ALLOW_SHORT)) {
721 if (! (flags & UTF8_CHECK_ONLY)) {
722 sv = sv_2mortal(Perl_newSVpvf(aTHX_ "%s (%d byte%s, need %d, after start byte 0x%02x)", malformed_text, (int)curlen, curlen == 1 ? "" : "s", (int)expectlen, *s0));
a0dbb045 723 }
eb83ed87
KW
724 goto malformed;
725 }
726 uv = UNICODE_REPLACEMENT;
2f8f112e 727 do_overlong_test = FALSE;
eb83ed87
KW
728 if (retlen) {
729 *retlen = curlen;
730 }
731 }
732
733#ifndef EBCDIC /* EBCDIC allows FE, FF, can't overflow */
2f8f112e 734 if ((*s0 & 0xFE) == 0xFE /* matches both FE, FF */
eb83ed87
KW
735 && (flags & (UTF8_WARN_FE_FF|UTF8_DISALLOW_FE_FF)))
736 {
737 /* By adding UTF8_CHECK_ONLY to the test, we avoid unnecessary
738 * generation of the sv, since no warnings are raised under CHECK */
739 if ((flags & (UTF8_WARN_FE_FF|UTF8_CHECK_ONLY)) == UTF8_WARN_FE_FF
740 && ckWARN_d(WARN_UTF8))
741 {
742 sv = sv_2mortal(Perl_newSVpvf(aTHX_ "%s Code point beginning with byte 0x%02X is not Unicode, and not portable", malformed_text, *s0));
743 pack_warn = packWARN(WARN_UTF8);
744 }
745 if (flags & UTF8_DISALLOW_FE_FF) {
746 goto malformed;
ba210ebe 747 }
ba210ebe 748 }
2f8f112e 749 if (overflowed) {
ba210ebe 750
eb83ed87
KW
751 /* If the first byte is FF, it will overflow a 32-bit word. If the
752 * first byte is FE, it will overflow a signed 32-bit word. The
753 * above preserves backward compatibility, since its message was used
754 * in earlier versions of this code in preference to overflow */
755 sv = sv_2mortal(Perl_newSVpvf(aTHX_ "%s (overflow at byte 0x%02x, after start byte 0x%02x)", malformed_text, overflow_byte, *s0));
ba210ebe 756 goto malformed;
eb83ed87
KW
757 }
758#endif
759
2f8f112e
KW
760 if (do_overlong_test
761 && expectlen > (STRLEN)UNISKIP(uv)
762 && ! (flags & UTF8_ALLOW_LONG))
763 {
eb83ed87
KW
764 /* The overlong malformation has lower precedence than the others.
765 * Note that if this malformation is allowed, we return the actual
766 * value, instead of the replacement character. This is because this
767 * value is actually well-defined. */
768 if (! (flags & UTF8_CHECK_ONLY)) {
769 sv = sv_2mortal(Perl_newSVpvf(aTHX_ "%s (%d byte%s, need %d, after start byte 0x%02x)", malformed_text, (int)expectlen, expectlen == 1 ? "": "s", UNISKIP(uv), *s0));
770 }
771 goto malformed;
772 }
773
774 /* Here, the input is considered to be well-formed , but could be a
775 * problematic code point that is not allowed by the input parameters. */
776 if (uv >= UNICODE_SURROGATE_FIRST /* isn't problematic if < this */
777 && (flags & (UTF8_DISALLOW_ILLEGAL_INTERCHANGE
778 |UTF8_WARN_ILLEGAL_INTERCHANGE)))
779 {
949cf498 780 if (UNICODE_IS_SURROGATE(uv)) {
eb83ed87
KW
781 if ((flags & (UTF8_WARN_SURROGATE|UTF8_CHECK_ONLY)) == UTF8_WARN_SURROGATE
782 && ckWARN2_d(WARN_UTF8, WARN_SURROGATE))
783 {
111d382d 784 sv = sv_2mortal(Perl_newSVpvf(aTHX_ "UTF-16 surrogate U+%04"UVXf"", uv));
eb83ed87 785 pack_warn = packWARN2(WARN_UTF8, WARN_SURROGATE);
949cf498
KW
786 }
787 if (flags & UTF8_DISALLOW_SURROGATE) {
788 goto disallowed;
789 }
790 }
791 else if (UNICODE_IS_NONCHAR(uv)) {
eb83ed87
KW
792 if ((flags & (UTF8_WARN_NONCHAR|UTF8_CHECK_ONLY)) == UTF8_WARN_NONCHAR
793 && ckWARN2_d(WARN_UTF8, WARN_NONCHAR))
794 {
111d382d 795 sv = sv_2mortal(Perl_newSVpvf(aTHX_ "Unicode non-character U+%04"UVXf" is illegal for open interchange", uv));
eb83ed87 796 pack_warn = packWARN2(WARN_UTF8, WARN_NONCHAR);
949cf498
KW
797 }
798 if (flags & UTF8_DISALLOW_NONCHAR) {
799 goto disallowed;
800 }
801 }
802 else if ((uv > PERL_UNICODE_MAX)) {
eb83ed87
KW
803 if ((flags & (UTF8_WARN_SUPER|UTF8_CHECK_ONLY)) == UTF8_WARN_SUPER
804 && ckWARN2_d(WARN_UTF8, WARN_NON_UNICODE))
805 {
111d382d 806 sv = sv_2mortal(Perl_newSVpvf(aTHX_ "Code point 0x%04"UVXf" is not Unicode, may not be portable", uv));
eb83ed87 807 pack_warn = packWARN2(WARN_UTF8, WARN_NON_UNICODE);
949cf498
KW
808 }
809 if (flags & UTF8_DISALLOW_SUPER) {
810 goto disallowed;
811 }
812 }
813
eb83ed87
KW
814 if (sv) {
815 outlier_ret = uv;
816 goto do_warn;
817 }
818
949cf498
KW
819 /* Here, this is not considered a malformed character, so drop through
820 * to return it */
a0ed51b3 821 }
ba210ebe 822
a0ed51b3 823 return uv;
ba210ebe 824
eb83ed87
KW
825 /* There are three cases which get to beyond this point. In all 3 cases:
826 * <sv> if not null points to a string to print as a warning.
827 * <curlen> is what <*retlen> should be set to if UTF8_CHECK_ONLY isn't
828 * set.
829 * <outlier_ret> is what return value to use if UTF8_CHECK_ONLY isn't set.
830 * This is done by initializing it to 0, and changing it only
831 * for case 1).
832 * The 3 cases are:
833 * 1) The input is valid but problematic, and to be warned about. The
834 * return value is the resultant code point; <*retlen> is set to
835 * <curlen>, the number of bytes that comprise the code point.
836 * <pack_warn> contains the result of packWARN() for the warning
837 * types. The entry point for this case is the label <do_warn>;
838 * 2) The input is a valid code point but disallowed by the parameters to
839 * this function. The return value is 0. If UTF8_CHECK_ONLY is set,
840 * <*relen> is -1; otherwise it is <curlen>, the number of bytes that
841 * comprise the code point. <pack_warn> contains the result of
842 * packWARN() for the warning types. The entry point for this case is
843 * the label <disallowed>.
844 * 3) The input is malformed. The return value is 0. If UTF8_CHECK_ONLY
845 * is set, <*relen> is -1; otherwise it is <curlen>, the number of
846 * bytes that comprise the malformation. All such malformations are
847 * assumed to be warning type <utf8>. The entry point for this case
848 * is the label <malformed>.
849 */
949cf498 850
ba210ebe
JH
851malformed:
852
eb83ed87
KW
853 if (sv && ckWARN_d(WARN_UTF8)) {
854 pack_warn = packWARN(WARN_UTF8);
855 }
856
857disallowed:
858
fcc8fcf6 859 if (flags & UTF8_CHECK_ONLY) {
ba210ebe 860 if (retlen)
10edeb5d 861 *retlen = ((STRLEN) -1);
ba210ebe
JH
862 return 0;
863 }
864
eb83ed87 865do_warn:
5b311467 866
eb83ed87
KW
867 if (pack_warn) { /* <pack_warn> was initialized to 0, and changed only
868 if warnings are to be raised. */
f555bc63 869 const char * const string = SvPVX_const(sv);
a0dbb045 870
f555bc63
KW
871 if (PL_op)
872 Perl_warner(aTHX_ pack_warn, "%s in %s", string, OP_DESC(PL_op));
873 else
874 Perl_warner(aTHX_ pack_warn, "%s", string);
a0dbb045
JH
875 }
876
eb83ed87
KW
877 if (retlen) {
878 *retlen = curlen;
879 }
ba210ebe 880
eb83ed87 881 return outlier_ret;
a0ed51b3
LW
882}
883
8e84507e 884/*
ec5f19d0
KW
885=for apidoc utf8_to_uvchr_buf
886
887Returns the native code point of the first character in the string C<s> which
888is assumed to be in UTF-8 encoding; C<send> points to 1 beyond the end of C<s>.
524080c4 889C<*retlen> will be set to the length, in bytes, of that character.
ec5f19d0 890
524080c4
KW
891If C<s> does not point to a well-formed UTF-8 character and UTF8 warnings are
892enabled, zero is returned and C<*retlen> is set (if C<retlen> isn't
893NULL) to -1. If those warnings are off, the computed value if well-defined (or
894the Unicode REPLACEMENT CHARACTER, if not) is silently returned, and C<*retlen>
895is set (if C<retlen> isn't NULL) so that (S<C<s> + C<*retlen>>) is the
896next possible position in C<s> that could begin a non-malformed character.
897See L</utf8n_to_uvuni> for details on when the REPLACEMENT CHARACTER is returned.
ec5f19d0
KW
898
899=cut
900*/
901
902
903UV
904Perl_utf8_to_uvchr_buf(pTHX_ const U8 *s, const U8 *send, STRLEN *retlen)
905{
906 PERL_ARGS_ASSERT_UTF8_TO_UVCHR_BUF;
907
908 assert(s < send);
909
910 return utf8n_to_uvchr(s, send - s, retlen,
911 ckWARN_d(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY);
912}
913
27d6c58a
KW
914/* Like L</utf8_to_uvchr_buf>(), but should only be called when it is known that
915 * there are no malformations in the input UTF-8 string C<s>. Currently, some
916 * malformations are checked for, but this checking likely will be removed in
917 * the future */
918
919UV
920Perl_valid_utf8_to_uvchr(pTHX_ const U8 *s, STRLEN *retlen)
921{
922 PERL_ARGS_ASSERT_VALID_UTF8_TO_UVCHR;
923
924 return utf8_to_uvchr_buf(s, s + UTF8_MAXBYTES, retlen);
925}
926
ec5f19d0 927/*
87cea99e 928=for apidoc utf8_to_uvchr
9041c2e3 929
977c1d31
KW
930DEPRECATED!
931
6ee84de2 932Returns the native code point of the first character in the string C<s>
1e54db1a 933which is assumed to be in UTF-8 encoding; C<retlen> will be set to the
9041c2e3
NIS
934length, in bytes, of that character.
935
4b88fb76 936Some, but not all, UTF-8 malformations are detected, and in fact, some
977c1d31
KW
937malformed input could cause reading beyond the end of the input buffer, which
938is why this function is deprecated. Use L</utf8_to_uvchr_buf> instead.
4b88fb76 939
524080c4
KW
940If C<s> points to one of the detected malformations, and UTF8 warnings are
941enabled, zero is returned and C<*retlen> is set (if C<retlen> isn't
942NULL) to -1. If those warnings are off, the computed value if well-defined (or
943the Unicode REPLACEMENT CHARACTER, if not) is silently returned, and C<*retlen>
944is set (if C<retlen> isn't NULL) so that (S<C<s> + C<*retlen>>) is the
945next possible position in C<s> that could begin a non-malformed character.
946See L</utf8n_to_uvuni> for details on when the REPLACEMENT CHARACTER is returned.
9041c2e3
NIS
947
948=cut
949*/
950
951UV
7fc63493 952Perl_utf8_to_uvchr(pTHX_ const U8 *s, STRLEN *retlen)
9041c2e3 953{
7918f24d
NC
954 PERL_ARGS_ASSERT_UTF8_TO_UVCHR;
955
4b88fb76 956 return valid_utf8_to_uvchr(s, retlen);
9041c2e3
NIS
957}
958
959/*
ec5f19d0
KW
960=for apidoc utf8_to_uvuni_buf
961
962Returns the Unicode code point of the first character in the string C<s> which
963is assumed to be in UTF-8 encoding; C<send> points to 1 beyond the end of C<s>.
964C<retlen> will be set to the length, in bytes, of that character.
965
966This function should only be used when the returned UV is considered
967an index into the Unicode semantic tables (e.g. swashes).
968
524080c4
KW
969If C<s> does not point to a well-formed UTF-8 character and UTF8 warnings are
970enabled, zero is returned and C<*retlen> is set (if C<retlen> isn't
971NULL) to -1. If those warnings are off, the computed value if well-defined (or
972the Unicode REPLACEMENT CHARACTER, if not) is silently returned, and C<*retlen>
973is set (if C<retlen> isn't NULL) so that (S<C<s> + C<*retlen>>) is the
974next possible position in C<s> that could begin a non-malformed character.
975See L</utf8n_to_uvuni> for details on when the REPLACEMENT CHARACTER is returned.
ec5f19d0
KW
976
977=cut
978*/
979
980UV
981Perl_utf8_to_uvuni_buf(pTHX_ const U8 *s, const U8 *send, STRLEN *retlen)
982{
983 PERL_ARGS_ASSERT_UTF8_TO_UVUNI_BUF;
984
985 assert(send > s);
986
987 /* Call the low level routine asking for checks */
988 return Perl_utf8n_to_uvuni(aTHX_ s, send -s, retlen,
989 ckWARN_d(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY);
990}
991
27d6c58a
KW
992/* Like L</utf8_to_uvuni_buf>(), but should only be called when it is known that
993 * there are no malformations in the input UTF-8 string C<s>. Currently, some
994 * malformations are checked for, but this checking likely will be removed in
995 * the future */
996
997UV
998Perl_valid_utf8_to_uvuni(pTHX_ const U8 *s, STRLEN *retlen)
999{
1000 PERL_ARGS_ASSERT_VALID_UTF8_TO_UVUNI;
1001
1002 return utf8_to_uvuni_buf(s, s + UTF8_MAXBYTES, retlen);
1003}
1004
ec5f19d0 1005/*
87cea99e 1006=for apidoc utf8_to_uvuni
9041c2e3 1007
977c1d31
KW
1008DEPRECATED!
1009
9041c2e3 1010Returns the Unicode code point of the first character in the string C<s>
1e54db1a 1011which is assumed to be in UTF-8 encoding; C<retlen> will be set to the
9041c2e3
NIS
1012length, in bytes, of that character.
1013
39e518fd
KW
1014This function should only be used when the returned UV is considered
1015an index into the Unicode semantic tables (e.g. swashes).
1016
4b88fb76 1017Some, but not all, UTF-8 malformations are detected, and in fact, some
977c1d31
KW
1018malformed input could cause reading beyond the end of the input buffer, which
1019is why this function is deprecated. Use L</utf8_to_uvuni_buf> instead.
9041c2e3 1020
524080c4
KW
1021If C<s> points to one of the detected malformations, and UTF8 warnings are
1022enabled, zero is returned and C<*retlen> is set (if C<retlen> doesn't point to
1023NULL) to -1. If those warnings are off, the computed value if well-defined (or
1024the Unicode REPLACEMENT CHARACTER, if not) is silently returned, and C<*retlen>
1025is set (if C<retlen> isn't NULL) so that (S<C<s> + C<*retlen>>) is the
1026next possible position in C<s> that could begin a non-malformed character.
1027See L</utf8n_to_uvuni> for details on when the REPLACEMENT CHARACTER is returned.
8e84507e
NIS
1028
1029=cut
1030*/
1031
1032UV
7fc63493 1033Perl_utf8_to_uvuni(pTHX_ const U8 *s, STRLEN *retlen)
8e84507e 1034{
7918f24d
NC
1035 PERL_ARGS_ASSERT_UTF8_TO_UVUNI;
1036
4b88fb76 1037 return valid_utf8_to_uvuni(s, retlen);
8e84507e
NIS
1038}
1039
b76347f2 1040/*
87cea99e 1041=for apidoc utf8_length
b76347f2
JH
1042
1043Return the length of the UTF-8 char encoded string C<s> in characters.
02eb7b47
JH
1044Stops at C<e> (inclusive). If C<e E<lt> s> or if the scan would end
1045up past C<e>, croaks.
b76347f2
JH
1046
1047=cut
1048*/
1049
1050STRLEN
35a4481c 1051Perl_utf8_length(pTHX_ const U8 *s, const U8 *e)
b76347f2 1052{
97aff369 1053 dVAR;
b76347f2
JH
1054 STRLEN len = 0;
1055
7918f24d
NC
1056 PERL_ARGS_ASSERT_UTF8_LENGTH;
1057
8850bf83
JH
1058 /* Note: cannot use UTF8_IS_...() too eagerly here since e.g.
1059 * the bitops (especially ~) can create illegal UTF-8.
1060 * In other words: in Perl UTF-8 is not just for Unicode. */
1061
a3b680e6
AL
1062 if (e < s)
1063 goto warn_and_return;
b76347f2 1064 while (s < e) {
8e91ec7f
AV
1065 if (!UTF8_IS_INVARIANT(*s))
1066 s += UTF8SKIP(s);
1067 else
1068 s++;
1069 len++;
1070 }
1071
1072 if (e != s) {
1073 len--;
1074 warn_and_return:
9b387841
NC
1075 if (PL_op)
1076 Perl_ck_warner_d(aTHX_ packWARN(WARN_UTF8),
1077 "%s in %s", unees, OP_DESC(PL_op));
1078 else
61a12c31 1079 Perl_ck_warner_d(aTHX_ packWARN(WARN_UTF8), "%s", unees);
b76347f2
JH
1080 }
1081
1082 return len;
1083}
1084
b06226ff 1085/*
87cea99e 1086=for apidoc utf8_distance
b06226ff 1087
1e54db1a 1088Returns the number of UTF-8 characters between the UTF-8 pointers C<a>
b06226ff
JH
1089and C<b>.
1090
1091WARNING: use only if you *know* that the pointers point inside the
1092same UTF-8 buffer.
1093
37607a96
PK
1094=cut
1095*/
a0ed51b3 1096
02eb7b47 1097IV
35a4481c 1098Perl_utf8_distance(pTHX_ const U8 *a, const U8 *b)
a0ed51b3 1099{
7918f24d
NC
1100 PERL_ARGS_ASSERT_UTF8_DISTANCE;
1101
bf1665bc 1102 return (a < b) ? -1 * (IV) utf8_length(a, b) : (IV) utf8_length(b, a);
a0ed51b3
LW
1103}
1104
b06226ff 1105/*
87cea99e 1106=for apidoc utf8_hop
b06226ff 1107
8850bf83
JH
1108Return the UTF-8 pointer C<s> displaced by C<off> characters, either
1109forward or backward.
b06226ff
JH
1110
1111WARNING: do not use the following unless you *know* C<off> is within
8850bf83
JH
1112the UTF-8 data pointed to by C<s> *and* that on entry C<s> is aligned
1113on the first byte of character or just after the last byte of a character.
b06226ff 1114
37607a96
PK
1115=cut
1116*/
a0ed51b3
LW
1117
1118U8 *
4373e329 1119Perl_utf8_hop(pTHX_ const U8 *s, I32 off)
a0ed51b3 1120{
7918f24d
NC
1121 PERL_ARGS_ASSERT_UTF8_HOP;
1122
96a5add6 1123 PERL_UNUSED_CONTEXT;
8850bf83
JH
1124 /* Note: cannot use UTF8_IS_...() too eagerly here since e.g
1125 * the bitops (especially ~) can create illegal UTF-8.
1126 * In other words: in Perl UTF-8 is not just for Unicode. */
1127
a0ed51b3
LW
1128 if (off >= 0) {
1129 while (off--)
1130 s += UTF8SKIP(s);
1131 }
1132 else {
1133 while (off++) {
1134 s--;
8850bf83
JH
1135 while (UTF8_IS_CONTINUATION(*s))
1136 s--;
a0ed51b3
LW
1137 }
1138 }
4373e329 1139 return (U8 *)s;
a0ed51b3
LW
1140}
1141
6940069f 1142/*
fed3ba5d
NC
1143=for apidoc bytes_cmp_utf8
1144
a1433954
KW
1145Compares the sequence of characters (stored as octets) in C<b>, C<blen> with the
1146sequence of characters (stored as UTF-8) in C<u>, C<ulen>. Returns 0 if they are
fed3ba5d
NC
1147equal, -1 or -2 if the first string is less than the second string, +1 or +2
1148if the first string is greater than the second string.
1149
1150-1 or +1 is returned if the shorter string was identical to the start of the
1151longer string. -2 or +2 is returned if the was a difference between characters
1152within the strings.
