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