1153
1154=cut
1155*/
1156
1157int
1158Perl_bytes_cmp_utf8(pTHX_ const U8 *b, STRLEN blen, const U8 *u, STRLEN ulen)
1159{
1160 const U8 *const bend = b + blen;
1161 const U8 *const uend = u + ulen;
1162
1163 PERL_ARGS_ASSERT_BYTES_CMP_UTF8;
1164
1165 PERL_UNUSED_CONTEXT;
1166
1167 while (b < bend && u < uend) {
1168 U8 c = *u++;
1169 if (!UTF8_IS_INVARIANT(c)) {
1170 if (UTF8_IS_DOWNGRADEABLE_START(c)) {
1171 if (u < uend) {
1172 U8 c1 = *u++;
1173 if (UTF8_IS_CONTINUATION(c1)) {
356979f4 1174 c = UNI_TO_NATIVE(TWO_BYTE_UTF8_TO_UNI(c, c1));
fed3ba5d
NC
1175 } else {
1176 Perl_ck_warner_d(aTHX_ packWARN(WARN_UTF8),
1177 "Malformed UTF-8 character "
1178 "(unexpected non-continuation byte 0x%02x"
1179 ", immediately after start byte 0x%02x)"
1180 /* Dear diag.t, it's in the pod. */
1181 "%s%s", c1, c,
1182 PL_op ? " in " : "",
1183 PL_op ? OP_DESC(PL_op) : "");
1184 return -2;
1185 }
1186 } else {
1187 if (PL_op)
1188 Perl_ck_warner_d(aTHX_ packWARN(WARN_UTF8),
1189 "%s in %s", unees, OP_DESC(PL_op));
1190 else
61a12c31 1191 Perl_ck_warner_d(aTHX_ packWARN(WARN_UTF8), "%s", unees);
fed3ba5d
NC
1192 return -2; /* Really want to return undef :-) */
1193 }
1194 } else {
1195 return -2;
1196 }
1197 }
1198 if (*b != c) {
1199 return *b < c ? -2 : +2;
1200 }
1201 ++b;
1202 }
1203
1204 if (b == bend && u == uend)
1205 return 0;
1206
1207 return b < bend ? +1 : -1;
1208}
1209
1210/*
87cea99e 1211=for apidoc utf8_to_bytes
6940069f 1212
2bbc8d55 1213Converts a string C<s> of length C<len> from UTF-8 into native byte encoding.
a1433954
KW
1214Unlike L</bytes_to_utf8>, this over-writes the original string, and
1215updates C<len> to contain the new length.
67e989fb 1216Returns zero on failure, setting C<len> to -1.
6940069f 1217
a1433954 1218If you need a copy of the string, see L</bytes_from_utf8>.
95be277c 1219
6940069f
GS
1220=cut
1221*/
1222
1223U8 *
37607a96 1224Perl_utf8_to_bytes(pTHX_ U8 *s, STRLEN *len)
6940069f 1225{
d4c19fe8
AL
1226 U8 * const save = s;
1227 U8 * const send = s + *len;
6940069f 1228 U8 *d;
246fae53 1229
7918f24d
NC
1230 PERL_ARGS_ASSERT_UTF8_TO_BYTES;
1231
1e54db1a 1232 /* ensure valid UTF-8 and chars < 256 before updating string */
d4c19fe8 1233 while (s < send) {
dcad2880
JH
1234 U8 c = *s++;
1235
1d72bdf6
NIS
1236 if (!UTF8_IS_INVARIANT(c) &&
1237 (!UTF8_IS_DOWNGRADEABLE_START(c) || (s >= send)
1238 || !(c = *s++) || !UTF8_IS_CONTINUATION(c))) {
10edeb5d 1239 *len = ((STRLEN) -1);
dcad2880
JH
1240 return 0;
1241 }
246fae53 1242 }
dcad2880
JH
1243
1244 d = s = save;
6940069f 1245 while (s < send) {
ed646e6e 1246 STRLEN ulen;
4b88fb76 1247 *d++ = (U8)utf8_to_uvchr_buf(s, send, &ulen);
ed646e6e 1248 s += ulen;
6940069f
GS
1249 }
1250 *d = '\0';
246fae53 1251 *len = d - save;
6940069f
GS
1252 return save;
1253}
1254
1255/*
87cea99e 1256=for apidoc bytes_from_utf8
f9a63242 1257
2bbc8d55 1258Converts a string C<s> of length C<len> from UTF-8 into native byte encoding.
a1433954 1259Unlike L</utf8_to_bytes> but like L</bytes_to_utf8>, returns a pointer to
ef9edfd0
JH
1260the newly-created string, and updates C<len> to contain the new
1261length. Returns the original string if no conversion occurs, C<len>
1262is unchanged. Do nothing if C<is_utf8> points to 0. Sets C<is_utf8> to
2bbc8d55
SP
12630 if C<s> is converted or consisted entirely of characters that are invariant
1264in utf8 (i.e., US-ASCII on non-EBCDIC machines).
f9a63242 1265
37607a96
PK
1266=cut
1267*/
f9a63242
JH
1268
1269U8 *
e1ec3a88 1270Perl_bytes_from_utf8(pTHX_ const U8 *s, STRLEN *len, bool *is_utf8)
f9a63242 1271{
f9a63242 1272 U8 *d;
e1ec3a88
AL
1273 const U8 *start = s;
1274 const U8 *send;
f9a63242
JH
1275 I32 count = 0;
1276
7918f24d
NC
1277 PERL_ARGS_ASSERT_BYTES_FROM_UTF8;
1278
96a5add6 1279 PERL_UNUSED_CONTEXT;
f9a63242 1280 if (!*is_utf8)
73d840c0 1281 return (U8 *)start;
f9a63242 1282
1e54db1a 1283 /* ensure valid UTF-8 and chars < 256 before converting string */
f9a63242 1284 for (send = s + *len; s < send;) {
e1ec3a88 1285 U8 c = *s++;
1d72bdf6 1286 if (!UTF8_IS_INVARIANT(c)) {
db42d148
NIS
1287 if (UTF8_IS_DOWNGRADEABLE_START(c) && s < send &&
1288 (c = *s++) && UTF8_IS_CONTINUATION(c))
1289 count++;
1290 else
73d840c0 1291 return (U8 *)start;
db42d148 1292 }
f9a63242
JH
1293 }
1294
35da51f7 1295 *is_utf8 = FALSE;
f9a63242 1296
212542aa 1297 Newx(d, (*len) - count + 1, U8);
ef9edfd0 1298 s = start; start = d;
f9a63242
JH
1299 while (s < send) {
1300 U8 c = *s++;
c4d5f83a
NIS
1301 if (!UTF8_IS_INVARIANT(c)) {
1302 /* Then it is two-byte encoded */
356979f4 1303 c = UNI_TO_NATIVE(TWO_BYTE_UTF8_TO_UNI(c, *s++));
c4d5f83a
NIS
1304 }
1305 *d++ = c;
f9a63242
JH
1306 }
1307 *d = '\0';
1308 *len = d - start;
73d840c0 1309 return (U8 *)start;
f9a63242
JH
1310}
1311
1312/*
87cea99e 1313=for apidoc bytes_to_utf8
6940069f 1314
ff97e5cf
KW
1315Converts a string C<s> of length C<len> bytes from the native encoding into
1316UTF-8.
6662521e 1317Returns a pointer to the newly-created string, and sets C<len> to
ff97e5cf 1318reflect the new length in bytes.
6940069f 1319
2bbc8d55
SP
1320A NUL character will be written after the end of the string.
1321
1322If you want to convert to UTF-8 from encodings other than
1323the native (Latin1 or EBCDIC),
a1433954 1324see L</sv_recode_to_utf8>().
c9ada85f 1325
497711e7 1326=cut
6940069f
GS
1327*/
1328
c682ebef
FC
1329/* This logic is duplicated in sv_catpvn_flags, so any bug fixes will
1330 likewise need duplication. */
1331
6940069f 1332U8*
35a4481c 1333Perl_bytes_to_utf8(pTHX_ const U8 *s, STRLEN *len)
6940069f 1334{
35a4481c 1335 const U8 * const send = s + (*len);
6940069f
GS
1336 U8 *d;
1337 U8 *dst;
7918f24d
NC
1338
1339 PERL_ARGS_ASSERT_BYTES_TO_UTF8;
96a5add6 1340 PERL_UNUSED_CONTEXT;
6940069f 1341
212542aa 1342 Newx(d, (*len) * 2 + 1, U8);
6940069f
GS
1343 dst = d;
1344
1345 while (s < send) {
35a4481c 1346 const UV uv = NATIVE_TO_ASCII(*s++);
c4d5f83a 1347 if (UNI_IS_INVARIANT(uv))
eb160463 1348 *d++ = (U8)UTF_TO_NATIVE(uv);
6940069f 1349 else {
eb160463
GS
1350 *d++ = (U8)UTF8_EIGHT_BIT_HI(uv);
1351 *d++ = (U8)UTF8_EIGHT_BIT_LO(uv);
6940069f
GS
1352 }
1353 }
1354 *d = '\0';
6662521e 1355 *len = d-dst;
6940069f
GS
1356 return dst;
1357}
1358
a0ed51b3 1359/*
dea0fc0b 1360 * Convert native (big-endian) or reversed (little-endian) UTF-16 to UTF-8.
a0ed51b3
LW
1361 *
1362 * Destination must be pre-extended to 3/2 source. Do not use in-place.
1363 * We optimize for native, for obvious reasons. */
1364
1365U8*
dea0fc0b 1366Perl_utf16_to_utf8(pTHX_ U8* p, U8* d, I32 bytelen, I32 *newlen)
a0ed51b3 1367{
dea0fc0b
JH
1368 U8* pend;
1369 U8* dstart = d;
1370
7918f24d
NC
1371 PERL_ARGS_ASSERT_UTF16_TO_UTF8;
1372
dea0fc0b 1373 if (bytelen & 1)
f5992bc4 1374 Perl_croak(aTHX_ "panic: utf16_to_utf8: odd bytelen %"UVuf, (UV)bytelen);
dea0fc0b
JH
1375
1376 pend = p + bytelen;
1377
a0ed51b3 1378 while (p < pend) {
dea0fc0b
JH
1379 UV uv = (p[0] << 8) + p[1]; /* UTF-16BE */
1380 p += 2;
a0ed51b3 1381 if (uv < 0x80) {
e294cc5d
JH
1382#ifdef EBCDIC
1383 *d++ = UNI_TO_NATIVE(uv);
1384#else
eb160463 1385 *d++ = (U8)uv;
e294cc5d 1386#endif
a0ed51b3
LW
1387 continue;
1388 }
1389 if (uv < 0x800) {
eb160463
GS
1390 *d++ = (U8)(( uv >> 6) | 0xc0);
1391 *d++ = (U8)(( uv & 0x3f) | 0x80);
a0ed51b3
LW
1392 continue;
1393 }
52b9aa85 1394 if (uv >= 0xd800 && uv <= 0xdbff) { /* surrogates */
01ea242b 1395 if (p >= pend) {
dea0fc0b 1396 Perl_croak(aTHX_ "Malformed UTF-16 surrogate");
01ea242b
NC
1397 } else {
1398 UV low = (p[0] << 8) + p[1];
1399 p += 2;
52b9aa85 1400 if (low < 0xdc00 || low > 0xdfff)
01ea242b
NC
1401 Perl_croak(aTHX_ "Malformed UTF-16 surrogate");
1402 uv = ((uv - 0xd800) << 10) + (low - 0xdc00) + 0x10000;
1403 }
dbde1951
NC
1404 } else if (uv >= 0xdc00 && uv <= 0xdfff) {
1405 Perl_croak(aTHX_ "Malformed UTF-16 surrogate");
a0ed51b3
LW
1406 }
1407 if (uv < 0x10000) {
eb160463
GS
1408 *d++ = (U8)(( uv >> 12) | 0xe0);
1409 *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
1410 *d++ = (U8)(( uv & 0x3f) | 0x80);
a0ed51b3
LW
1411 continue;
1412 }
1413 else {
eb160463
GS
1414 *d++ = (U8)(( uv >> 18) | 0xf0);
1415 *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80);
1416 *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
1417 *d++ = (U8)(( uv & 0x3f) | 0x80);
a0ed51b3
LW
1418 continue;
1419 }
1420 }
dea0fc0b 1421 *newlen = d - dstart;
a0ed51b3
LW
1422 return d;
1423}
1424
1425/* Note: this one is slightly destructive of the source. */
1426
1427U8*
dea0fc0b 1428Perl_utf16_to_utf8_reversed(pTHX_ U8* p, U8* d, I32 bytelen, I32 *newlen)
a0ed51b3
LW
1429{
1430 U8* s = (U8*)p;
d4c19fe8 1431 U8* const send = s + bytelen;
7918f24d
NC
1432
1433 PERL_ARGS_ASSERT_UTF16_TO_UTF8_REVERSED;
1434
e0ea5e2d
NC
1435 if (bytelen & 1)
1436 Perl_croak(aTHX_ "panic: utf16_to_utf8_reversed: odd bytelen %"UVuf,
1437 (UV)bytelen);
1438
a0ed51b3 1439 while (s < send) {
d4c19fe8 1440 const U8 tmp = s[0];
a0ed51b3
LW
1441 s[0] = s[1];
1442 s[1] = tmp;
1443 s += 2;
1444 }
dea0fc0b 1445 return utf16_to_utf8(p, d, bytelen, newlen);
a0ed51b3
LW
1446}
1447
c3fd2246
KW
1448/* for now these are all defined (inefficiently) in terms of the utf8 versions.
1449 * Note that the macros in handy.h that call these short-circuit calling them
1450 * for Latin-1 range inputs */
a0ed51b3
LW
1451
1452bool
84afefe6 1453Perl_is_uni_alnum(pTHX_ UV c)
a0ed51b3 1454{
89ebb4a3 1455 U8 tmpbuf[UTF8_MAXBYTES+1];
230880c1 1456 uvchr_to_utf8(tmpbuf, c);
a0ed51b3
LW
1457 return is_utf8_alnum(tmpbuf);
1458}
1459
1460bool
84afefe6 1461Perl_is_uni_idfirst(pTHX_ UV c)
a0ed51b3 1462{
89ebb4a3 1463 U8 tmpbuf[UTF8_MAXBYTES+1];
230880c1 1464 uvchr_to_utf8(tmpbuf, c);
a0ed51b3
LW
1465 return is_utf8_idfirst(tmpbuf);
1466}
1467
1468bool
84afefe6 1469Perl_is_uni_alpha(pTHX_ UV c)
a0ed51b3 1470{
89ebb4a3 1471 U8 tmpbuf[UTF8_MAXBYTES+1];
230880c1 1472 uvchr_to_utf8(tmpbuf, c);
a0ed51b3
LW
1473 return is_utf8_alpha(tmpbuf);
1474}
1475
1476bool
84afefe6 1477Perl_is_uni_ascii(pTHX_ UV c)
4d61ec05 1478{
bc39fe24 1479 return isASCII(c);
4d61ec05
GS
1480}
1481
1482bool
84afefe6 1483Perl_is_uni_space(pTHX_ UV c)
a0ed51b3 1484{
89ebb4a3 1485 U8 tmpbuf[UTF8_MAXBYTES+1];
230880c1 1486 uvchr_to_utf8(tmpbuf, c);
a0ed51b3
LW
1487 return is_utf8_space(tmpbuf);
1488}
1489
1490bool
84afefe6 1491Perl_is_uni_digit(pTHX_ UV c)
a0ed51b3 1492{
89ebb4a3 1493 U8 tmpbuf[UTF8_MAXBYTES+1];
230880c1 1494 uvchr_to_utf8(tmpbuf, c);
a0ed51b3
LW
1495 return is_utf8_digit(tmpbuf);
1496}
1497
1498bool
84afefe6 1499Perl_is_uni_upper(pTHX_ UV c)
a0ed51b3 1500{
89ebb4a3 1501 U8 tmpbuf[UTF8_MAXBYTES+1];
230880c1 1502 uvchr_to_utf8(tmpbuf, c);
a0ed51b3
LW
1503 return is_utf8_upper(tmpbuf);
1504}
1505
1506bool
84afefe6 1507Perl_is_uni_lower(pTHX_ UV c)
a0ed51b3 1508{
89ebb4a3 1509 U8 tmpbuf[UTF8_MAXBYTES+1];
230880c1 1510 uvchr_to_utf8(tmpbuf, c);
a0ed51b3
LW
1511 return is_utf8_lower(tmpbuf);
1512}
1513
1514bool
84afefe6 1515Perl_is_uni_cntrl(pTHX_ UV c)
b8c5462f 1516{
7b952154 1517 return isCNTRL_L1(c);
b8c5462f
JH
1518}
1519
1520bool
84afefe6 1521Perl_is_uni_graph(pTHX_ UV c)
b8c5462f 1522{
89ebb4a3 1523 U8 tmpbuf[UTF8_MAXBYTES+1];
230880c1 1524 uvchr_to_utf8(tmpbuf, c);
b8c5462f
JH
1525 return is_utf8_graph(tmpbuf);
1526}
1527
1528bool
84afefe6 1529Perl_is_uni_print(pTHX_ UV c)
a0ed51b3 1530{
89ebb4a3 1531 U8 tmpbuf[UTF8_MAXBYTES+1];
230880c1 1532 uvchr_to_utf8(tmpbuf, c);
a0ed51b3
LW
1533 return is_utf8_print(tmpbuf);
1534}
1535
b8c5462f 1536bool
84afefe6 1537Perl_is_uni_punct(pTHX_ UV c)
b8c5462f 1538{
89ebb4a3 1539 U8 tmpbuf[UTF8_MAXBYTES+1];
230880c1 1540 uvchr_to_utf8(tmpbuf, c);
b8c5462f
JH
1541 return is_utf8_punct(tmpbuf);
1542}
1543
4d61ec05 1544bool
84afefe6 1545Perl_is_uni_xdigit(pTHX_ UV c)
4d61ec05 1546{
89ebb4a3 1547 U8 tmpbuf[UTF8_MAXBYTES_CASE+1];
230880c1 1548 uvchr_to_utf8(tmpbuf, c);
4d61ec05
GS
1549 return is_utf8_xdigit(tmpbuf);
1550}
1551
3a4c58c9
KW
1552UV
1553Perl__to_upper_title_latin1(pTHX_ const U8 c, U8* p, STRLEN *lenp, const char S_or_s)
1554{
1555 /* We have the latin1-range values compiled into the core, so just use
1556 * those, converting the result to utf8. The only difference between upper
1557 * and title case in this range is that LATIN_SMALL_LETTER_SHARP_S is
1558 * either "SS" or "Ss". Which one to use is passed into the routine in
1559 * 'S_or_s' to avoid a test */
1560
1561 UV converted = toUPPER_LATIN1_MOD(c);
1562
1563 PERL_ARGS_ASSERT__TO_UPPER_TITLE_LATIN1;
1564
1565 assert(S_or_s == 'S' || S_or_s == 's');
1566
1567 if (UNI_IS_INVARIANT(converted)) { /* No difference between the two for
1568 characters in this range */
1569 *p = (U8) converted;
1570 *lenp = 1;
1571 return converted;
1572 }
1573
1574 /* toUPPER_LATIN1_MOD gives the correct results except for three outliers,
1575 * which it maps to one of them, so as to only have to have one check for
1576 * it in the main case */
1577 if (UNLIKELY(converted == LATIN_SMALL_LETTER_Y_WITH_DIAERESIS)) {
1578 switch (c) {
1579 case LATIN_SMALL_LETTER_Y_WITH_DIAERESIS:
1580 converted = LATIN_CAPITAL_LETTER_Y_WITH_DIAERESIS;
1581 break;
1582 case MICRO_SIGN:
1583 converted = GREEK_CAPITAL_LETTER_MU;
1584 break;
1585 case LATIN_SMALL_LETTER_SHARP_S:
1586 *(p)++ = 'S';
1587 *p = S_or_s;
1588 *lenp = 2;
1589 return 'S';
1590 default:
1591 Perl_croak(aTHX_ "panic: to_upper_title_latin1 did not expect '%c' to map to '%c'", c, LATIN_SMALL_LETTER_Y_WITH_DIAERESIS);
1592 /* NOTREACHED */
1593 }
1594 }
1595
1596 *(p)++ = UTF8_TWO_BYTE_HI(converted);
1597 *p = UTF8_TWO_BYTE_LO(converted);
1598 *lenp = 2;
1599
1600 return converted;
1601}
1602
50bda2c3
KW
1603/* Call the function to convert a UTF-8 encoded character to the specified case.
1604 * Note that there may be more than one character in the result.
1605 * INP is a pointer to the first byte of the input character
1606 * OUTP will be set to the first byte of the string of changed characters. It
1607 * needs to have space for UTF8_MAXBYTES_CASE+1 bytes
1608 * LENP will be set to the length in bytes of the string of changed characters
1609 *
1610 * The functions return the ordinal of the first character in the string of OUTP */
f90a9a02
KW
1611#define CALL_UPPER_CASE(INP, OUTP, LENP) Perl_to_utf8_case(aTHX_ INP, OUTP, LENP, &PL_utf8_toupper, "ToUc", "utf8::ToSpecUc")
1612#define CALL_TITLE_CASE(INP, OUTP, LENP) Perl_to_utf8_case(aTHX_ INP, OUTP, LENP, &PL_utf8_totitle, "ToTc", "utf8::ToSpecTc")
1613#define CALL_LOWER_CASE(INP, OUTP, LENP) Perl_to_utf8_case(aTHX_ INP, OUTP, LENP, &PL_utf8_tolower, "ToLc", "utf8::ToSpecLc")
50bda2c3
KW
1614
1615/* This additionally has the input parameter SPECIALS, which if non-zero will
1616 * cause this to use the SPECIALS hash for folding (meaning get full case
1617 * folding); otherwise, when zero, this implies a simple case fold */
f90a9a02 1618#define CALL_FOLD_CASE(INP, OUTP, LENP, SPECIALS) Perl_to_utf8_case(aTHX_ INP, OUTP, LENP, &PL_utf8_tofold, "ToCf", (SPECIALS) ? "utf8::ToSpecCf" : NULL)
c3fd2246 1619
84afefe6
JH
1620UV
1621Perl_to_uni_upper(pTHX_ UV c, U8* p, STRLEN *lenp)
a0ed51b3 1622{
3a4c58c9
KW
1623 dVAR;
1624
a1433954
KW
1625 /* Convert the Unicode character whose ordinal is <c> to its uppercase
1626 * version and store that in UTF-8 in <p> and its length in bytes in <lenp>.
1627 * Note that the <p> needs to be at least UTF8_MAXBYTES_CASE+1 bytes since
c3fd2246
KW
1628 * the changed version may be longer than the original character.
1629 *
1630 * The ordinal of the first character of the changed version is returned
1631 * (but note, as explained above, that there may be more.) */
1632
7918f24d
NC
1633 PERL_ARGS_ASSERT_TO_UNI_UPPER;
1634
3a4c58c9
KW
1635 if (c < 256) {
1636 return _to_upper_title_latin1((U8) c, p, lenp, 'S');
1637 }
1638
0ebc6274 1639 uvchr_to_utf8(p, c);
3a4c58c9 1640 return CALL_UPPER_CASE(p, p, lenp);
a0ed51b3
LW
1641}
1642
84afefe6
JH
1643UV
1644Perl_to_uni_title(pTHX_ UV c, U8* p, STRLEN *lenp)
a0ed51b3 1645{
3a4c58c9
KW
1646 dVAR;
1647
7918f24d
NC
1648 PERL_ARGS_ASSERT_TO_UNI_TITLE;
1649
3a4c58c9
KW
1650 if (c < 256) {
1651 return _to_upper_title_latin1((U8) c, p, lenp, 's');
1652 }
1653
0ebc6274 1654 uvchr_to_utf8(p, c);
3a4c58c9 1655 return CALL_TITLE_CASE(p, p, lenp);
a0ed51b3
LW
1656}
1657
afc16117
KW
1658STATIC U8
1659S_to_lower_latin1(pTHX_ const U8 c, U8* p, STRLEN *lenp)
1660{
1661 /* We have the latin1-range values compiled into the core, so just use
1662 * those, converting the result to utf8. Since the result is always just
a1433954 1663 * one character, we allow <p> to be NULL */
afc16117
KW
1664
1665 U8 converted = toLOWER_LATIN1(c);
1666
1667 if (p != NULL) {
1668 if (UNI_IS_INVARIANT(converted)) {
1669 *p = converted;
1670 *lenp = 1;
1671 }
1672 else {
1673 *p = UTF8_TWO_BYTE_HI(converted);
1674 *(p+1) = UTF8_TWO_BYTE_LO(converted);
1675 *lenp = 2;
1676 }
1677 }
1678 return converted;
1679}
1680
84afefe6
JH
1681UV
1682Perl_to_uni_lower(pTHX_ UV c, U8* p, STRLEN *lenp)
a0ed51b3 1683{
968c5e6a
KW
1684 dVAR;
1685
7918f24d
NC
1686 PERL_ARGS_ASSERT_TO_UNI_LOWER;
1687
afc16117
KW
1688 if (c < 256) {
1689 return to_lower_latin1((U8) c, p, lenp);
bca00c02
KW
1690 }
1691
afc16117 1692 uvchr_to_utf8(p, c);
968c5e6a 1693 return CALL_LOWER_CASE(p, p, lenp);
a0ed51b3
LW
1694}
1695
84afefe6 1696UV
f673fad4 1697Perl__to_fold_latin1(pTHX_ const U8 c, U8* p, STRLEN *lenp, const bool flags)
a1dde8de 1698{
a1433954 1699 /* Corresponds to to_lower_latin1(), <flags> is TRUE if to use full case
f673fad4
KW
1700 * folding */
1701
a1dde8de
KW
1702 UV converted;
1703
1704 PERL_ARGS_ASSERT__TO_FOLD_LATIN1;
1705
1706 if (c == MICRO_SIGN) {
1707 converted = GREEK_SMALL_LETTER_MU;
1708 }
1709 else if (flags && c == LATIN_SMALL_LETTER_SHARP_S) {
1710 *(p)++ = 's';
1711 *p = 's';
1712 *lenp = 2;
1713 return 's';
1714 }
1715 else { /* In this range the fold of all other characters is their lower
1716 case */
1717 converted = toLOWER_LATIN1(c);
1718 }
1719
1720 if (UNI_IS_INVARIANT(converted)) {
1721 *p = (U8) converted;
1722 *lenp = 1;
1723 }
1724 else {
1725 *(p)++ = UTF8_TWO_BYTE_HI(converted);
1726 *p = UTF8_TWO_BYTE_LO(converted);
1727 *lenp = 2;
1728 }
1729
1730 return converted;
1731}
1732
1733UV
f673fad4 1734Perl__to_uni_fold_flags(pTHX_ UV c, U8* p, STRLEN *lenp, const bool flags)
84afefe6 1735{
4b593389
KW
1736
1737 /* Not currently externally documented, and subject to change, <flags> is
f673fad4 1738 * TRUE iff full folding is to be used */
4b593389 1739
36bb2ab6 1740 PERL_ARGS_ASSERT__TO_UNI_FOLD_FLAGS;
7918f24d 1741
a1dde8de
KW
1742 if (c < 256) {
1743 return _to_fold_latin1((U8) c, p, lenp, flags);
1744 }
1745
0ebc6274 1746 uvchr_to_utf8(p, c);
a1dde8de 1747 return CALL_FOLD_CASE(p, p, lenp, flags);
84afefe6
JH
1748}
1749
ea317ccb
KW
1750/* for now these all assume no locale info available for Unicode > 255; and
1751 * the corresponding macros in handy.h (like isALNUM_LC_uvchr) should have been
1752 * called instead, so that these don't get called for < 255 */
a0ed51b3
LW
1753
1754bool
84afefe6 1755Perl_is_uni_alnum_lc(pTHX_ UV c)
a0ed51b3
LW
1756{
1757 return is_uni_alnum(c); /* XXX no locale support yet */
1758}
1759
1760bool
84afefe6 1761Perl_is_uni_idfirst_lc(pTHX_ UV c)
a0ed51b3
LW
1762{
1763 return is_uni_idfirst(c); /* XXX no locale support yet */
1764}
1765
1766bool
84afefe6 1767Perl_is_uni_alpha_lc(pTHX_ UV c)
a0ed51b3
LW
1768{
1769 return is_uni_alpha(c); /* XXX no locale support yet */
1770}
1771
1772bool
84afefe6 1773Perl_is_uni_ascii_lc(pTHX_ UV c)
4d61ec05
GS
1774{
1775 return is_uni_ascii(c); /* XXX no locale support yet */
1776}
1777
1778bool
84afefe6 1779Perl_is_uni_space_lc(pTHX_ UV c)
a0ed51b3
LW
1780{
1781 return is_uni_space(c); /* XXX no locale support yet */
1782}
1783
1784bool
84afefe6 1785Perl_is_uni_digit_lc(pTHX_ UV c)
a0ed51b3
LW
1786{
1787 return is_uni_digit(c); /* XXX no locale support yet */
1788}
1789
1790bool
84afefe6 1791Perl_is_uni_upper_lc(pTHX_ UV c)
a0ed51b3
LW
1792{
1793 return is_uni_upper(c); /* XXX no locale support yet */
1794}
1795
1796bool
84afefe6 1797Perl_is_uni_lower_lc(pTHX_ UV c)
a0ed51b3
LW
1798{
1799 return is_uni_lower(c); /* XXX no locale support yet */
1800}
1801
1802bool
84afefe6 1803Perl_is_uni_cntrl_lc(pTHX_ UV c)
b8c5462f
JH
1804{
1805 return is_uni_cntrl(c); /* XXX no locale support yet */
1806}
1807
1808bool
84afefe6 1809Perl_is_uni_graph_lc(pTHX_ UV c)
b8c5462f
JH
1810{
1811 return is_uni_graph(c); /* XXX no locale support yet */
1812}
1813
1814bool
84afefe6 1815Perl_is_uni_print_lc(pTHX_ UV c)
a0ed51b3
LW
1816{
1817 return is_uni_print(c); /* XXX no locale support yet */
1818}
1819
b8c5462f 1820bool
84afefe6 1821Perl_is_uni_punct_lc(pTHX_ UV c)
b8c5462f
JH
1822{
1823 return is_uni_punct(c); /* XXX no locale support yet */
1824}
1825
4d61ec05 1826bool
84afefe6 1827Perl_is_uni_xdigit_lc(pTHX_ UV c)
4d61ec05
GS
1828{
1829 return is_uni_xdigit(c); /* XXX no locale support yet */
1830}
1831
b7ac61fa
JH
1832U32
1833Perl_to_uni_upper_lc(pTHX_ U32 c)
1834{
ee099d14
JH
1835 /* XXX returns only the first character -- do not use XXX */
1836 /* XXX no locale support yet */
1837 STRLEN len;
89ebb4a3 1838 U8 tmpbuf[UTF8_MAXBYTES_CASE+1];
ee099d14 1839 return (U32)to_uni_upper(c, tmpbuf, &len);
b7ac61fa
JH
1840}
1841
1842U32
1843Perl_to_uni_title_lc(pTHX_ U32 c)
1844{
ee099d14
JH
1845 /* XXX returns only the first character XXX -- do not use XXX */
1846 /* XXX no locale support yet */
1847 STRLEN len;
89ebb4a3 1848 U8 tmpbuf[UTF8_MAXBYTES_CASE+1];
ee099d14 1849 return (U32)to_uni_title(c, tmpbuf, &len);
b7ac61fa
JH
1850}
1851
1852U32
1853Perl_to_uni_lower_lc(pTHX_ U32 c)
1854{
ee099d14
JH
1855 /* XXX returns only the first character -- do not use XXX */
1856 /* XXX no locale support yet */
1857 STRLEN len;
89ebb4a3 1858 U8 tmpbuf[UTF8_MAXBYTES_CASE+1];
ee099d14 1859 return (U32)to_uni_lower(c, tmpbuf, &len);
b7ac61fa
JH
1860}
1861
7452cf6a 1862static bool
5141f98e 1863S_is_utf8_common(pTHX_ const U8 *const p, SV **swash,
bde6a22d
NC
1864 const char *const swashname)
1865{
ea317ccb
KW
1866 /* returns a boolean giving whether or not the UTF8-encoded character that
1867 * starts at <p> is in the swash indicated by <swashname>. <swash>
1868 * contains a pointer to where the swash indicated by <swashname>
1869 * is to be stored; which this routine will do, so that future calls will
1870 * look at <*swash> and only generate a swash if it is not null
1871 *
1872 * Note that it is assumed that the buffer length of <p> is enough to
1873 * contain all the bytes that comprise the character. Thus, <*p> should
1874 * have been checked before this call for mal-formedness enough to assure
1875 * that. */
1876
97aff369 1877 dVAR;
7918f24d
NC
1878
1879 PERL_ARGS_ASSERT_IS_UTF8_COMMON;
1880
492a624f
KW
1881 /* The API should have included a length for the UTF-8 character in <p>,
1882 * but it doesn't. We therefor assume that p has been validated at least
1883 * as far as there being enough bytes available in it to accommodate the
1884 * character without reading beyond the end, and pass that number on to the
1885 * validating routine */
1886 if (!is_utf8_char_buf(p, p + UTF8SKIP(p)))
bde6a22d
NC
1887 return FALSE;
1888 if (!*swash)
711a919c 1889 *swash = swash_init("utf8", swashname, &PL_sv_undef, 1, 0);
bde6a22d
NC
1890 return swash_fetch(*swash, p, TRUE) != 0;
1891}
1892
1893bool
7fc63493 1894Perl_is_utf8_alnum(pTHX_ const U8 *p)
a0ed51b3 1895{
97aff369 1896 dVAR;
7918f24d
NC
1897
1898 PERL_ARGS_ASSERT_IS_UTF8_ALNUM;
1899
671c33bf
NC
1900 /* NOTE: "IsWord", not "IsAlnum", since Alnum is a true
1901 * descendant of isalnum(3), in other words, it doesn't
1902 * contain the '_'. --jhi */
d4c19fe8 1903 return is_utf8_common(p, &PL_utf8_alnum, "IsWord");
a0ed51b3
LW
1904}
1905
1906bool
7fc63493 1907Perl_is_utf8_idfirst(pTHX_ const U8 *p) /* The naming is historical. */
a0ed51b3 1908{
97aff369 1909 dVAR;
7918f24d
NC
1910
1911 PERL_ARGS_ASSERT_IS_UTF8_IDFIRST;
1912
82686b01
JH
1913 if (*p == '_')
1914 return TRUE;
bde6a22d 1915 /* is_utf8_idstart would be more logical. */
d4c19fe8 1916 return is_utf8_common(p, &PL_utf8_idstart, "IdStart");
82686b01
JH
1917}
1918
1919bool
c11ff943
KW
1920Perl_is_utf8_xidfirst(pTHX_ const U8 *p) /* The naming is historical. */
1921{
1922 dVAR;
1923
1924 PERL_ARGS_ASSERT_IS_UTF8_XIDFIRST;
1925
1926 if (*p == '_')
1927 return TRUE;
1928 /* is_utf8_idstart would be more logical. */
1929 return is_utf8_common(p, &PL_utf8_xidstart, "XIdStart");
1930}
1931
1932bool
b6912c02
KW
1933Perl__is_utf8__perl_idstart(pTHX_ const U8 *p)
1934{
1935 dVAR;
1936
1937 PERL_ARGS_ASSERT__IS_UTF8__PERL_IDSTART;
1938
1939 return is_utf8_common(p, &PL_utf8_perl_idstart, "_Perl_IDStart");
1940}
1941
1942bool
7fc63493 1943Perl_is_utf8_idcont(pTHX_ const U8 *p)
82686b01 1944{
97aff369 1945 dVAR;
7918f24d
NC
1946
1947 PERL_ARGS_ASSERT_IS_UTF8_IDCONT;
1948
d4c19fe8 1949 return is_utf8_common(p, &PL_utf8_idcont, "IdContinue");
a0ed51b3
LW
1950}
1951
1952bool
c11ff943
KW
1953Perl_is_utf8_xidcont(pTHX_ const U8 *p)
1954{
1955 dVAR;
1956
1957 PERL_ARGS_ASSERT_IS_UTF8_XIDCONT;
1958
c11ff943
KW
1959 return is_utf8_common(p, &PL_utf8_idcont, "XIdContinue");
1960}
1961
1962bool
7fc63493 1963Perl_is_utf8_alpha(pTHX_ const U8 *p)
a0ed51b3 1964{
97aff369 1965 dVAR;
7918f24d
NC
1966
1967 PERL_ARGS_ASSERT_IS_UTF8_ALPHA;
1968
d4c19fe8 1969 return is_utf8_common(p, &PL_utf8_alpha, "IsAlpha");
a0ed51b3
LW
1970}
1971
1972bool
7fc63493 1973Perl_is_utf8_ascii(pTHX_ const U8 *p)
b8c5462f 1974{
97aff369 1975 dVAR;
7918f24d
NC
1976
1977 PERL_ARGS_ASSERT_IS_UTF8_ASCII;
1978
bc39fe24
KW
1979 /* ASCII characters are the same whether in utf8 or not. So the macro
1980 * works on both utf8 and non-utf8 representations. */
1981 return isASCII(*p);
b8c5462f
JH
1982}
1983
1984bool
7fc63493 1985Perl_is_utf8_space(pTHX_ const U8 *p)
a0ed51b3 1986{
97aff369 1987 dVAR;
7918f24d
NC
1988
1989 PERL_ARGS_ASSERT_IS_UTF8_SPACE;
1990
a34094a9 1991 return is_utf8_common(p, &PL_utf8_space, "IsXPerlSpace");
a0ed51b3
LW
1992}
1993
1994bool
d1eb3177
YO
1995Perl_is_utf8_perl_space(pTHX_ const U8 *p)
1996{
1997 dVAR;
1998
1999 PERL_ARGS_ASSERT_IS_UTF8_PERL_SPACE;
2000
c4428693
KW
2001 /* Only true if is an ASCII space-like character, and ASCII is invariant
2002 * under utf8, so can just use the macro */
2003 return isSPACE_A(*p);
d1eb3177
YO
2004}
2005
2006bool
2007Perl_is_utf8_perl_word(pTHX_ const U8 *p)
2008{
2009 dVAR;
2010
2011 PERL_ARGS_ASSERT_IS_UTF8_PERL_WORD;
2012
c4428693
KW
2013 /* Only true if is an ASCII word character, and ASCII is invariant
2014 * under utf8, so can just use the macro */
2015 return isWORDCHAR_A(*p);
d1eb3177
YO
2016}
2017
2018bool
7fc63493 2019Perl_is_utf8_digit(pTHX_ const U8 *p)
a0ed51b3 2020{
97aff369 2021 dVAR;
7918f24d
NC
2022
2023 PERL_ARGS_ASSERT_IS_UTF8_DIGIT;
2024
d4c19fe8 2025 return is_utf8_common(p, &PL_utf8_digit, "IsDigit");
a0ed51b3
LW
2026}
2027
2028bool
d1eb3177
YO
2029Perl_is_utf8_posix_digit(pTHX_ const U8 *p)
2030{
2031 dVAR;
2032
2033 PERL_ARGS_ASSERT_IS_UTF8_POSIX_DIGIT;
2034
c4428693
KW
2035 /* Only true if is an ASCII digit character, and ASCII is invariant
2036 * under utf8, so can just use the macro */
2037 return isDIGIT_A(*p);
d1eb3177
YO
2038}
2039
2040bool
7fc63493 2041Perl_is_utf8_upper(pTHX_ const U8 *p)
a0ed51b3 2042{
97aff369 2043 dVAR;
7918f24d
NC
2044
2045 PERL_ARGS_ASSERT_IS_UTF8_UPPER;
2046
d4c19fe8 2047 return is_utf8_common(p, &PL_utf8_upper, "IsUppercase");
a0ed51b3
LW
2048}
2049
2050bool
7fc63493 2051Perl_is_utf8_lower(pTHX_ const U8 *p)
a0ed51b3 2052{
97aff369 2053 dVAR;
7918f24d
NC
2054
2055 PERL_ARGS_ASSERT_IS_UTF8_LOWER;
2056
d4c19fe8 2057 return is_utf8_common(p, &PL_utf8_lower, "IsLowercase");
a0ed51b3
LW
2058}
2059
2060bool
7fc63493 2061Perl_is_utf8_cntrl(pTHX_ const U8 *p)
b8c5462f 2062{
97aff369 2063 dVAR;
7918f24d
NC
2064
2065 PERL_ARGS_ASSERT_IS_UTF8_CNTRL;
2066
7b952154
KW
2067 if (isASCII(*p)) {
2068 return isCNTRL_A(*p);
2069 }
2070
2071 /* All controls are in Latin1 */
2072 if (! UTF8_IS_DOWNGRADEABLE_START(*p)) {
2073 return 0;
2074 }
2075 return isCNTRL_L1(TWO_BYTE_UTF8_TO_UNI(*p, *(p+1)));
b8c5462f
JH
2076}
2077
2078bool
7fc63493 2079Perl_is_utf8_graph(pTHX_ const U8 *p)
b8c5462f 2080{
97aff369 2081 dVAR;
7918f24d
NC
2082
2083 PERL_ARGS_ASSERT_IS_UTF8_GRAPH;
2084
d4c19fe8 2085 return is_utf8_common(p, &PL_utf8_graph, "IsGraph");
b8c5462f
JH
2086}
2087
2088bool
7fc63493 2089Perl_is_utf8_print(pTHX_ const U8 *p)
a0ed51b3 2090{
97aff369 2091 dVAR;
7918f24d
NC
2092
2093 PERL_ARGS_ASSERT_IS_UTF8_PRINT;
2094
d4c19fe8 2095 return is_utf8_common(p, &PL_utf8_print, "IsPrint");
a0ed51b3
LW
2096}
2097
2098bool
7fc63493 2099Perl_is_utf8_punct(pTHX_ const U8 *p)
b8c5462f 2100{
97aff369 2101 dVAR;
7918f24d
NC
2102
2103 PERL_ARGS_ASSERT_IS_UTF8_PUNCT;
2104
d4c19fe8 2105 return is_utf8_common(p, &PL_utf8_punct, "IsPunct");
b8c5462f
JH
2106}
2107
2108bool
7fc63493 2109Perl_is_utf8_xdigit(pTHX_ const U8 *p)
b8c5462f 2110{
97aff369 2111 dVAR;
7918f24d
NC
2112
2113 PERL_ARGS_ASSERT_IS_UTF8_XDIGIT;
2114
d1eb3177 2115 return is_utf8_common(p, &PL_utf8_xdigit, "IsXDigit");
b8c5462f
JH
2116}
2117
2118bool
7fc63493 2119Perl_is_utf8_mark(pTHX_ const U8 *p)
a0ed51b3 2120{
97aff369 2121 dVAR;
7918f24d
NC
2122
2123 PERL_ARGS_ASSERT_IS_UTF8_MARK;
2124
d4c19fe8 2125 return is_utf8_common(p, &PL_utf8_mark, "IsM");
a0ed51b3
LW
2126}
2127
37e2e78e
KW
2128bool
2129Perl_is_utf8_X_begin(pTHX_ const U8 *p)
2130{
2131 dVAR;
2132
2133 PERL_ARGS_ASSERT_IS_UTF8_X_BEGIN;
2134
2135 return is_utf8_common(p, &PL_utf8_X_begin, "_X_Begin");
2136}
2137
2138bool
2139Perl_is_utf8_X_extend(pTHX_ const U8 *p)
2140{
2141 dVAR;
2142
2143 PERL_ARGS_ASSERT_IS_UTF8_X_EXTEND;
2144
2145 return is_utf8_common(p, &PL_utf8_X_extend, "_X_Extend");
2146}
2147
2148bool
2149Perl_is_utf8_X_prepend(pTHX_ const U8 *p)
2150{
2151 dVAR;
2152
2153 PERL_ARGS_ASSERT_IS_UTF8_X_PREPEND;
2154
2155 return is_utf8_common(p, &PL_utf8_X_prepend, "GCB=Prepend");
2156}
2157
2158bool
2159Perl_is_utf8_X_non_hangul(pTHX_ const U8 *p)
2160{
2161 dVAR;
2162
2163 PERL_ARGS_ASSERT_IS_UTF8_X_NON_HANGUL;
2164
2165 return is_utf8_common(p, &PL_utf8_X_non_hangul, "HST=Not_Applicable");
2166}
2167
2168bool
2169Perl_is_utf8_X_L(pTHX_ const U8 *p)
2170{
2171 dVAR;
2172
2173 PERL_ARGS_ASSERT_IS_UTF8_X_L;
2174
2175 return is_utf8_common(p, &PL_utf8_X_L, "GCB=L");
2176}
2177
2178bool
2179Perl_is_utf8_X_LV(pTHX_ const U8 *p)
2180{
2181 dVAR;
2182
2183 PERL_ARGS_ASSERT_IS_UTF8_X_LV;
2184
2185 return is_utf8_common(p, &PL_utf8_X_LV, "GCB=LV");
2186}
2187
2188bool
2189Perl_is_utf8_X_LVT(pTHX_ const U8 *p)
2190{
2191 dVAR;
2192
2193 PERL_ARGS_ASSERT_IS_UTF8_X_LVT;
2194
2195 return is_utf8_common(p, &PL_utf8_X_LVT, "GCB=LVT");
2196}
2197
2198bool
2199Perl_is_utf8_X_T(pTHX_ const U8 *p)
2200{
2201 dVAR;
2202
2203 PERL_ARGS_ASSERT_IS_UTF8_X_T;
2204
2205 return is_utf8_common(p, &PL_utf8_X_T, "GCB=T");
2206}
2207
2208bool
2209Perl_is_utf8_X_V(pTHX_ const U8 *p)
2210{
2211 dVAR;
2212
2213 PERL_ARGS_ASSERT_IS_UTF8_X_V;
2214
2215 return is_utf8_common(p, &PL_utf8_X_V, "GCB=V");
2216}
2217
2218bool
2219Perl_is_utf8_X_LV_LVT_V(pTHX_ const U8 *p)
2220{
2221 dVAR;
2222
2223 PERL_ARGS_ASSERT_IS_UTF8_X_LV_LVT_V;
2224
2225 return is_utf8_common(p, &PL_utf8_X_LV_LVT_V, "_X_LV_LVT_V");
2226}
2227
2e2b2571
KW
2228bool
2229Perl__is_utf8_quotemeta(pTHX_ const U8 *p)
2230{
2231 /* For exclusive use of pp_quotemeta() */
2232
2233 dVAR;
2234
2235 PERL_ARGS_ASSERT__IS_UTF8_QUOTEMETA;
2236
2237 return is_utf8_common(p, &PL_utf8_quotemeta, "_Perl_Quotemeta");
2238}
2239
6b5c0936 2240/*
87cea99e 2241=for apidoc to_utf8_case
6b5c0936 2242
a1433954
KW
2243The C<p> contains the pointer to the UTF-8 string encoding
2244the character that is being converted. This routine assumes that the character
2245at C<p> is well-formed.
6b5c0936 2246
a1433954
KW
2247The C<ustrp> is a pointer to the character buffer to put the
2248conversion result to. The C<lenp> is a pointer to the length
6b5c0936
JH
2249of the result.
2250
a1433954 2251The C<swashp> is a pointer to the swash to use.
6b5c0936 2252
a1433954
KW
2253Both the special and normal mappings are stored in F<lib/unicore/To/Foo.pl>,
2254and loaded by SWASHNEW, using F<lib/utf8_heavy.pl>. The C<special> (usually,
0134edef 2255but not always, a multicharacter mapping), is tried first.
6b5c0936 2256
a1433954 2257The C<special> is a string like "utf8::ToSpecLower", which means the
0134edef
JH
2258hash %utf8::ToSpecLower. The access to the hash is through
2259Perl_to_utf8_case().
6b5c0936 2260
a1433954 2261The C<normal> is a string like "ToLower" which means the swash
0134edef
JH
2262%utf8::ToLower.
2263
2264=cut */
6b5c0936 2265
2104c8d9 2266UV
9a957fbc
AL
2267Perl_to_utf8_case(pTHX_ const U8 *p, U8* ustrp, STRLEN *lenp,
2268 SV **swashp, const char *normal, const char *special)
a0ed51b3 2269{
97aff369 2270 dVAR;
89ebb4a3 2271 U8 tmpbuf[UTF8_MAXBYTES_CASE+1];
0134edef 2272 STRLEN len = 0;
4b88fb76 2273 const UV uv0 = valid_utf8_to_uvchr(p, NULL);
1feea2c7
JH
2274 /* The NATIVE_TO_UNI() and UNI_TO_NATIVE() mappings
2275 * are necessary in EBCDIC, they are redundant no-ops
2276 * in ASCII-ish platforms, and hopefully optimized away. */
f54cb97a 2277 const UV uv1 = NATIVE_TO_UNI(uv0);
7918f24d
NC
2278
2279 PERL_ARGS_ASSERT_TO_UTF8_CASE;
2280
9ae3ac1a
KW
2281 /* Note that swash_fetch() doesn't output warnings for these because it
2282 * assumes we will */
8457b38f 2283 if (uv1 >= UNICODE_SURROGATE_FIRST) {
9ae3ac1a 2284 if (uv1 <= UNICODE_SURROGATE_LAST) {
8457b38f
KW
2285 if (ckWARN_d(WARN_SURROGATE)) {
2286 const char* desc = (PL_op) ? OP_DESC(PL_op) : normal;
2287 Perl_warner(aTHX_ packWARN(WARN_SURROGATE),
2288 "Operation \"%s\" returns its argument for UTF-16 surrogate U+%04"UVXf"", desc, uv1);
2289 }
9ae3ac1a
KW
2290 }
2291 else if (UNICODE_IS_SUPER(uv1)) {
8457b38f
KW
2292 if (ckWARN_d(WARN_NON_UNICODE)) {
2293 const char* desc = (PL_op) ? OP_DESC(PL_op) : normal;
2294 Perl_warner(aTHX_ packWARN(WARN_NON_UNICODE),
2295 "Operation \"%s\" returns its argument for non-Unicode code point 0x%04"UVXf"", desc, uv1);
2296 }
9ae3ac1a
KW
2297 }
2298
2299 /* Note that non-characters are perfectly legal, so no warning should
2300 * be given */
2301 }
2302
1feea2c7 2303 uvuni_to_utf8(tmpbuf, uv1);
0134edef
JH
2304
2305 if (!*swashp) /* load on-demand */
2306 *swashp = swash_init("utf8", normal, &PL_sv_undef, 4, 0);
2307
a6f87d8c 2308 if (special) {
0134edef 2309 /* It might be "special" (sometimes, but not always,
2a37f04d 2310 * a multicharacter mapping) */
6673a63c 2311 HV * const hv = get_hv(special, 0);
b08cf34e
JH
2312 SV **svp;
2313
35da51f7 2314 if (hv &&
b08cf34e
JH
2315 (svp = hv_fetch(hv, (const char*)tmpbuf, UNISKIP(uv1), FALSE)) &&
2316 (*svp)) {
cfd0369c 2317 const char *s;
47654450 2318
cfd0369c 2319 s = SvPV_const(*svp, len);
47654450
JH
2320 if (len == 1)
2321 len = uvuni_to_utf8(ustrp, NATIVE_TO_UNI(*(U8*)s)) - ustrp;
2a37f04d 2322 else {
2f9475ad
JH
2323#ifdef EBCDIC
2324 /* If we have EBCDIC we need to remap the characters
2325 * since any characters in the low 256 are Unicode
2326 * code points, not EBCDIC. */
7cda7a3d 2327 U8 *t = (U8*)s, *tend = t + len, *d;
2f9475ad
JH
2328
2329 d = tmpbuf;
b08cf34e 2330 if (SvUTF8(*svp)) {
2f9475ad
JH
2331 STRLEN tlen = 0;
2332
2333 while (t < tend) {
6bd1c396 2334 const UV c = utf8_to_uvchr_buf(t, tend, &tlen);
2f9475ad
JH
2335 if (tlen > 0) {
2336 d = uvchr_to_utf8(d, UNI_TO_NATIVE(c));
2337 t += tlen;
2338 }
2339 else
2340 break;
2341 }
2342 }
2343 else {
36fec512
JH
2344 while (t < tend) {
2345 d = uvchr_to_utf8(d, UNI_TO_NATIVE(*t));
2346 t++;
2347 }
2f9475ad
JH
2348 }
2349 len = d - tmpbuf;
2350 Copy(tmpbuf, ustrp, len, U8);
2351#else
d2dcd0fb 2352 Copy(s, ustrp, len, U8);
2f9475ad 2353#endif
29e98929 2354 }
983ffd37 2355 }
0134edef
JH
2356 }
2357
2358 if (!len && *swashp) {
d4c19fe8
AL
2359 const UV uv2 = swash_fetch(*swashp, tmpbuf, TRUE);
2360
0134edef
JH
2361 if (uv2) {
2362 /* It was "normal" (a single character mapping). */
d4c19fe8 2363 const UV uv3 = UNI_TO_NATIVE(uv2);
e9101d72 2364 len = uvchr_to_utf8(ustrp, uv3) - ustrp;
2a37f04d
JH
2365 }
2366 }
1feea2c7 2367
37e2e78e
KW
2368 if (!len) /* Neither: just copy. In other words, there was no mapping
2369 defined, which means that the code point maps to itself */
0134edef
JH
2370 len = uvchr_to_utf8(ustrp, uv0) - ustrp;
2371
2a37f04d
JH
2372 if (lenp)
2373 *lenp = len;
2374
6bd1c396 2375 return len ? valid_utf8_to_uvchr(ustrp, 0) : 0;
a0ed51b3
LW
2376}
2377
051a06d4
KW
2378STATIC UV
2379S_check_locale_boundary_crossing(pTHX_ const U8* const p, const UV result, U8* const ustrp, STRLEN *lenp)
2380{
2381 /* This is called when changing the case of a utf8-encoded character above
2382 * the Latin1 range, and the operation is in locale. If the result
2383 * contains a character that crosses the 255/256 boundary, disallow the
2384 * change, and return the original code point. See L<perlfunc/lc> for why;
2385 *
a1433954
KW
2386 * p points to the original string whose case was changed; assumed
2387 * by this routine to be well-formed
051a06d4
KW
2388 * result the code point of the first character in the changed-case string
2389 * ustrp points to the changed-case string (<result> represents its first char)
2390 * lenp points to the length of <ustrp> */
2391
2392 UV original; /* To store the first code point of <p> */
2393
2394 PERL_ARGS_ASSERT_CHECK_LOCALE_BOUNDARY_CROSSING;
2395
2396 assert(! UTF8_IS_INVARIANT(*p) && ! UTF8_IS_DOWNGRADEABLE_START(*p));
2397
2398 /* We know immediately if the first character in the string crosses the
2399 * boundary, so can skip */
2400 if (result > 255) {
2401
2402 /* Look at every character in the result; if any cross the
2403 * boundary, the whole thing is disallowed */
2404 U8* s = ustrp + UTF8SKIP(ustrp);
2405 U8* e = ustrp + *lenp;
2406 while (s < e) {
2407 if (UTF8_IS_INVARIANT(*s) || UTF8_IS_DOWNGRADEABLE_START(*s))
2408 {
2409 goto bad_crossing;
2410 }
2411 s += UTF8SKIP(s);
2412 }
2413
2414 /* Here, no characters crossed, result is ok as-is */
2415 return result;
2416 }
2417
2418bad_crossing:
2419
2420 /* Failed, have to return the original */
4b88fb76 2421 original = valid_utf8_to_uvchr(p, lenp);
051a06d4
KW
2422 Copy(p, ustrp, *lenp, char);
2423 return original;
2424}
2425
d3e79532 2426/*
87cea99e 2427=for apidoc to_utf8_upper
d3e79532 2428
a1433954
KW
2429Convert the UTF-8 encoded character at C<p> to its uppercase version and
2430store that in UTF-8 in C<ustrp> and its length in bytes in C<lenp>. Note
89ebb4a3
JH
2431that the ustrp needs to be at least UTF8_MAXBYTES_CASE+1 bytes since
2432the uppercase version may be longer than the original character.
d3e79532
JH
2433
2434The first character of the uppercased version is returned
2435(but note, as explained above, that there may be more.)
2436
a1433954
KW
2437The character at C<p> is assumed by this routine to be well-formed.
2438
d3e79532
JH
2439=cut */
2440
051a06d4
KW
2441/* Not currently externally documented, and subject to change:
2442 * <flags> is set iff locale semantics are to be used for code points < 256
2443 * <tainted_ptr> if non-null, *tainted_ptr will be set TRUE iff locale rules
2444 * were used in the calculation; otherwise unchanged. */
2445
2104c8d9 2446UV
051a06d4 2447Perl__to_utf8_upper_flags(pTHX_ const U8 *p, U8* ustrp, STRLEN *lenp, const bool flags, bool* tainted_ptr)
a0ed51b3 2448{
97aff369 2449 dVAR;
7918f24d 2450
051a06d4
KW
2451 UV result;
2452
2453 PERL_ARGS_ASSERT__TO_UTF8_UPPER_FLAGS;
7918f24d 2454
3a4c58c9 2455 if (UTF8_IS_INVARIANT(*p)) {
051a06d4
KW
2456 if (flags) {
2457 result = toUPPER_LC(*p);
2458 }
2459 else {
81c6c7ce 2460 return _to_upper_title_latin1(*p, ustrp, lenp, 'S');
051a06d4 2461 }
3a4c58c9
KW
2462 }
2463 else if UTF8_IS_DOWNGRADEABLE_START(*p) {
051a06d4
KW
2464 if (flags) {
2465 result = toUPPER_LC(TWO_BYTE_UTF8_TO_UNI(*p, *(p+1)));
2466 }
2467 else {
81c6c7ce
KW
2468 return _to_upper_title_latin1(TWO_BYTE_UTF8_TO_UNI(*p, *(p+1)),
2469 ustrp, lenp, 'S');
051a06d4
KW
2470 }
2471 }
2472 else { /* utf8, ord above 255 */
2473 result = CALL_UPPER_CASE(p, ustrp, lenp);
2474
2475 if (flags) {
2476 result = check_locale_boundary_crossing(p, result, ustrp, lenp);
2477 }
2478 return result;
2479 }
2480
2481 /* Here, used locale rules. Convert back to utf8 */
2482 if (UTF8_IS_INVARIANT(result)) {
2483 *ustrp = (U8) result;
2484 *lenp = 1;
2485 }
2486 else {
2487 *ustrp = UTF8_EIGHT_BIT_HI(result);
2488 *(ustrp + 1) = UTF8_EIGHT_BIT_LO(result);
2489 *lenp = 2;
3a4c58c9
KW
2490 }
2491
051a06d4
KW
2492 if (tainted_ptr) {
2493 *tainted_ptr = TRUE;
2494 }
2495 return result;
983ffd37 2496}
a0ed51b3 2497
d3e79532 2498/*
87cea99e 2499=for apidoc to_utf8_title
d3e79532 2500
a1433954
KW
2501Convert the UTF-8 encoded character at C<p> to its titlecase version and
2502store that in UTF-8 in C<ustrp> and its length in bytes in C<lenp>. Note
2503that the C<ustrp> needs to be at least UTF8_MAXBYTES_CASE+1 bytes since the
89ebb4a3 2504titlecase version may be longer than the original character.
d3e79532
JH
2505
2506The first character of the titlecased version is returned
2507(but note, as explained above, that there may be more.)
2508
a1433954
KW
2509The character at C<p> is assumed by this routine to be well-formed.
2510
d3e79532
JH
2511=cut */
2512
051a06d4
KW
2513/* Not currently externally documented, and subject to change:
2514 * <flags> is set iff locale semantics are to be used for code points < 256
2515 * Since titlecase is not defined in POSIX, uppercase is used instead
2516 * for these/
2517 * <tainted_ptr> if non-null, *tainted_ptr will be set TRUE iff locale rules
2518 * were used in the calculation; otherwise unchanged. */
2519
983ffd37 2520UV
051a06d4 2521Perl__to_utf8_title_flags(pTHX_ const U8 *p, U8* ustrp, STRLEN *lenp, const bool flags, bool* tainted_ptr)
983ffd37 2522{
97aff369 2523 dVAR;
7918f24d 2524
051a06d4
KW
2525 UV result;
2526
2527 PERL_ARGS_ASSERT__TO_UTF8_TITLE_FLAGS;
7918f24d 2528
3a4c58c9 2529 if (UTF8_IS_INVARIANT(*p)) {
051a06d4
KW
2530 if (flags) {
2531 result = toUPPER_LC(*p);
2532 }
2533 else {
81c6c7ce 2534 return _to_upper_title_latin1(*p, ustrp, lenp, 's');
051a06d4 2535 }
3a4c58c9
KW
2536 }
2537 else if UTF8_IS_DOWNGRADEABLE_START(*p) {
051a06d4
KW
2538 if (flags) {
2539 result = toUPPER_LC(TWO_BYTE_UTF8_TO_UNI(*p, *(p+1)));
2540 }
2541 else {
81c6c7ce
KW
2542 return _to_upper_title_latin1(TWO_BYTE_UTF8_TO_UNI(*p, *(p+1)),
2543 ustrp, lenp, 's');
051a06d4
KW
2544 }
2545 }
2546 else { /* utf8, ord above 255 */
2547 result = CALL_TITLE_CASE(p, ustrp, lenp);
2548
2549 if (flags) {
2550 result = check_locale_boundary_crossing(p, result, ustrp, lenp);
2551 }
2552 return result;
2553 }
2554
2555 /* Here, used locale rules. Convert back to utf8 */
2556 if (UTF8_IS_INVARIANT(result)) {
2557 *ustrp = (U8) result;
2558 *lenp = 1;
2559 }
2560 else {
2561 *ustrp = UTF8_EIGHT_BIT_HI(result);
2562 *(ustrp + 1) = UTF8_EIGHT_BIT_LO(result);
2563 *lenp = 2;
3a4c58c9
KW
2564 }
2565
051a06d4
KW
2566 if (tainted_ptr) {
2567 *tainted_ptr = TRUE;
2568 }
2569 return result;
a0ed51b3
LW
2570}
2571
d3e79532 2572/*
87cea99e 2573=for apidoc to_utf8_lower
d3e79532 2574
a1433954
KW
2575Convert the UTF-8 encoded character at C<p> to its lowercase version and
2576store that in UTF-8 in ustrp and its length in bytes in C<lenp>. Note
2577that the C<ustrp> needs to be at least UTF8_MAXBYTES_CASE+1 bytes since the
89ebb4a3 2578lowercase version may be longer than the original character.
d3e79532
JH
2579
2580The first character of the lowercased version is returned
2581(but note, as explained above, that there may be more.)
2582
a1433954
KW
2583The character at C<p> is assumed by this routine to be well-formed.
2584
d3e79532
JH
2585=cut */
2586
051a06d4
KW
2587/* Not currently externally documented, and subject to change:
2588 * <flags> is set iff locale semantics are to be used for code points < 256
2589 * <tainted_ptr> if non-null, *tainted_ptr will be set TRUE iff locale rules
2590 * were used in the calculation; otherwise unchanged. */
2591
2104c8d9 2592UV
051a06d4 2593Perl__to_utf8_lower_flags(pTHX_ const U8 *p, U8* ustrp, STRLEN *lenp, const bool flags, bool* tainted_ptr)
a0ed51b3 2594{
051a06d4
KW
2595 UV result;
2596
97aff369 2597 dVAR;
7918f24d 2598
051a06d4 2599 PERL_ARGS_ASSERT__TO_UTF8_LOWER_FLAGS;
7918f24d 2600
968c5e6a 2601 if (UTF8_IS_INVARIANT(*p)) {
051a06d4
KW
2602 if (flags) {
2603 result = toLOWER_LC(*p);
2604 }
2605 else {
81c6c7ce 2606 return to_lower_latin1(*p, ustrp, lenp);
051a06d4 2607 }
968c5e6a
KW
2608 }
2609 else if UTF8_IS_DOWNGRADEABLE_START(*p) {
051a06d4
KW
2610 if (flags) {
2611 result = toLOWER_LC(TWO_BYTE_UTF8_TO_UNI(*p, *(p+1)));
2612 }
2613 else {
81c6c7ce
KW
2614 return to_lower_latin1(TWO_BYTE_UTF8_TO_UNI(*p, *(p+1)),
2615 ustrp, lenp);
051a06d4 2616 }
968c5e6a 2617 }
051a06d4
KW
2618 else { /* utf8, ord above 255 */
2619 result = CALL_LOWER_CASE(p, ustrp, lenp);
2620
2621 if (flags) {
2622 result = check_locale_boundary_crossing(p, result, ustrp, lenp);
2623 }
968c5e6a 2624
051a06d4
KW
2625 return result;
2626 }
2627
2628 /* Here, used locale rules. Convert back to utf8 */
2629 if (UTF8_IS_INVARIANT(result)) {
2630 *ustrp = (U8) result;
2631 *lenp = 1;
2632 }
2633 else {
2634 *ustrp = UTF8_EIGHT_BIT_HI(result);
2635 *(ustrp + 1) = UTF8_EIGHT_BIT_LO(result);
2636 *lenp = 2;
2637 }
2638
2639 if (tainted_ptr) {
2640 *tainted_ptr = TRUE;
2641 }
2642 return result;
b4e400f9
JH
2643}
2644
d3e79532 2645/*
87cea99e 2646=for apidoc to_utf8_fold
d3e79532 2647
a1433954
KW
2648Convert the UTF-8 encoded character at C<p> to its foldcase version and
2649store that in UTF-8 in C<ustrp> and its length in bytes in C<lenp>. Note
2650that the C<ustrp> needs to be at least UTF8_MAXBYTES_CASE+1 bytes since the
d3e79532
JH
2651foldcase version may be longer than the original character (up to
2652three characters).
2653
2654The first character of the foldcased version is returned
2655(but note, as explained above, that there may be more.)
2656
a1433954
KW
2657The character at C<p> is assumed by this routine to be well-formed.
2658
d3e79532
JH
2659=cut */
2660
051a06d4
KW
2661/* Not currently externally documented, and subject to change,
2662 * in <flags>
2663 * bit FOLD_FLAGS_LOCALE is set iff locale semantics are to be used for code
2664 * points < 256. Since foldcase is not defined in
2665 * POSIX, lowercase is used instead
2666 * bit FOLD_FLAGS_FULL is set iff full case folds are to be used;
2667 * otherwise simple folds
2668 * <tainted_ptr> if non-null, *tainted_ptr will be set TRUE iff locale rules
2669 * were used in the calculation; otherwise unchanged. */
36bb2ab6 2670
b4e400f9 2671UV
051a06d4 2672Perl__to_utf8_fold_flags(pTHX_ const U8 *p, U8* ustrp, STRLEN *lenp, U8 flags, bool* tainted_ptr)
b4e400f9 2673{
97aff369 2674 dVAR;
7918f24d 2675
051a06d4
KW
2676 UV result;
2677
36bb2ab6 2678 PERL_ARGS_ASSERT__TO_UTF8_FOLD_FLAGS;
7918f24d 2679
a1dde8de 2680 if (UTF8_IS_INVARIANT(*p)) {
051a06d4
KW
2681 if (flags & FOLD_FLAGS_LOCALE) {
2682 result = toLOWER_LC(*p);
2683 }
2684 else {
81c6c7ce
KW
2685 return _to_fold_latin1(*p, ustrp, lenp,
2686 cBOOL(flags & FOLD_FLAGS_FULL));
051a06d4 2687 }
a1dde8de
KW
2688 }
2689 else if UTF8_IS_DOWNGRADEABLE_START(*p) {
051a06d4
KW
2690 if (flags & FOLD_FLAGS_LOCALE) {
2691 result = toLOWER_LC(TWO_BYTE_UTF8_TO_UNI(*p, *(p+1)));
2692 }
2693 else {
81c6c7ce 2694 return _to_fold_latin1(TWO_BYTE_UTF8_TO_UNI(*p, *(p+1)),
051a06d4
KW
2695 ustrp, lenp, cBOOL(flags & FOLD_FLAGS_FULL));
2696 }
a1dde8de 2697 }
051a06d4
KW
2698 else { /* utf8, ord above 255 */
2699 result = CALL_FOLD_CASE(p, ustrp, lenp, flags);
a1dde8de 2700
051a06d4
KW
2701 if ((flags & FOLD_FLAGS_LOCALE)) {
2702 result = check_locale_boundary_crossing(p, result, ustrp, lenp);
2703 }
2704
2705 return result;
2706 }
2707
2708 /* Here, used locale rules. Convert back to utf8 */
2709 if (UTF8_IS_INVARIANT(result)) {
2710 *ustrp = (U8) result;
2711 *lenp = 1;
2712 }
2713 else {
2714 *ustrp = UTF8_EIGHT_BIT_HI(result);
2715 *(ustrp + 1) = UTF8_EIGHT_BIT_LO(result);
2716 *lenp = 2;
2717 }
2718
2719 if (tainted_ptr) {
2720 *tainted_ptr = TRUE;
2721 }
2722 return result;
a0ed51b3
LW
2723}
2724
711a919c 2725/* Note:
f90a9a02 2726 * Returns a "swash" which is a hash described in utf8.c:Perl_swash_fetch().
711a919c
TS
2727 * C<pkg> is a pointer to a package name for SWASHNEW, should be "utf8".
2728 * For other parameters, see utf8::SWASHNEW in lib/utf8_heavy.pl.
2729 */
c4a5db0c 2730
a0ed51b3 2731SV*
7fc63493 2732Perl_swash_init(pTHX_ const char* pkg, const char* name, SV *listsv, I32 minbits, I32 none)
a0ed51b3 2733{
c4a5db0c
KW
2734 PERL_ARGS_ASSERT_SWASH_INIT;
2735
2736 /* Returns a copy of a swash initiated by the called function. This is the
2737 * public interface, and returning a copy prevents others from doing
2738 * mischief on the original */
2739
9a53f6cf 2740 return newSVsv(_core_swash_init(pkg, name, listsv, minbits, none, FALSE, NULL, FALSE));
c4a5db0c
KW
2741}
2742
2743SV*
9a53f6cf 2744Perl__core_swash_init(pTHX_ const char* pkg, const char* name, SV *listsv, I32 minbits, I32 none, bool return_if_undef, SV* invlist, bool passed_in_invlist_has_user_defined_property)
c4a5db0c
KW
2745{
2746 /* Initialize and return a swash, creating it if necessary. It does this
9a53f6cf 2747 * by calling utf8_heavy.pl in the general case.
c4a5db0c
KW
2748 *
2749 * This interface should only be used by functions that won't destroy or
2750 * adversely change the swash, as doing so affects all other uses of the
2751 * swash in the program; the general public should use 'Perl_swash_init'
2752 * instead.
2753 *
2754 * pkg is the name of the package that <name> should be in.
2755 * name is the name of the swash to find. Typically it is a Unicode
2756 * property name, including user-defined ones
2757 * listsv is a string to initialize the swash with. It must be of the form
2758 * documented as the subroutine return value in
2759 * L<perlunicode/User-Defined Character Properties>
2760 * minbits is the number of bits required to represent each data element.
2761 * It is '1' for binary properties.
2762 * none I (khw) do not understand this one, but it is used only in tr///.
9a53f6cf
KW
2763 * return_if_undef is TRUE if the routine shouldn't croak if it can't find
2764 * the requested property
2765 * invlist is an inversion list to initialize the swash with (or NULL)
2766 * has_user_defined_property is TRUE if <invlist> has some component that
2767 * came from a user-defined property
2768 *
2769 * Thus there are three possible inputs to find the swash: <name>,
2770 * <listsv>, and <invlist>. At least one must be specified. The result
2771 * will be the union of the specified ones, although <listsv>'s various
2772 * actions can intersect, etc. what <name> gives.
2773 *
2774 * <invlist> is only valid for binary properties */
c4a5db0c 2775
27da23d5 2776 dVAR;
c4a5db0c 2777 SV* retval = &PL_sv_undef;
9a53f6cf
KW
2778
2779 assert(listsv != &PL_sv_undef || strNE(name, "") || invlist);
2780 assert(! invlist || minbits == 1);
2781
2782 /* If data was passed in to go out to utf8_heavy to find the swash of, do
2783 * so */
2784 if (listsv != &PL_sv_undef || strNE(name, "")) {
69794297
KW
2785 dSP;
2786 const size_t pkg_len = strlen(pkg);
2787 const size_t name_len = strlen(name);
2788 HV * const stash = gv_stashpvn(pkg, pkg_len, 0);
2789 SV* errsv_save;
2790 GV *method;
2791
2792 PERL_ARGS_ASSERT__CORE_SWASH_INIT;
2793
2794 PUSHSTACKi(PERLSI_MAGIC);
ce3b816e 2795 ENTER;
69794297
KW
2796 SAVEHINTS();
2797 save_re_context();
2798 if (PL_parser && PL_parser->error_count)
2799 SAVEI8(PL_parser->error_count), PL_parser->error_count = 0;
2800 method = gv_fetchmeth(stash, "SWASHNEW", 8, -1);
2801 if (!method) { /* demand load utf8 */
2802 ENTER;
2803 errsv_save = newSVsv(ERRSV);
2804 /* It is assumed that callers of this routine are not passing in
2805 * any user derived data. */
2806 /* Need to do this after save_re_context() as it will set
2807 * PL_tainted to 1 while saving $1 etc (see the code after getrx:
2808 * in Perl_magic_get). Even line to create errsv_save can turn on
2809 * PL_tainted. */
2810 SAVEBOOL(PL_tainted);
2811 PL_tainted = 0;
2812 Perl_load_module(aTHX_ PERL_LOADMOD_NOIMPORT, newSVpvn(pkg,pkg_len),
2813 NULL);
2814 if (!SvTRUE(ERRSV))
2815 sv_setsv(ERRSV, errsv_save);
2816 SvREFCNT_dec(errsv_save);
2817 LEAVE;
2818 }
2819 SPAGAIN;
2820 PUSHMARK(SP);
2821 EXTEND(SP,5);
2822 mPUSHp(pkg, pkg_len);
2823 mPUSHp(name, name_len);
2824 PUSHs(listsv);
2825 mPUSHi(minbits);
2826 mPUSHi(none);
2827 PUTBACK;
f8be5cf0 2828 errsv_save = newSVsv(ERRSV);
69794297
KW
2829 /* If we already have a pointer to the method, no need to use
2830 * call_method() to repeat the lookup. */
2831 if (method ? call_sv(MUTABLE_SV(method), G_SCALAR)
2832 : call_sv(newSVpvs_flags("SWASHNEW", SVs_TEMP), G_SCALAR | G_METHOD))
2833 {
2834 retval = *PL_stack_sp--;
2835 SvREFCNT_inc(retval);
2836 }
f8be5cf0
JH
2837 if (!SvTRUE(ERRSV))
2838 sv_setsv(ERRSV, errsv_save);
2839 SvREFCNT_dec(errsv_save);
ce3b816e 2840 LEAVE;
69794297
KW
2841 POPSTACK;
2842 if (IN_PERL_COMPILETIME) {
2843 CopHINTS_set(PL_curcop, PL_hints);
2844 }
2845 if (!SvROK(retval) || SvTYPE(SvRV(retval)) != SVt_PVHV) {
2846 if (SvPOK(retval))
2847
2848 /* If caller wants to handle missing properties, let them */
2849 if (return_if_undef) {
2850 return NULL;
2851 }
2852 Perl_croak(aTHX_
2853 "Can't find Unicode property definition \"%"SVf"\"",
2854 SVfARG(retval));
2855 Perl_croak(aTHX_ "SWASHNEW didn't return an HV ref");
2856 }
9a53f6cf 2857 } /* End of calling the module to find the swash */
36eb48b4
KW
2858
2859 /* Make sure there is an inversion list for binary properties */
2860 if (minbits == 1) {
2861 SV** swash_invlistsvp = NULL;
2862 SV* swash_invlist = NULL;
9a53f6cf 2863 bool invlist_in_swash_is_valid = FALSE;
88d45d28 2864 HV* swash_hv = NULL;
36eb48b4 2865
9a53f6cf
KW
2866 /* If this operation fetched a swash, get its already existing
2867 * inversion list or create one for it */
2868 if (retval != &PL_sv_undef) {
36eb48b4
KW
2869 swash_hv = MUTABLE_HV(SvRV(retval));
2870
2871 swash_invlistsvp = hv_fetchs(swash_hv, "INVLIST", FALSE);
9a53f6cf
KW
2872 if (swash_invlistsvp) {
2873 swash_invlist = *swash_invlistsvp;
2874 invlist_in_swash_is_valid = TRUE;
2875 }
2876 else {
36eb48b4 2877 swash_invlist = _swash_to_invlist(retval);
9a53f6cf
KW
2878 }
2879 }
2880
2881 /* If an inversion list was passed in, have to include it */
2882 if (invlist) {
2883
2884 /* Any fetched swash will by now have an inversion list in it;
2885 * otherwise <swash_invlist> will be NULL, indicating that we
2886 * didn't fetch a swash */
2887 if (swash_invlist) {
2888
2889 /* Add the passed-in inversion list, which invalidates the one
2890 * already stored in the swash */
2891 invlist_in_swash_is_valid = FALSE;
2892 _invlist_union(invlist, swash_invlist, &swash_invlist);
2893 }
2894 else {
2895
2896 /* Here, there is no swash already. Set up a minimal one */
2897 swash_hv = newHV();
2898 retval = newRV_inc(MUTABLE_SV(swash_hv));
2899 swash_invlist = invlist;
2900 }
2901
2902 if (passed_in_invlist_has_user_defined_property) {
2903 if (! hv_stores(swash_hv, "USER_DEFINED", newSVuv(1))) {
2904 Perl_croak(aTHX_ "panic: hv_store() unexpectedly failed");
2905 }
2906 }
2907 }
2908
2909 /* Here, we have computed the union of all the passed-in data. It may
2910 * be that there was an inversion list in the swash which didn't get
2911 * touched; otherwise save the one computed one */
2912 if (! invlist_in_swash_is_valid) {
69794297
KW
2913 if (! hv_stores(MUTABLE_HV(SvRV(retval)), "INVLIST", swash_invlist))
2914 {
2915 Perl_croak(aTHX_ "panic: hv_store() unexpectedly failed");
2916 }
9a53f6cf 2917 }
36eb48b4
KW
2918 }
2919
a0ed51b3
LW
2920 return retval;
2921}
2922
035d37be
JH
2923
2924/* This API is wrong for special case conversions since we may need to
2925 * return several Unicode characters for a single Unicode character
2926 * (see lib/unicore/SpecCase.txt) The SWASHGET in lib/utf8_heavy.pl is
2927 * the lower-level routine, and it is similarly broken for returning
38684baa
KW
2928 * multiple values. --jhi
2929 * For those, you should use to_utf8_case() instead */
b0e3252e 2930/* Now SWASHGET is recasted into S_swatch_get in this file. */
680c470c
TS
2931
2932/* Note:
2933 * Returns the value of property/mapping C<swash> for the first character
2934 * of the string C<ptr>. If C<do_utf8> is true, the string C<ptr> is
2935 * assumed to be in utf8. If C<do_utf8> is false, the string C<ptr> is
2936 * assumed to be in native 8-bit encoding. Caches the swatch in C<swash>.
af2af982
KW
2937 *
2938 * A "swash" is a hash which contains initially the keys/values set up by
2939 * SWASHNEW. The purpose is to be able to completely represent a Unicode
2940 * property for all possible code points. Things are stored in a compact form
2941 * (see utf8_heavy.pl) so that calculation is required to find the actual
2942 * property value for a given code point. As code points are looked up, new
2943 * key/value pairs are added to the hash, so that the calculation doesn't have
2944 * to ever be re-done. Further, each calculation is done, not just for the
2945 * desired one, but for a whole block of code points adjacent to that one.
2946 * For binary properties on ASCII machines, the block is usually for 64 code
2947 * points, starting with a code point evenly divisible by 64. Thus if the
2948 * property value for code point 257 is requested, the code goes out and
2949 * calculates the property values for all 64 code points between 256 and 319,
2950 * and stores these as a single 64-bit long bit vector, called a "swatch",
2951 * under the key for code point 256. The key is the UTF-8 encoding for code
2952 * point 256, minus the final byte. Thus, if the length of the UTF-8 encoding
2953 * for a code point is 13 bytes, the key will be 12 bytes long. If the value
2954 * for code point 258 is then requested, this code realizes that it would be
2955 * stored under the key for 256, and would find that value and extract the
2956 * relevant bit, offset from 256.
2957 *
2958 * Non-binary properties are stored in as many bits as necessary to represent
2959 * their values (32 currently, though the code is more general than that), not
2960 * as single bits, but the principal is the same: the value for each key is a
2961 * vector that encompasses the property values for all code points whose UTF-8
2962 * representations are represented by the key. That is, for all code points
2963 * whose UTF-8 representations are length N bytes, and the key is the first N-1
2964 * bytes of that.
680c470c 2965 */
a0ed51b3 2966UV
680c470c 2967Perl_swash_fetch(pTHX_ SV *swash, const U8 *ptr, bool do_utf8)
a0ed51b3 2968{
27da23d5 2969 dVAR;
ef8f7699 2970 HV *const hv = MUTABLE_HV(SvRV(swash));
3568d838
JH
2971 U32 klen;
2972 U32 off;
a0ed51b3 2973 STRLEN slen;
7d85a32c 2974 STRLEN needents;
cfd0369c 2975 const U8 *tmps = NULL;
a0ed51b3 2976 U32 bit;
979f2922 2977 SV *swatch;
3568d838 2978 U8 tmputf8[2];
35da51f7 2979 const UV c = NATIVE_TO_ASCII(*ptr);
3568d838 2980
7918f24d
NC
2981 PERL_ARGS_ASSERT_SWASH_FETCH;
2982
dbe7a391 2983 /* Convert to utf8 if not already */
3568d838 2984 if (!do_utf8 && !UNI_IS_INVARIANT(c)) {
979f2922
TS
2985 tmputf8[0] = (U8)UTF8_EIGHT_BIT_HI(c);
2986 tmputf8[1] = (U8)UTF8_EIGHT_BIT_LO(c);
2987 ptr = tmputf8;
3568d838
JH
2988 }
2989 /* Given a UTF-X encoded char 0xAA..0xYY,0xZZ
37e2e78e 2990 * then the "swatch" is a vec() for all the chars which start
3568d838
JH
2991 * with 0xAA..0xYY
2992 * So the key in the hash (klen) is length of encoded char -1
2993 */
2994 klen = UTF8SKIP(ptr) - 1;
2995 off = ptr[klen];
a0ed51b3 2996
979f2922 2997 if (klen == 0) {
37e2e78e 2998 /* If char is invariant then swatch is for all the invariant chars
1e54db1a 2999 * In both UTF-8 and UTF-8-MOD that happens to be UTF_CONTINUATION_MARK
7d85a32c 3000 */
979f2922
TS
3001 needents = UTF_CONTINUATION_MARK;
3002 off = NATIVE_TO_UTF(ptr[klen]);
3003 }
3004 else {
7d85a32c 3005 /* If char is encoded then swatch is for the prefix */
979f2922
TS
3006 needents = (1 << UTF_ACCUMULATION_SHIFT);
3007 off = NATIVE_TO_UTF(ptr[klen]) & UTF_CONTINUATION_MASK;
3008 }
7d85a32c 3009
a0ed51b3
LW
3010 /*
3011 * This single-entry cache saves about 1/3 of the utf8 overhead in test
3012 * suite. (That is, only 7-8% overall over just a hash cache. Still,
3013 * it's nothing to sniff at.) Pity we usually come through at least
3014 * two function calls to get here...
3015 *
3016 * NB: this code assumes that swatches are never modified, once generated!
3017 */
3018
3568d838 3019 if (hv == PL_last_swash_hv &&
a0ed51b3 3020 klen == PL_last_swash_klen &&
27da23d5 3021 (!klen || memEQ((char *)ptr, (char *)PL_last_swash_key, klen)) )
a0ed51b3
LW
3022 {
3023 tmps = PL_last_swash_tmps;
3024 slen = PL_last_swash_slen;
3025 }
3026 else {
3027 /* Try our second-level swatch cache, kept in a hash. */
e1ec3a88 3028 SV** svp = hv_fetch(hv, (const char*)ptr, klen, FALSE);
a0ed51b3 3029
b0e3252e 3030 /* If not cached, generate it via swatch_get */
979f2922
TS
3031 if (!svp || !SvPOK(*svp)
3032 || !(tmps = (const U8*)SvPV_const(*svp, slen))) {
2b9d42f0
NIS
3033 /* We use utf8n_to_uvuni() as we want an index into
3034 Unicode tables, not a native character number.
3035 */
aec46f14 3036 const UV code_point = utf8n_to_uvuni(ptr, UTF8_MAXBYTES, 0,
872c91ae
JH
3037 ckWARN(WARN_UTF8) ?
3038 0 : UTF8_ALLOW_ANY);
b0e3252e 3039 swatch = swatch_get(swash,
979f2922 3040 /* On EBCDIC & ~(0xA0-1) isn't a useful thing to do */
361ee0fe 3041 (klen) ? (code_point & ~((UV)needents - 1)) : 0,
979f2922
TS
3042 needents);
3043
923e4eb5 3044 if (IN_PERL_COMPILETIME)
623e6609 3045 CopHINTS_set(PL_curcop, PL_hints);
a0ed51b3 3046
979f2922 3047 svp = hv_store(hv, (const char *)ptr, klen, swatch, 0);
a0ed51b3 3048
979f2922
TS
3049 if (!svp || !(tmps = (U8*)SvPV(*svp, slen))
3050 || (slen << 3) < needents)
5637ef5b
NC
3051 Perl_croak(aTHX_ "panic: swash_fetch got improper swatch, "
3052 "svp=%p, tmps=%p, slen=%"UVuf", needents=%"UVuf,
3053 svp, tmps, (UV)slen, (UV)needents);
a0ed51b3
LW
3054 }
3055
3056 PL_last_swash_hv = hv;
16d8f38a 3057 assert(klen <= sizeof(PL_last_swash_key));
eac04b2e 3058 PL_last_swash_klen = (U8)klen;
cfd0369c
NC
3059 /* FIXME change interpvar.h? */
3060 PL_last_swash_tmps = (U8 *) tmps;
a0ed51b3
LW
3061 PL_last_swash_slen = slen;
3062 if (klen)
3063 Copy(ptr, PL_last_swash_key, klen, U8);
3064 }
3065
a410ec23
KW
3066 if (UTF8_IS_SUPER(ptr) && ckWARN_d(WARN_NON_UNICODE)) {
3067 SV** const bitssvp = hv_fetchs(hv, "BITS", FALSE);
3068
3069 /* This outputs warnings for binary properties only, assuming that
3070 * to_utf8_case() will output any for non-binary. Also, surrogates
3071 * aren't checked for, as that would warn on things like /\p{Gc=Cs}/ */
3072
9a53f6cf 3073 if (! bitssvp || SvUV(*bitssvp) == 1) {
a410ec23
KW
3074 /* User-defined properties can silently match above-Unicode */
3075 SV** const user_defined_svp = hv_fetchs(hv, "USER_DEFINED", FALSE);
3076 if (! user_defined_svp || ! SvUV(*user_defined_svp)) {
3077 const UV code_point = utf8n_to_uvuni(ptr, UTF8_MAXBYTES, 0, 0);
3078 Perl_warner(aTHX_ packWARN(WARN_NON_UNICODE),
3079 "Code point 0x%04"UVXf" is not Unicode, all \\p{} matches fail; all \\P{} matches succeed", code_point);
3080 }
3081 }
3082 }
3083
9faf8d75 3084 switch ((int)((slen << 3) / needents)) {
a0ed51b3
LW
3085 case 1:
3086 bit = 1 << (off & 7);
3087 off >>= 3;
3088 return (tmps[off] & bit) != 0;
3089 case 8:
3090 return tmps[off];
3091 case 16:
3092 off <<= 1;
3093 return (tmps[off] << 8) + tmps[off + 1] ;
3094 case 32:
3095 off <<= 2;
3096 return (tmps[off] << 24) + (tmps[off+1] << 16) + (tmps[off+2] << 8) + tmps[off + 3] ;
3097 }
5637ef5b
NC
3098 Perl_croak(aTHX_ "panic: swash_fetch got swatch of unexpected bit width, "
3099 "slen=%"UVuf", needents=%"UVuf, (UV)slen, (UV)needents);
670f1322 3100 NORETURN_FUNCTION_END;
a0ed51b3 3101}
2b9d42f0 3102
319009ee
KW
3103/* Read a single line of the main body of the swash input text. These are of
3104 * the form:
3105 * 0053 0056 0073
3106 * where each number is hex. The first two numbers form the minimum and
3107 * maximum of a range, and the third is the value associated with the range.
3108 * Not all swashes should have a third number
3109 *
3110 * On input: l points to the beginning of the line to be examined; it points
3111 * to somewhere in the string of the whole input text, and is
3112 * terminated by a \n or the null string terminator.
3113 * lend points to the null terminator of that string
3114 * wants_value is non-zero if the swash expects a third number
3115 * typestr is the name of the swash's mapping, like 'ToLower'
3116 * On output: *min, *max, and *val are set to the values read from the line.
3117 * returns a pointer just beyond the line examined. If there was no
3118 * valid min number on the line, returns lend+1
3119 */
3120
3121STATIC U8*
3122S_swash_scan_list_line(pTHX_ U8* l, U8* const lend, UV* min, UV* max, UV* val,
3123 const bool wants_value, const U8* const typestr)
3124{
3125 const int typeto = typestr[0] == 'T' && typestr[1] == 'o';
3126 STRLEN numlen; /* Length of the number */
02470786
KW
3127 I32 flags = PERL_SCAN_SILENT_ILLDIGIT
3128 | PERL_SCAN_DISALLOW_PREFIX
3129 | PERL_SCAN_SILENT_NON_PORTABLE;
319009ee
KW
3130
3131 /* nl points to the next \n in the scan */
3132 U8* const nl = (U8*)memchr(l, '\n', lend - l);
3133
3134 /* Get the first number on the line: the range minimum */
3135 numlen = lend - l;
3136 *min = grok_hex((char *)l, &numlen, &flags, NULL);
3137 if (numlen) /* If found a hex number, position past it */
3138 l += numlen;
3139 else if (nl) { /* Else, go handle next line, if any */
3140 return nl + 1; /* 1 is length of "\n" */
3141 }
3142 else { /* Else, no next line */
3143 return lend + 1; /* to LIST's end at which \n is not found */
3144 }
3145
3146 /* The max range value follows, separated by a BLANK */
3147 if (isBLANK(*l)) {
3148 ++l;
02470786
KW
3149 flags = PERL_SCAN_SILENT_ILLDIGIT
3150 | PERL_SCAN_DISALLOW_PREFIX
3151 | PERL_SCAN_SILENT_NON_PORTABLE;
319009ee
KW
3152 numlen = lend - l;
3153 *max = grok_hex((char *)l, &numlen, &flags, NULL);
3154 if (numlen)
3155 l += numlen;
3156 else /* If no value here, it is a single element range */
3157 *max = *min;
3158
3159 /* Non-binary tables have a third entry: what the first element of the
3160 * range maps to */
3161 if (wants_value) {
3162 if (isBLANK(*l)) {
3163 ++l;
f90a9a02
KW
3164
3165 /* The ToLc, etc table mappings are not in hex, and must be
3166 * corrected by adding the code point to them */
3167 if (typeto) {
3168 char *after_strtol = (char *) lend;
3169 *val = Strtol((char *)l, &after_strtol, 10);
3170 l = (U8 *) after_strtol;
f90a9a02
KW
3171 }
3172 else { /* Other tables are in hex, and are the correct result
3173 without tweaking */
a9d188b3
KW
3174 flags = PERL_SCAN_SILENT_ILLDIGIT
3175 | PERL_SCAN_DISALLOW_PREFIX
3176 | PERL_SCAN_SILENT_NON_PORTABLE;
3177 numlen = lend - l;
3178 *val = grok_hex((char *)l, &numlen, &flags, NULL);
3179 if (numlen)
3180 l += numlen;
3181 else
3182 *val = 0;
f90a9a02 3183 }
319009ee
KW
3184 }
3185 else {
3186 *val = 0;
3187 if (typeto) {
dcbac5bb 3188 /* diag_listed_as: To%s: illegal mapping '%s' */
319009ee
KW
3189 Perl_croak(aTHX_ "%s: illegal mapping '%s'",
3190 typestr, l);
3191 }
3192 }
3193 }
3194 else
3195 *val = 0; /* bits == 1, then any val should be ignored */
3196 }
3197 else { /* Nothing following range min, should be single element with no
3198 mapping expected */
3199 *max = *min;
3200 if (wants_value) {
3201 *val = 0;
3202 if (typeto) {
dcbac5bb 3203 /* diag_listed_as: To%s: illegal mapping '%s' */
319009ee
KW
3204 Perl_croak(aTHX_ "%s: illegal mapping '%s'", typestr, l);
3205 }
3206 }
3207 else
3208 *val = 0; /* bits == 1, then val should be ignored */
3209 }
3210
3211 /* Position to next line if any, or EOF */
3212 if (nl)
3213 l = nl + 1;
3214 else
3215 l = lend;
3216
3217 return l;
3218}
3219
979f2922
TS
3220/* Note:
3221 * Returns a swatch (a bit vector string) for a code point sequence
3222 * that starts from the value C<start> and comprises the number C<span>.
3223 * A C<swash> must be an object created by SWASHNEW (see lib/utf8_heavy.pl).
3224 * Should be used via swash_fetch, which will cache the swatch in C<swash>.
3225 */
3226STATIC SV*
b0e3252e 3227S_swatch_get(pTHX_ SV* swash, UV start, UV span)
979f2922
TS
3228{
3229 SV *swatch;
77f9f126 3230 U8 *l, *lend, *x, *xend, *s, *send;
979f2922 3231 STRLEN lcur, xcur, scur;
ef8f7699 3232 HV *const hv = MUTABLE_HV(SvRV(swash));
36eb48b4
KW
3233 SV** const invlistsvp = hv_fetchs(hv, "INVLIST", FALSE);
3234
88d45d28
KW
3235 SV** listsvp = NULL; /* The string containing the main body of the table */
3236 SV** extssvp = NULL;
3237 SV** invert_it_svp = NULL;
3238 U8* typestr = NULL;
786861f5
KW
3239 STRLEN bits;
3240 STRLEN octets; /* if bits == 1, then octets == 0 */
3241 UV none;
3242 UV end = start + span;
972dd592 3243
36eb48b4 3244 if (invlistsvp == NULL) {
786861f5
KW
3245 SV** const bitssvp = hv_fetchs(hv, "BITS", FALSE);
3246 SV** const nonesvp = hv_fetchs(hv, "NONE", FALSE);
3247 SV** const typesvp = hv_fetchs(hv, "TYPE", FALSE);
3248 extssvp = hv_fetchs(hv, "EXTRAS", FALSE);
3249 listsvp = hv_fetchs(hv, "LIST", FALSE);
3250 invert_it_svp = hv_fetchs(hv, "INVERT_IT", FALSE);
3251
3252 bits = SvUV(*bitssvp);
3253 none = SvUV(*nonesvp);
3254 typestr = (U8*)SvPV_nolen(*typesvp);
3255 }
36eb48b4
KW
3256 else {
3257 bits = 1;
3258 none = 0;
3259 }
786861f5 3260 octets = bits >> 3; /* if bits == 1, then octets == 0 */
979f2922 3261
b0e3252e 3262 PERL_ARGS_ASSERT_SWATCH_GET;
7918f24d 3263
979f2922 3264 if (bits != 1 && bits != 8 && bits != 16 && bits != 32) {
b0e3252e 3265 Perl_croak(aTHX_ "panic: swatch_get doesn't expect bits %"UVuf,
660a4616 3266 (UV)bits);
979f2922
TS
3267 }
3268
84ea5ef6
KW
3269 /* If overflowed, use the max possible */
3270 if (end < start) {
3271 end = UV_MAX;
3272 span = end - start;
3273 }
3274
979f2922 3275 /* create and initialize $swatch */
979f2922 3276 scur = octets ? (span * octets) : (span + 7) / 8;
e524fe40
NC
3277 swatch = newSV(scur);
3278 SvPOK_on(swatch);
979f2922
TS
3279 s = (U8*)SvPVX(swatch);
3280 if (octets && none) {
0bd48802 3281 const U8* const e = s + scur;
979f2922
TS
3282 while (s < e) {
3283 if (bits == 8)
3284 *s++ = (U8)(none & 0xff);
3285 else if (bits == 16) {
3286 *s++ = (U8)((none >> 8) & 0xff);
3287 *s++ = (U8)( none & 0xff);
3288 }
3289 else if (bits == 32) {
3290 *s++ = (U8)((none >> 24) & 0xff);
3291 *s++ = (U8)((none >> 16) & 0xff);
3292 *s++ = (U8)((none >> 8) & 0xff);
3293 *s++ = (U8)( none & 0xff);
3294 }
3295 }
3296 *s = '\0';
3297 }
3298 else {
3299 (void)memzero((U8*)s, scur + 1);
3300 }
3301 SvCUR_set(swatch, scur);
3302 s = (U8*)SvPVX(swatch);
3303
36eb48b4
KW
3304 if (invlistsvp) { /* If has an inversion list set up use that */
3305 _invlist_populate_swatch(*invlistsvp, start, end, s);
3306 return swatch;
3307 }
3308
3309 /* read $swash->{LIST} */
979f2922
TS
3310 l = (U8*)SvPV(*listsvp, lcur);
3311 lend = l + lcur;
3312 while (l < lend) {
8ed25d53 3313 UV min, max, val, upper;
319009ee
KW
3314 l = S_swash_scan_list_line(aTHX_ l, lend, &min, &max, &val,
3315 cBOOL(octets), typestr);
3316 if (l > lend) {
979f2922
TS
3317 break;
3318 }
3319
972dd592 3320 /* If looking for something beyond this range, go try the next one */
979f2922
TS
3321 if (max < start)
3322 continue;
3323
8ed25d53
KW
3324 /* <end> is generally 1 beyond where we want to set things, but at the
3325 * platform's infinity, where we can't go any higher, we want to
3326 * include the code point at <end> */
3327 upper = (max < end)
3328 ? max
3329 : (max != UV_MAX || end != UV_MAX)
3330 ? end - 1
3331 : end;
3332
979f2922 3333 if (octets) {
35da51f7 3334 UV key;
979f2922
TS
3335 if (min < start) {
3336 if (!none || val < none) {
3337 val += start - min;
3338 }
3339 min = start;
3340 }
8ed25d53 3341 for (key = min; key <= upper; key++) {
979f2922 3342 STRLEN offset;
979f2922
TS
3343 /* offset must be non-negative (start <= min <= key < end) */
3344 offset = octets * (key - start);
3345 if (bits == 8)
3346 s[offset] = (U8)(val & 0xff);
3347 else if (bits == 16) {
3348 s[offset ] = (U8)((val >> 8) & 0xff);
3349 s[offset + 1] = (U8)( val & 0xff);
3350 }
3351 else if (bits == 32) {
3352 s[offset ] = (U8)((val >> 24) & 0xff);
3353 s[offset + 1] = (U8)((val >> 16) & 0xff);
3354 s[offset + 2] = (U8)((val >> 8) & 0xff);
3355 s[offset + 3] = (U8)( val & 0xff);
3356 }
3357
3358 if (!none || val < none)
3359 ++val;
3360 }
3361 }
711a919c 3362 else { /* bits == 1, then val should be ignored */
35da51f7 3363 UV key;
979f2922
TS
3364 if (min < start)
3365 min = start;
6cb05c12 3366
8ed25d53 3367 for (key = min; key <= upper; key++) {
0bd48802 3368 const STRLEN offset = (STRLEN)(key - start);
979f2922
TS
3369 s[offset >> 3] |= 1 << (offset & 7);
3370 }
3371 }
3372 } /* while */
979f2922 3373
9479a769 3374 /* Invert if the data says it should be. Assumes that bits == 1 */
77f9f126 3375 if (invert_it_svp && SvUV(*invert_it_svp)) {
0bda3001
KW
3376
3377 /* Unicode properties should come with all bits above PERL_UNICODE_MAX
3378 * be 0, and their inversion should also be 0, as we don't succeed any
3379 * Unicode property matches for non-Unicode code points */
3380 if (start <= PERL_UNICODE_MAX) {
3381
3382 /* The code below assumes that we never cross the
3383 * Unicode/above-Unicode boundary in a range, as otherwise we would
3384 * have to figure out where to stop flipping the bits. Since this
3385 * boundary is divisible by a large power of 2, and swatches comes
3386 * in small powers of 2, this should be a valid assumption */
3387 assert(start + span - 1 <= PERL_UNICODE_MAX);
3388
507a8485
KW
3389 send = s + scur;
3390 while (s < send) {
3391 *s = ~(*s);
3392 s++;
3393 }
0bda3001 3394 }
77f9f126
KW
3395 }
3396
d73c39c5
KW
3397 /* read $swash->{EXTRAS}
3398 * This code also copied to swash_to_invlist() below */
979f2922
TS
3399 x = (U8*)SvPV(*extssvp, xcur);
3400 xend = x + xcur;
3401 while (x < xend) {
3402 STRLEN namelen;
3403 U8 *namestr;
3404 SV** othersvp;
3405 HV* otherhv;
3406 STRLEN otherbits;
3407 SV **otherbitssvp, *other;
711a919c 3408 U8 *s, *o, *nl;
979f2922
TS
3409 STRLEN slen, olen;
3410
35da51f7 3411 const U8 opc = *x++;
979f2922
TS
3412 if (opc == '\n')
3413 continue;
3414
3415 nl = (U8*)memchr(x, '\n', xend - x);
3416
3417 if (opc != '-' && opc != '+' && opc != '!' && opc != '&') {
3418 if (nl) {
3419 x = nl + 1; /* 1 is length of "\n" */
3420 continue;
3421 }
3422 else {
3423 x = xend; /* to EXTRAS' end at which \n is not found */
3424 break;
3425 }
3426 }
3427
3428 namestr = x;
3429 if (nl) {
3430 namelen = nl - namestr;
3431 x = nl + 1;
3432 }
3433 else {
3434 namelen = xend - namestr;
3435 x = xend;
3436 }
3437
3438 othersvp = hv_fetch(hv, (char *)namestr, namelen, FALSE);
ef8f7699 3439 otherhv = MUTABLE_HV(SvRV(*othersvp));
017a3ce5 3440 otherbitssvp = hv_fetchs(otherhv, "BITS", FALSE);
979f2922
TS
3441 otherbits = (STRLEN)SvUV(*otherbitssvp);
3442 if (bits < otherbits)
5637ef5b
NC
3443 Perl_croak(aTHX_ "panic: swatch_get found swatch size mismatch, "
3444 "bits=%"UVuf", otherbits=%"UVuf, (UV)bits, (UV)otherbits);
979f2922
TS
3445
3446 /* The "other" swatch must be destroyed after. */
b0e3252e 3447 other = swatch_get(*othersvp, start, span);
979f2922
TS
3448 o = (U8*)SvPV(other, olen);
3449
3450 if (!olen)
b0e3252e 3451 Perl_croak(aTHX_ "panic: swatch_get got improper swatch");
979f2922
TS
3452
3453 s = (U8*)SvPV(swatch, slen);
3454 if (bits == 1 && otherbits == 1) {
3455 if (slen != olen)
5637ef5b
NC
3456 Perl_croak(aTHX_ "panic: swatch_get found swatch length "
3457 "mismatch, slen=%"UVuf", olen=%"UVuf,
3458 (UV)slen, (UV)olen);
979f2922
TS
3459
3460 switch (opc) {
3461 case '+':
3462 while (slen--)
3463 *s++ |= *o++;
3464 break;
3465 case '!':
3466 while (slen--)
3467 *s++ |= ~*o++;
3468 break;
3469 case '-':
3470 while (slen--)
3471 *s++ &= ~*o++;
3472 break;
3473 case '&':
3474 while (slen--)
3475 *s++ &= *o++;
3476 break;
3477 default:
3478 break;
3479 }
3480 }
711a919c 3481 else {
979f2922
TS
3482 STRLEN otheroctets = otherbits >> 3;
3483 STRLEN offset = 0;
35da51f7 3484 U8* const send = s + slen;
979f2922
TS
3485
3486 while (s < send) {
3487 UV otherval = 0;
3488
3489 if (otherbits == 1) {
3490 otherval = (o[offset >> 3] >> (offset & 7)) & 1;
3491 ++offset;
3492 }
3493 else {
3494 STRLEN vlen = otheroctets;
3495 otherval = *o++;
3496 while (--vlen) {
3497 otherval <<= 8;
3498 otherval |= *o++;
3499 }
3500 }
3501
711a919c 3502 if (opc == '+' && otherval)
6f207bd3 3503 NOOP; /* replace with otherval */
979f2922
TS
3504 else if (opc == '!' && !otherval)
3505 otherval = 1;
3506 else if (opc == '-' && otherval)
3507 otherval = 0;
3508 else if (opc == '&' && !otherval)
3509 otherval = 0;
3510 else {
711a919c 3511 s += octets; /* no replacement */
979f2922
TS
3512 continue;
3513 }
3514
3515 if (bits == 8)
3516 *s++ = (U8)( otherval & 0xff);
3517 else if (bits == 16) {
3518 *s++ = (U8)((otherval >> 8) & 0xff);
3519 *s++ = (U8)( otherval & 0xff);
3520 }
3521 else if (bits == 32) {
3522 *s++ = (U8)((otherval >> 24) & 0xff);
3523 *s++ = (U8)((otherval >> 16) & 0xff);
3524 *s++ = (U8)((otherval >> 8) & 0xff);
3525 *s++ = (U8)( otherval & 0xff);
3526 }
3527 }
3528 }
3529 sv_free(other); /* through with it! */
3530 } /* while */
3531 return swatch;
3532}
3533
064c021d 3534HV*
4c2e1131 3535Perl__swash_inversion_hash(pTHX_ SV* const swash)
064c021d
KW
3536{
3537
5662e334
KW
3538 /* Subject to change or removal. For use only in one place in regcomp.c.
3539 * Can't be used on a property that is subject to user override, as it
3540 * relies on the value of SPECIALS in the swash which would be set by
3541 * utf8_heavy.pl to the hash in the non-overriden file, and hence is not set
3542 * for overridden properties
064c021d
KW
3543 *
3544 * Returns a hash which is the inversion and closure of a swash mapping.
3545 * For example, consider the input lines:
3546 * 004B 006B
3547 * 004C 006C
3548 * 212A 006B
3549 *
3550 * The returned hash would have two keys, the utf8 for 006B and the utf8 for
3551 * 006C. The value for each key is an array. For 006C, the array would
3552 * have a two elements, the utf8 for itself, and for 004C. For 006B, there
3553 * would be three elements in its array, the utf8 for 006B, 004B and 212A.
3554 *
3555 * Essentially, for any code point, it gives all the code points that map to
3556 * it, or the list of 'froms' for that point.
3557 *
5662e334
KW
3558 * Currently it ignores any additions or deletions from other swashes,
3559 * looking at just the main body of the swash, and if there are SPECIALS
3560 * in the swash, at that hash
3561 *
3562 * The specials hash can be extra code points, and most likely consists of
3563 * maps from single code points to multiple ones (each expressed as a string
3564 * of utf8 characters). This function currently returns only 1-1 mappings.
3565 * However consider this possible input in the specials hash:
3566 * "\xEF\xAC\x85" => "\x{0073}\x{0074}", # U+FB05 => 0073 0074
3567 * "\xEF\xAC\x86" => "\x{0073}\x{0074}", # U+FB06 => 0073 0074
3568 *
3569 * Both FB05 and FB06 map to the same multi-char sequence, which we don't
3570 * currently handle. But it also means that FB05 and FB06 are equivalent in
3571 * a 1-1 mapping which we should handle, and this relationship may not be in
3572 * the main table. Therefore this function examines all the multi-char
3573 * sequences and adds the 1-1 mappings that come out of that. */
064c021d
KW
3574
3575 U8 *l, *lend;
3576 STRLEN lcur;
3577 HV *const hv = MUTABLE_HV(SvRV(swash));
3578
3579 /* The string containing the main body of the table */
3580 SV** const listsvp = hv_fetchs(hv, "LIST", FALSE);
3581
3582 SV** const typesvp = hv_fetchs(hv, "TYPE", FALSE);
3583 SV** const bitssvp = hv_fetchs(hv, "BITS", FALSE);
3584 SV** const nonesvp = hv_fetchs(hv, "NONE", FALSE);
3585 /*SV** const extssvp = hv_fetchs(hv, "EXTRAS", FALSE);*/
3586 const U8* const typestr = (U8*)SvPV_nolen(*typesvp);
3587 const STRLEN bits = SvUV(*bitssvp);
3588 const STRLEN octets = bits >> 3; /* if bits == 1, then octets == 0 */
3589 const UV none = SvUV(*nonesvp);
5662e334 3590 SV **specials_p = hv_fetchs(hv, "SPECIALS", 0);
064c021d
KW
3591
3592 HV* ret = newHV();
3593
3594 PERL_ARGS_ASSERT__SWASH_INVERSION_HASH;
3595
3596 /* Must have at least 8 bits to get the mappings */
3597 if (bits != 8 && bits != 16 && bits != 32) {
3598 Perl_croak(aTHX_ "panic: swash_inversion_hash doesn't expect bits %"UVuf,
3599 (UV)bits);
3600 }
3601
5662e334
KW
3602 if (specials_p) { /* It might be "special" (sometimes, but not always, a
3603 mapping to more than one character */
3604
3605 /* Construct an inverse mapping hash for the specials */
3606 HV * const specials_hv = MUTABLE_HV(SvRV(*specials_p));
3607 HV * specials_inverse = newHV();
3608 char *char_from; /* the lhs of the map */
3609 I32 from_len; /* its byte length */
3610 char *char_to; /* the rhs of the map */
3611 I32 to_len; /* its byte length */
3612 SV *sv_to; /* and in a sv */
3613 AV* from_list; /* list of things that map to each 'to' */
3614
3615 hv_iterinit(specials_hv);
3616
3617 /* The keys are the characters (in utf8) that map to the corresponding
3618 * utf8 string value. Iterate through the list creating the inverse
3619 * list. */
3620 while ((sv_to = hv_iternextsv(specials_hv, &char_from, &from_len))) {
3621 SV** listp;
3622 if (! SvPOK(sv_to)) {
5637ef5b
NC
3623 Perl_croak(aTHX_ "panic: value returned from hv_iternextsv() "
3624 "unexpectedly is not a string, flags=%lu",
3625 (unsigned long)SvFLAGS(sv_to));
5662e334 3626 }
4b88fb76 3627 /*DEBUG_U(PerlIO_printf(Perl_debug_log, "Found mapping from %"UVXf", First char of to is %"UVXf"\n", valid_utf8_to_uvchr((U8*) char_from, 0), valid_utf8_to_uvchr((U8*) SvPVX(sv_to), 0)));*/
5662e334
KW
3628
3629 /* Each key in the inverse list is a mapped-to value, and the key's
3630 * hash value is a list of the strings (each in utf8) that map to
3631 * it. Those strings are all one character long */
3632 if ((listp = hv_fetch(specials_inverse,
3633 SvPVX(sv_to),
3634 SvCUR(sv_to), 0)))
3635 {
3636 from_list = (AV*) *listp;
3637 }
3638 else { /* No entry yet for it: create one */
3639 from_list = newAV();
3640 if (! hv_store(specials_inverse,
3641 SvPVX(sv_to),
3642 SvCUR(sv_to),
3643 (SV*) from_list, 0))
3644 {
3645 Perl_croak(aTHX_ "panic: hv_store() unexpectedly failed");
3646 }
3647 }
3648
3649 /* Here have the list associated with this 'to' (perhaps newly
3650 * created and empty). Just add to it. Note that we ASSUME that
3651 * the input is guaranteed to not have duplications, so we don't
3652 * check for that. Duplications just slow down execution time. */
3653 av_push(from_list, newSVpvn_utf8(char_from, from_len, TRUE));
3654 }
3655
3656 /* Here, 'specials_inverse' contains the inverse mapping. Go through
3657 * it looking for cases like the FB05/FB06 examples above. There would
3658 * be an entry in the hash like
3659 * 'st' => [ FB05, FB06 ]
3660 * In this example we will create two lists that get stored in the
3661 * returned hash, 'ret':
3662 * FB05 => [ FB05, FB06 ]
3663 * FB06 => [ FB05, FB06 ]
3664 *
3665 * Note that there is nothing to do if the array only has one element.
3666 * (In the normal 1-1 case handled below, we don't have to worry about
3667 * two lists, as everything gets tied to the single list that is
3668 * generated for the single character 'to'. But here, we are omitting
3669 * that list, ('st' in the example), so must have multiple lists.) */
3670 while ((from_list = (AV *) hv_iternextsv(specials_inverse,
3671 &char_to, &to_len)))
3672 {
3673 if (av_len(from_list) > 0) {
3674 int i;
3675
3676 /* We iterate over all combinations of i,j to place each code
3677 * point on each list */
3678 for (i = 0; i <= av_len(from_list); i++) {
3679 int j;
3680 AV* i_list = newAV();
3681 SV** entryp = av_fetch(from_list, i, FALSE);
3682 if (entryp == NULL) {
3683 Perl_croak(aTHX_ "panic: av_fetch() unexpectedly failed");
3684 }
3685 if (hv_fetch(ret, SvPVX(*entryp), SvCUR(*entryp), FALSE)) {
3686 Perl_croak(aTHX_ "panic: unexpected entry for %s", SvPVX(*entryp));
3687 }
3688 if (! hv_store(ret, SvPVX(*entryp), SvCUR(*entryp),
3689 (SV*) i_list, FALSE))
3690 {
3691 Perl_croak(aTHX_ "panic: hv_store() unexpectedly failed");
3692 }
3693
4b88fb76 3694 /* For debugging: UV u = valid_utf8_to_uvchr((U8*) SvPVX(*entryp), 0);*/
5662e334
KW
3695 for (j = 0; j <= av_len(from_list); j++) {
3696 entryp = av_fetch(from_list, j, FALSE);
3697 if (entryp == NULL) {
3698 Perl_croak(aTHX_ "panic: av_fetch() unexpectedly failed");
3699 }
3700
3701 /* When i==j this adds itself to the list */
4b88fb76
KW
3702 av_push(i_list, newSVuv(utf8_to_uvchr_buf(
3703 (U8*) SvPVX(*entryp),
3704 (U8*) SvPVX(*entryp) + SvCUR(*entryp),
3705 0)));
3706 /*DEBUG_U(PerlIO_printf(Perl_debug_log, "Adding %"UVXf" to list for %"UVXf"\n", valid_utf8_to_uvchr((U8*) SvPVX(*entryp), 0), u));*/
5662e334
KW
3707 }
3708 }
3709 }
3710 }
3711 SvREFCNT_dec(specials_inverse); /* done with it */
3712 } /* End of specials */
3713
064c021d
KW
3714 /* read $swash->{LIST} */
3715 l = (U8*)SvPV(*listsvp, lcur);
3716 lend = l + lcur;
3717
3718 /* Go through each input line */
3719 while (l < lend) {
3720 UV min, max, val;
3721 UV inverse;
3722 l = S_swash_scan_list_line(aTHX_ l, lend, &min, &max, &val,
3723 cBOOL(octets), typestr);
3724 if (l > lend) {
3725 break;
3726 }
3727
3728 /* Each element in the range is to be inverted */
3729 for (inverse = min; inverse <= max; inverse++) {
3730 AV* list;
064c021d
KW
3731 SV** listp;
3732 IV i;
3733 bool found_key = FALSE;
5662e334 3734 bool found_inverse = FALSE;
064c021d
KW
3735
3736 /* The key is the inverse mapping */
3737 char key[UTF8_MAXBYTES+1];
3738 char* key_end = (char *) uvuni_to_utf8((U8*) key, val);
3739 STRLEN key_len = key_end - key;
3740
064c021d
KW
3741 /* Get the list for the map */
3742 if ((listp = hv_fetch(ret, key, key_len, FALSE))) {
3743 list = (AV*) *listp;
3744 }
3745 else { /* No entry yet for it: create one */
3746 list = newAV();
3747 if (! hv_store(ret, key, key_len, (SV*) list, FALSE)) {
3748 Perl_croak(aTHX_ "panic: hv_store() unexpectedly failed");
3749 }
3750 }
3751
5662e334
KW
3752 /* Look through list to see if this inverse mapping already is
3753 * listed, or if there is a mapping to itself already */
508f7cfa 3754 for (i = 0; i <= av_len(list); i++) {
064c021d
KW
3755 SV** entryp = av_fetch(list, i, FALSE);
3756 SV* entry;
3757 if (entryp == NULL) {
3758 Perl_croak(aTHX_ "panic: av_fetch() unexpectedly failed");
3759 }
3760 entry = *entryp;
5662e334 3761 /*DEBUG_U(PerlIO_printf(Perl_debug_log, "list for %"UVXf" contains %"UVXf"\n", val, SvUV(entry)));*/
56ca34ca 3762 if (SvUV(entry) == val) {
064c021d 3763 found_key = TRUE;
5662e334
KW
3764 }
3765 if (SvUV(entry) == inverse) {
3766 found_inverse = TRUE;
3767 }
3768
3769 /* No need to continue searching if found everything we are
3770 * looking for */
3771 if (found_key && found_inverse) {
064c021d
KW
3772 break;
3773 }
3774 }
56ca34ca
KW
3775
3776 /* Make sure there is a mapping to itself on the list */
064c021d 3777 if (! found_key) {
d397ff6a 3778 av_push(list, newSVuv(val));
5662e334 3779 /*DEBUG_U(PerlIO_printf(Perl_debug_log, "Adding %"UVXf" to list for %"UVXf"\n", val, val));*/
064c021d
KW
3780 }
3781
3782
3783 /* Simply add the value to the list */
5662e334
KW
3784 if (! found_inverse) {
3785 av_push(list, newSVuv(inverse));
3786 /*DEBUG_U(PerlIO_printf(Perl_debug_log, "Adding %"UVXf" to list for %"UVXf"\n", inverse, val));*/
3787 }
064c021d 3788
b0e3252e 3789 /* swatch_get() increments the value of val for each element in the
064c021d
KW
3790 * range. That makes more compact tables possible. You can
3791 * express the capitalization, for example, of all consecutive
3792 * letters with a single line: 0061\t007A\t0041 This maps 0061 to
3793 * 0041, 0062 to 0042, etc. I (khw) have never understood 'none',
bd3f2f94 3794 * and it's not documented; it appears to be used only in
b0e3252e 3795 * implementing tr//; I copied the semantics from swatch_get(), just
bd3f2f94 3796 * in case */
064c021d
KW
3797 if (!none || val < none) {
3798 ++val;
3799 }
3800 }
3801 }
3802
3803 return ret;
3804}
3805
a25abddc 3806SV*
d764b54e
KW
3807Perl__swash_to_invlist(pTHX_ SV* const swash)
3808{
3809
3810 /* Subject to change or removal. For use only in one place in regcomp.c */
3811
3812 U8 *l, *lend;
3813 char *loc;
3814 STRLEN lcur;
3815 HV *const hv = MUTABLE_HV(SvRV(swash));
3816 UV elements = 0; /* Number of elements in the inversion list */
b443038a 3817 U8 empty[] = "";
d764b54e
KW
3818
3819 /* The string containing the main body of the table */
3820 SV** const listsvp = hv_fetchs(hv, "LIST", FALSE);
3821 SV** const typesvp = hv_fetchs(hv, "TYPE", FALSE);
3822 SV** const bitssvp = hv_fetchs(hv, "BITS", FALSE);
d73c39c5 3823 SV** const extssvp = hv_fetchs(hv, "EXTRAS", FALSE);
77f9f126 3824 SV** const invert_it_svp = hv_fetchs(hv, "INVERT_IT", FALSE);
d764b54e
KW
3825
3826 const U8* const typestr = (U8*)SvPV_nolen(*typesvp);
3827 const STRLEN bits = SvUV(*bitssvp);
3828 const STRLEN octets = bits >> 3; /* if bits == 1, then octets == 0 */
d73c39c5
KW
3829 U8 *x, *xend;
3830 STRLEN xcur;
d764b54e 3831
a25abddc 3832 SV* invlist;
d764b54e
KW
3833
3834 PERL_ARGS_ASSERT__SWASH_TO_INVLIST;
3835
3836 /* read $swash->{LIST} */
b443038a
KW
3837 if (SvPOK(*listsvp)) {
3838 l = (U8*)SvPV(*listsvp, lcur);
3839 }
3840 else {
3841 /* LIST legitimately doesn't contain a string during compilation phases
3842 * of Perl itself, before the Unicode tables are generated. In this
3843 * case, just fake things up by creating an empty list */
3844 l = empty;
3845 lcur = 0;
3846 }
d764b54e
KW
3847 loc = (char *) l;
3848 lend = l + lcur;
3849
3850 /* Scan the input to count the number of lines to preallocate array size
3851 * based on worst possible case, which is each line in the input creates 2
3852 * elements in the inversion list: 1) the beginning of a range in the list;
3853 * 2) the beginning of a range not in the list. */
3854 while ((loc = (strchr(loc, '\n'))) != NULL) {
3855 elements += 2;
3856 loc++;
3857 }
3858
3859 /* If the ending is somehow corrupt and isn't a new line, add another
3860 * element for the final range that isn't in the inversion list */
fd05e003
KW
3861 if (! (*lend == '\n'
3862 || (*lend == '\0' && (lcur == 0 || *(lend - 1) == '\n'))))
3863 {
d764b54e
KW
3864 elements++;
3865 }
3866
3867 invlist = _new_invlist(elements);
3868
3869 /* Now go through the input again, adding each range to the list */
3870 while (l < lend) {
3871 UV start, end;
3872 UV val; /* Not used by this function */
3873
3874 l = S_swash_scan_list_line(aTHX_ l, lend, &start, &end, &val,
3875 cBOOL(octets), typestr);
3876
3877 if (l > lend) {
3878 break;
3879 }
3880
9d501133 3881 invlist = _add_range_to_invlist(invlist, start, end);
d764b54e
KW
3882 }
3883
77f9f126
KW
3884 /* Invert if the data says it should be */
3885 if (invert_it_svp && SvUV(*invert_it_svp)) {
45bb2768 3886 _invlist_invert_prop(invlist);
77f9f126
KW
3887 }
3888
b0e3252e 3889 /* This code is copied from swatch_get()
d73c39c5
KW
3890 * read $swash->{EXTRAS} */
3891 x = (U8*)SvPV(*extssvp, xcur);
3892 xend = x + xcur;
3893 while (x < xend) {
3894 STRLEN namelen;
3895 U8 *namestr;
3896 SV** othersvp;
3897 HV* otherhv;
3898 STRLEN otherbits;
3899 SV **otherbitssvp, *other;
3900 U8 *nl;
3901
3902 const U8 opc = *x++;
3903 if (opc == '\n')
3904 continue;
3905
3906 nl = (U8*)memchr(x, '\n', xend - x);
3907
3908 if (opc != '-' && opc != '+' && opc != '!' && opc != '&') {
3909 if (nl) {
3910 x = nl + 1; /* 1 is length of "\n" */
3911 continue;
3912 }
3913 else {
3914 x = xend; /* to EXTRAS' end at which \n is not found */
3915 break;
3916 }
3917 }
3918
3919 namestr = x;
3920 if (nl) {
3921 namelen = nl - namestr;
3922 x = nl + 1;
3923 }
3924 else {
3925 namelen = xend - namestr;
3926 x = xend;
3927 }
3928
3929 othersvp = hv_fetch(hv, (char *)namestr, namelen, FALSE);
3930 otherhv = MUTABLE_HV(SvRV(*othersvp));
3931 otherbitssvp = hv_fetchs(otherhv, "BITS", FALSE);
3932 otherbits = (STRLEN)SvUV(*otherbitssvp);
3933
3934 if (bits != otherbits || bits != 1) {
5637ef5b
NC
3935 Perl_croak(aTHX_ "panic: _swash_to_invlist only operates on boolean "
3936 "properties, bits=%"UVuf", otherbits=%"UVuf,
3937 (UV)bits, (UV)otherbits);
d73c39c5
KW
3938 }
3939
3940 /* The "other" swatch must be destroyed after. */
3941 other = _swash_to_invlist((SV *)*othersvp);
3942
b0e3252e 3943 /* End of code copied from swatch_get() */
d73c39c5
KW
3944 switch (opc) {
3945 case '+':
3946 _invlist_union(invlist, other, &invlist);
3947 break;
3948 case '!':
3949 _invlist_invert(other);
3950 _invlist_union(invlist, other, &invlist);
3951 break;
3952 case '-':
3953 _invlist_subtract(invlist, other, &invlist);
3954 break;
3955 case '&':
3956 _invlist_intersection(invlist, other, &invlist);
3957 break;
3958 default:
3959 break;
3960 }
3961 sv_free(other); /* through with it! */
3962 }
3963
d764b54e
KW
3964 return invlist;
3965}
3966
0f830e0b 3967/*
87cea99e 3968=for apidoc uvchr_to_utf8
0f830e0b 3969
6ee84de2 3970Adds the UTF-8 representation of the Native code point C<uv> to the end
e0aa61c6 3971of the string C<d>; C<d> should have at least C<UTF8_MAXBYTES+1> free
0f830e0b
NC
3972bytes available. The return value is the pointer to the byte after the
3973end of the new character. In other words,
3974
3975 d = uvchr_to_utf8(d, uv);
3976
3977is the recommended wide native character-aware way of saying
3978
3979 *(d++) = uv;
3980
3981=cut
3982*/
3983
3984/* On ASCII machines this is normally a macro but we want a
3985 real function in case XS code wants it
3986*/
0f830e0b
NC
3987U8 *
3988Perl_uvchr_to_utf8(pTHX_ U8 *d, UV uv)
3989{
7918f24d
NC
3990 PERL_ARGS_ASSERT_UVCHR_TO_UTF8;
3991
0f830e0b
NC
3992 return Perl_uvuni_to_utf8_flags(aTHX_ d, NATIVE_TO_UNI(uv), 0);
3993}
3994
b851fbc1
JH
3995U8 *
3996Perl_uvchr_to_utf8_flags(pTHX_ U8 *d, UV uv, UV flags)
3997{
7918f24d
NC
3998 PERL_ARGS_ASSERT_UVCHR_TO_UTF8_FLAGS;
3999
b851fbc1
JH
4000 return Perl_uvuni_to_utf8_flags(aTHX_ d, NATIVE_TO_UNI(uv), flags);
4001}
2b9d42f0
NIS
4002
4003/*
87cea99e 4004=for apidoc utf8n_to_uvchr
0f830e0b 4005
48ef279e 4006Returns the native character value of the first character in the string
0f830e0b
NC
4007C<s>
4008which is assumed to be in UTF-8 encoding; C<retlen> will be set to the
4009length, in bytes, of that character.
4010
a1433954 4011C<length> and C<flags> are the same as L</utf8n_to_uvuni>().
0f830e0b
NC
4012
4013=cut
4014*/
4015/* On ASCII machines this is normally a macro but we want
4016 a real function in case XS code wants it
4017*/
0f830e0b 4018UV
48ef279e 4019Perl_utf8n_to_uvchr(pTHX_ const U8 *s, STRLEN curlen, STRLEN *retlen,
0f830e0b
NC
4020U32 flags)
4021{
4022 const UV uv = Perl_utf8n_to_uvuni(aTHX_ s, curlen, retlen, flags);
7918f24d
NC
4023
4024 PERL_ARGS_ASSERT_UTF8N_TO_UVCHR;
4025
0f830e0b
NC
4026 return UNI_TO_NATIVE(uv);
4027}
4028
0876b9a0
KW
4029bool
4030Perl_check_utf8_print(pTHX_ register const U8* s, const STRLEN len)
4031{
4032 /* May change: warns if surrogates, non-character code points, or
af2af982
KW
4033 * non-Unicode code points are in s which has length len bytes. Returns
4034 * TRUE if none found; FALSE otherwise. The only other validity check is
4035 * to make sure that this won't exceed the string's length */
0876b9a0
KW
4036
4037 const U8* const e = s + len;
4038 bool ok = TRUE;
4039
4040 PERL_ARGS_ASSERT_CHECK_UTF8_PRINT;
4041
4042 while (s < e) {
4043 if (UTF8SKIP(s) > len) {
4044 Perl_ck_warner_d(aTHX_ packWARN(WARN_UTF8),
4045 "%s in %s", unees, PL_op ? OP_DESC(PL_op) : "print");
4046 return FALSE;
4047 }
732fbc05 4048 if (UNLIKELY(*s >= UTF8_FIRST_PROBLEMATIC_CODE_POINT_FIRST_BYTE)) {
0876b9a0
KW
4049 STRLEN char_len;
4050 if (UTF8_IS_SUPER(s)) {
8457b38f 4051 if (ckWARN_d(WARN_NON_UNICODE)) {
4b88fb76 4052 UV uv = utf8_to_uvchr_buf(s, e, &char_len);
8457b38f
KW
4053 Perl_warner(aTHX_ packWARN(WARN_NON_UNICODE),
4054 "Code point 0x%04"UVXf" is not Unicode, may not be portable", uv);
4055 ok = FALSE;
4056 }
0876b9a0
KW
4057 }
4058 else if (UTF8_IS_SURROGATE(s)) {
8457b38f 4059 if (ckWARN_d(WARN_SURROGATE)) {
4b88fb76 4060 UV uv = utf8_to_uvchr_buf(s, e, &char_len);
8457b38f
KW
4061 Perl_warner(aTHX_ packWARN(WARN_SURROGATE),
4062 "Unicode surrogate U+%04"UVXf" is illegal in UTF-8", uv);
4063 ok = FALSE;
4064 }
0876b9a0
KW
4065 }
4066 else if
8457b38f
KW
4067 ((UTF8_IS_NONCHAR_GIVEN_THAT_NON_SUPER_AND_GE_PROBLEMATIC(s))
4068 && (ckWARN_d(WARN_NONCHAR)))
0876b9a0 4069 {
4b88fb76 4070 UV uv = utf8_to_uvchr_buf(s, e, &char_len);
8457b38f 4071 Perl_warner(aTHX_ packWARN(WARN_NONCHAR),
0876b9a0
KW
4072 "Unicode non-character U+%04"UVXf" is illegal for open interchange", uv);
4073 ok = FALSE;
4074 }
4075 }
4076 s += UTF8SKIP(s);
4077 }
4078
4079 return ok;
4080}
4081
0f830e0b 4082/*
87cea99e 4083=for apidoc pv_uni_display
d2cc3551 4084
a1433954
KW
4085Build to the scalar C<dsv> a displayable version of the string C<spv>,
4086length C<len>, the displayable version being at most C<pvlim> bytes long
d2cc3551 4087(if longer, the rest is truncated and "..." will be appended).
0a2ef054 4088
a1433954 4089The C<flags> argument can have UNI_DISPLAY_ISPRINT set to display
00e86452 4090isPRINT()able characters as themselves, UNI_DISPLAY_BACKSLASH
0a2ef054
JH
4091to display the \\[nrfta\\] as the backslashed versions (like '\n')
4092(UNI_DISPLAY_BACKSLASH is preferred over UNI_DISPLAY_ISPRINT for \\).
4093UNI_DISPLAY_QQ (and its alias UNI_DISPLAY_REGEX) have both
4094UNI_DISPLAY_BACKSLASH and UNI_DISPLAY_ISPRINT turned on.
4095
a1433954 4096The pointer to the PV of the C<dsv> is returned.
d2cc3551
JH
4097
4098=cut */
e6b2e755 4099char *
e1ec3a88 4100Perl_pv_uni_display(pTHX_ SV *dsv, const U8 *spv, STRLEN len, STRLEN pvlim, UV flags)
e6b2e755
JH
4101{
4102 int truncated = 0;
e1ec3a88 4103 const char *s, *e;
e6b2e755 4104
7918f24d
NC
4105 PERL_ARGS_ASSERT_PV_UNI_DISPLAY;
4106
76f68e9b 4107 sv_setpvs(dsv, "");
7fddd944 4108 SvUTF8_off(dsv);
e1ec3a88 4109 for (s = (const char *)spv, e = s + len; s < e; s += UTF8SKIP(s)) {
e6b2e755 4110 UV u;
a49f32c6
NC
4111 /* This serves double duty as a flag and a character to print after
4112 a \ when flags & UNI_DISPLAY_BACKSLASH is true.
4113 */
4114 char ok = 0;
c728cb41 4115
e6b2e755
JH
4116 if (pvlim && SvCUR(dsv) >= pvlim) {
4117 truncated++;
4118 break;
4119 }
4b88fb76 4120 u = utf8_to_uvchr_buf((U8*)s, (U8*)e, 0);
c728cb41 4121 if (u < 256) {
a3b680e6 4122 const unsigned char c = (unsigned char)u & 0xFF;
0bd48802 4123 if (flags & UNI_DISPLAY_BACKSLASH) {
a49f32c6 4124 switch (c) {
c728cb41 4125 case '\n':
a49f32c6 4126 ok = 'n'; break;
c728cb41 4127 case '\r':
a49f32c6 4128 ok = 'r'; break;
c728cb41 4129 case '\t':
a49f32c6 4130 ok = 't'; break;
c728cb41 4131 case '\f':
a49f32c6 4132 ok = 'f'; break;
c728cb41 4133 case '\a':
a49f32c6 4134 ok = 'a'; break;
c728cb41 4135 case '\\':
a49f32c6 4136 ok = '\\'; break;
c728cb41
JH
4137 default: break;
4138 }
a49f32c6 4139 if (ok) {
88c9ea1e 4140 const char string = ok;
76f68e9b 4141 sv_catpvs(dsv, "\\");
5e7aa789 4142 sv_catpvn(dsv, &string, 1);
a49f32c6 4143 }
c728cb41 4144 }
00e86452 4145 /* isPRINT() is the locale-blind version. */
a49f32c6 4146 if (!ok && (flags & UNI_DISPLAY_ISPRINT) && isPRINT(c)) {
88c9ea1e 4147 const char string = c;
5e7aa789 4148 sv_catpvn(dsv, &string, 1);
a49f32c6 4149 ok = 1;
0a2ef054 4150 }
c728cb41
JH
4151 }
4152 if (!ok)
9e55ce06 4153 Perl_sv_catpvf(aTHX_ dsv, "\\x{%"UVxf"}", u);
e6b2e755
JH
4154 }
4155 if (truncated)
396482e1 4156 sv_catpvs(dsv, "...");
48ef279e 4157
e6b2e755
JH
4158 return SvPVX(dsv);
4159}
2b9d42f0 4160
d2cc3551 4161/*
87cea99e 4162=for apidoc sv_uni_display
d2cc3551 4163
a1433954
KW
4164Build to the scalar C<dsv> a displayable version of the scalar C<sv>,
4165the displayable version being at most C<pvlim> bytes long
d2cc3551 4166(if longer, the rest is truncated and "..." will be appended).
0a2ef054 4167
a1433954 4168The C<flags> argument is as in L</pv_uni_display>().
0a2ef054 4169
a1433954 4170The pointer to the PV of the C<dsv> is returned.
d2cc3551 4171
d4c19fe8
AL
4172=cut
4173*/
e6b2e755
JH
4174char *
4175Perl_sv_uni_display(pTHX_ SV *dsv, SV *ssv, STRLEN pvlim, UV flags)
4176{
7918f24d
NC
4177 PERL_ARGS_ASSERT_SV_UNI_DISPLAY;
4178
cfd0369c
NC
4179 return Perl_pv_uni_display(aTHX_ dsv, (const U8*)SvPVX_const(ssv),
4180 SvCUR(ssv), pvlim, flags);
701a277b
JH
4181}
4182
d2cc3551 4183/*
e6226b18 4184=for apidoc foldEQ_utf8
d2cc3551 4185
a1433954 4186Returns true if the leading portions of the strings C<s1> and C<s2> (either or both
e6226b18 4187of which may be in UTF-8) are the same case-insensitively; false otherwise.
d51c1b21 4188How far into the strings to compare is determined by other input parameters.
8b35872c 4189
a1433954
KW
4190If C<u1> is true, the string C<s1> is assumed to be in UTF-8-encoded Unicode;
4191otherwise it is assumed to be in native 8-bit encoding. Correspondingly for C<u2>
4192with respect to C<s2>.
8b35872c 4193
a1433954
KW
4194If the byte length C<l1> is non-zero, it says how far into C<s1> to check for fold
4195equality. In other words, C<s1>+C<l1> will be used as a goal to reach. The
8b35872c 4196scan will not be considered to be a match unless the goal is reached, and
a1433954
KW
4197scanning won't continue past that goal. Correspondingly for C<l2> with respect to
4198C<s2>.
4199
4200If C<pe1> is non-NULL and the pointer it points to is not NULL, that pointer is
4201considered an end pointer beyond which scanning of C<s1> will not continue under
4202any circumstances. This means that if both C<l1> and C<pe1> are specified, and
4203C<pe1>
4204is less than C<s1>+C<l1>, the match will never be successful because it can
4205never
d51c1b21 4206get as far as its goal (and in fact is asserted against). Correspondingly for
a1433954 4207C<pe2> with respect to C<s2>.
8b35872c 4208
a1433954
KW
4209At least one of C<s1> and C<s2> must have a goal (at least one of C<l1> and
4210C<l2> must be non-zero), and if both do, both have to be
8b35872c
KW
4211reached for a successful match. Also, if the fold of a character is multiple
4212characters, all of them must be matched (see tr21 reference below for
4213'folding').
4214
a1433954
KW
4215Upon a successful match, if C<pe1> is non-NULL,
4216it will be set to point to the beginning of the I<next> character of C<s1>
4217beyond what was matched. Correspondingly for C<pe2> and C<s2>.
d2cc3551
JH
4218
4219For case-insensitiveness, the "casefolding" of Unicode is used
4220instead of upper/lowercasing both the characters, see
a1433954 4221L<http://www.unicode.org/unicode/reports/tr21/> (Case Mappings).
d2cc3551
JH
4222
4223=cut */
a33c29bc
KW
4224
4225/* A flags parameter has been added which may change, and hence isn't
4226 * externally documented. Currently it is:
4227 * 0 for as-documented above
4228 * FOLDEQ_UTF8_NOMIX_ASCII meaning that if a non-ASCII character folds to an
4229 ASCII one, to not match
5e64d0fa
KW
4230 * FOLDEQ_UTF8_LOCALE meaning that locale rules are to be used for code
4231 * points below 256; unicode rules for above 255; and
4232 * folds that cross those boundaries are disallowed,
4233 * like the NOMIX_ASCII option
18f762c3
KW
4234 * FOLDEQ_S1_ALREADY_FOLDED s1 has already been folded before calling this
4235 * routine. This allows that step to be skipped.
4236 * FOLDEQ_S2_ALREADY_FOLDED Similarly.
a33c29bc 4237 */
701a277b 4238I32
eda9cac1 4239Perl_foldEQ_utf8_flags(pTHX_ const char *s1, char **pe1, register UV l1, bool u1, const char *s2, char **pe2, register UV l2, bool u2, U32 flags)
332ddc25 4240{
8b35872c
KW
4241 dVAR;
4242 register const U8 *p1 = (const U8*)s1; /* Point to current char */
4243 register const U8 *p2 = (const U8*)s2;
48ef279e 4244 register const U8 *g1 = NULL; /* goal for s1 */
8b35872c 4245 register const U8 *g2 = NULL;
48ef279e
KW
4246 register const U8 *e1 = NULL; /* Don't scan s1 past this */
4247 register U8 *f1 = NULL; /* Point to current folded */
8b35872c
KW
4248 register const U8 *e2 = NULL;
4249 register U8 *f2 = NULL;
48ef279e 4250 STRLEN n1 = 0, n2 = 0; /* Number of bytes in current char */
8b35872c
KW
4251 U8 foldbuf1[UTF8_MAXBYTES_CASE+1];
4252 U8 foldbuf2[UTF8_MAXBYTES_CASE+1];
8b35872c 4253
eda9cac1 4254 PERL_ARGS_ASSERT_FOLDEQ_UTF8_FLAGS;
8b35872c 4255
18f762c3
KW
4256 /* The algorithm requires that input with the flags on the first line of
4257 * the assert not be pre-folded. */
4258 assert( ! ((flags & (FOLDEQ_UTF8_NOMIX_ASCII | FOLDEQ_UTF8_LOCALE))
4259 && (flags & (FOLDEQ_S1_ALREADY_FOLDED | FOLDEQ_S2_ALREADY_FOLDED))));
4260
8b35872c 4261 if (pe1) {
48ef279e 4262 e1 = *(U8**)pe1;
8b35872c
KW
4263 }
4264
4265 if (l1) {
48ef279e 4266 g1 = (const U8*)s1 + l1;
8b35872c
KW
4267 }
4268
4269 if (pe2) {
48ef279e 4270 e2 = *(U8**)pe2;
8b35872c
KW
4271 }
4272
4273 if (l2) {
48ef279e 4274 g2 = (const U8*)s2 + l2;
8b35872c
KW
4275 }
4276
4277 /* Must have at least one goal */
4278 assert(g1 || g2);
4279
4280 if (g1) {
4281
48ef279e
KW
4282 /* Will never match if goal is out-of-bounds */
4283 assert(! e1 || e1 >= g1);
8b35872c 4284
48ef279e
KW
4285 /* Here, there isn't an end pointer, or it is beyond the goal. We
4286 * only go as far as the goal */
4287 e1 = g1;
8b35872c 4288 }
313b38e5
NC
4289 else {
4290 assert(e1); /* Must have an end for looking at s1 */
4291 }
8b35872c
KW
4292
4293 /* Same for goal for s2 */
4294 if (g2) {
48ef279e
KW
4295 assert(! e2 || e2 >= g2);
4296 e2 = g2;
8b35872c 4297 }
313b38e5
NC
4298 else {
4299 assert(e2);
4300 }
8b35872c 4301
18f762c3
KW
4302 /* If both operands are already folded, we could just do a memEQ on the
4303 * whole strings at once, but it would be better if the caller realized
4304 * this and didn't even call us */
4305
8b35872c
KW
4306 /* Look through both strings, a character at a time */
4307 while (p1 < e1 && p2 < e2) {
4308
d51c1b21 4309 /* If at the beginning of a new character in s1, get its fold to use
5e64d0fa
KW
4310 * and the length of the fold. (exception: locale rules just get the
4311 * character to a single byte) */
48ef279e 4312 if (n1 == 0) {
18f762c3
KW
4313 if (flags & FOLDEQ_S1_ALREADY_FOLDED) {
4314 f1 = (U8 *) p1;
4315 n1 = UTF8SKIP(f1);
18f762c3 4316 }
a6d5f321 4317
18f762c3 4318 else {
a6d5f321
KW
4319 /* If in locale matching, we use two sets of rules, depending
4320 * on if the code point is above or below 255. Here, we test
4321 * for and handle locale rules */
1b4059d5
KW
4322 if ((flags & FOLDEQ_UTF8_LOCALE)
4323 && (! u1 || UTF8_IS_INVARIANT(*p1)
4324 || UTF8_IS_DOWNGRADEABLE_START(*p1)))
5e64d0fa 4325 {
1b4059d5
KW
4326 /* There is no mixing of code points above and below 255. */
4327 if (u2 && (! UTF8_IS_INVARIANT(*p2)
4328 && ! UTF8_IS_DOWNGRADEABLE_START(*p2)))
4329 {
4330 return 0;
4331 }
5e64d0fa 4332
1b4059d5
KW
4333 /* We handle locale rules by converting, if necessary, the
4334 * code point to a single byte. */
4335 if (! u1 || UTF8_IS_INVARIANT(*p1)) {
4336 *foldbuf1 = *p1;
4337 }
4338 else {
4339 *foldbuf1 = TWO_BYTE_UTF8_TO_UNI(*p1, *(p1 + 1));
4340 }
4341 n1 = 1;
5e64d0fa 4342 }
a6d5f321
KW
4343 else if (isASCII(*p1)) { /* Note, that here won't be both
4344 ASCII and using locale rules */
1b4059d5
KW
4345
4346 /* If trying to mix non- with ASCII, and not supposed to,
4347 * fail */
4348 if ((flags & FOLDEQ_UTF8_NOMIX_ASCII) && ! isASCII(*p2)) {
4349 return 0;
4350 }
4351 n1 = 1;
4352 *foldbuf1 = toLOWER(*p1); /* Folds in the ASCII range are
4353 just lowercased */
5e64d0fa 4354 }
1b4059d5
KW
4355 else if (u1) {
4356 to_utf8_fold(p1, foldbuf1, &n1);
a33c29bc 4357 }
9cef8533
KW
4358 else { /* Not utf8, get utf8 fold */
4359 to_uni_fold(NATIVE_TO_UNI(*p1), foldbuf1, &n1);
1b4059d5
KW
4360 }
4361 f1 = foldbuf1;
18f762c3 4362 }
48ef279e 4363 }
8b35872c 4364
48ef279e 4365 if (n2 == 0) { /* Same for s2 */
18f762c3
KW
4366 if (flags & FOLDEQ_S2_ALREADY_FOLDED) {
4367 f2 = (U8 *) p2;
4368 n2 = UTF8SKIP(f2);
4369 }
4370 else {
227968da
KW
4371 if ((flags & FOLDEQ_UTF8_LOCALE)
4372 && (! u2 || UTF8_IS_INVARIANT(*p2) || UTF8_IS_DOWNGRADEABLE_START(*p2)))
5e64d0fa 4373 {
227968da
KW
4374 /* Here, the next char in s2 is < 256. We've already
4375 * worked on s1, and if it isn't also < 256, can't match */
4376 if (u1 && (! UTF8_IS_INVARIANT(*p1)
4377 && ! UTF8_IS_DOWNGRADEABLE_START(*p1)))
4378 {
4379 return 0;
4380 }
4381 if (! u2 || UTF8_IS_INVARIANT(*p2)) {
4382 *foldbuf2 = *p2;
4383 }
4384 else {
4385 *foldbuf2 = TWO_BYTE_UTF8_TO_UNI(*p2, *(p2 + 1));
4386 }
4387
4388 /* Use another function to handle locale rules. We've made
4389 * sure that both characters to compare are single bytes */
4390 if (! foldEQ_locale((char *) f1, (char *) foldbuf2, 1)) {
4391 return 0;
4392 }
4393 n1 = n2 = 0;
5e64d0fa 4394 }
227968da 4395 else if (isASCII(*p2)) {
ba9114af 4396 if ((flags & FOLDEQ_UTF8_NOMIX_ASCII) && ! isASCII(*p1)) {
227968da
KW
4397 return 0;
4398 }
4399 n2 = 1;
4400 *foldbuf2 = toLOWER(*p2);
5e64d0fa 4401 }
227968da
KW
4402 else if (u2) {
4403 to_utf8_fold(p2, foldbuf2, &n2);
5001101e 4404 }
227968da 4405 else {
9cef8533 4406 to_uni_fold(NATIVE_TO_UNI(*p2), foldbuf2, &n2);
a33c29bc 4407 }
227968da 4408 f2 = foldbuf2;
18f762c3 4409 }
48ef279e 4410 }
8b35872c 4411
5001101e 4412 /* Here f1 and f2 point to the beginning of the strings to compare.
227968da
KW
4413 * These strings are the folds of the next character from each input
4414 * string, stored in utf8. */
5e64d0fa 4415
48ef279e
KW
4416 /* While there is more to look for in both folds, see if they
4417 * continue to match */
4418 while (n1 && n2) {
4419 U8 fold_length = UTF8SKIP(f1);
4420 if (fold_length != UTF8SKIP(f2)
4421 || (fold_length == 1 && *f1 != *f2) /* Short circuit memNE
4422 function call for single
a6d5f321 4423 byte */
48ef279e
KW
4424 || memNE((char*)f1, (char*)f2, fold_length))
4425 {
e6226b18 4426 return 0; /* mismatch */
48ef279e
KW
4427 }
4428
4429 /* Here, they matched, advance past them */
4430 n1 -= fold_length;
4431 f1 += fold_length;
4432 n2 -= fold_length;
4433 f2 += fold_length;
4434 }
8b35872c 4435
48ef279e
KW
4436 /* When reach the end of any fold, advance the input past it */
4437 if (n1 == 0) {
4438 p1 += u1 ? UTF8SKIP(p1) : 1;
4439 }
4440 if (n2 == 0) {
4441 p2 += u2 ? UTF8SKIP(p2) : 1;
4442 }
8b35872c
KW
4443 } /* End of loop through both strings */
4444
4445 /* A match is defined by each scan that specified an explicit length
4446 * reaching its final goal, and the other not having matched a partial
4447 * character (which can happen when the fold of a character is more than one
4448 * character). */
4449 if (! ((g1 == 0 || p1 == g1) && (g2 == 0 || p2 == g2)) || n1 || n2) {
e6226b18 4450 return 0;
8b35872c
KW
4451 }
4452
4453 /* Successful match. Set output pointers */
4454 if (pe1) {
48ef279e 4455 *pe1 = (char*)p1;
8b35872c
KW
4456 }
4457 if (pe2) {
48ef279e 4458 *pe2 = (char*)p2;
8b35872c 4459 }
e6226b18 4460 return 1;
e6b2e755 4461}
701a277b 4462
a49f32c6
NC
4463/*
4464 * Local variables:
4465 * c-indentation-style: bsd
4466 * c-basic-offset: 4
4467 * indent-tabs-mode: t
4468 * End:
4469 *
37442d52
RGS
4470 * ex: set ts=8 sts=4 sw=4 noet:
4471 */