1 /* $NetBSD: citrus_viqr.c,v 1.3 2006/11/22 20:11:03 tnozaki Exp $ */ 2 3 /*- 4 * Copyright (c)2006 Citrus Project, 5 * All rights reserved. 6 * 7 * Redistribution and use in source and binary forms, with or without 8 * modification, are permitted provided that the following conditions 9 * are met: 10 * 1. Redistributions of source code must retain the above copyright 11 * notice, this list of conditions and the following disclaimer. 12 * 2. Redistributions in binary form must reproduce the above copyright 13 * notice, this list of conditions and the following disclaimer in the 14 * documentation and/or other materials provided with the distribution. 15 * 16 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND 17 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 18 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 19 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE 20 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 21 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 22 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 23 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 24 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 25 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 26 * SUCH DAMAGE. 27 * 28 */ 29 30 #include <sys/cdefs.h> 31 #if defined(LIBC_SCCS) && !defined(lint) 32 __RCSID("$NetBSD: citrus_viqr.c,v 1.3 2006/11/22 20:11:03 tnozaki Exp $"); 33 #endif /* LIBC_SCCS and not lint */ 34 35 #include <sys/queue.h> 36 #include <sys/types.h> 37 #include <assert.h> 38 #include <errno.h> 39 #include <string.h> 40 #include <stdint.h> 41 #include <stdlib.h> 42 #include <stddef.h> 43 #include <locale.h> 44 #include <wchar.h> 45 #include <limits.h> 46 47 #include "citrus_namespace.h" 48 #include "citrus_types.h" 49 #include "citrus_bcs.h" 50 #include "citrus_module.h" 51 #include "citrus_ctype.h" 52 #include "citrus_stdenc.h" 53 #include "citrus_viqr.h" 54 55 #define ESCAPE '\\' 56 57 /* 58 * this table generated from RFC 1456. 59 */ 60 static const char *mnemonic_rfc1456[0x100] = { 61 NULL , NULL , "A(?", NULL , NULL , "A(~", "A^~", NULL , 62 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 63 NULL , NULL , NULL , NULL , "Y?" , NULL , NULL , NULL , 64 NULL , "Y~" , NULL , NULL , NULL , NULL , "Y." , NULL , 65 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 66 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 67 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 68 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 69 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 70 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 71 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 72 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 73 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 74 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 75 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 76 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 77 "A." , "A('", "A(`", "A(.", "A^'", "A^`", "A^?", "A^.", 78 "E~" , "E." , "E^'", "E^`", "E^?", "E^~", "E^.", "O^'", 79 "O^`", "O^?", "O^~", "O^.", "O+.", "O+'", "O+`", "O+?", 80 "I." , "O?" , "O." , "I?" , "U?" , "U~" , "U." , "Y`" , 81 "O~" , "a('", "a(`", "a(.", "a^'", "a^`", "a^?", "a^.", 82 "e~" , "e." , "e^'", "e^`", "e^?", "e^~", "e^.", "o^'", 83 "o^`", "o^?", "o^~", "O+~", "O+" , "o^.", "o+`", "o+?", 84 "i." , "U+.", "U+'", "U+`", "U+?", "o+" , "o+'", "U+" , 85 "A`" , "A'" , "A^" , "A~" , "A?" , "A(" , "a(?", "a(~", 86 "E`" , "E'" , "E^" , "E?" , "I`" , "I'" , "I~" , "y`" , 87 "DD" , "u+'", "O`" , "O'" , "O^" , "a." , "y?" , "u+`", 88 "u+?", "U`" , "U'" , "y~" , "y." , "Y'" , "o+~", "u+" , 89 "a`" , "a'" , "a^" , "a~" , "a?" , "a(" , "u+~", "a^~", 90 "e`" , "e'" , "e^" , "e?" , "i`" , "i'" , "i~" , "i?" , 91 "dd" , "u+.", "o`" , "o'" , "o^" , "o~" , "o?" , "o." , 92 "u." , "u`" , "u'" , "u~" , "u?" , "y'" , "o+.", "U+~", 93 }; 94 95 typedef struct { 96 const char *name; 97 wchar_t value; 98 } mnemonic_def_t; 99 100 static const mnemonic_def_t mnemonic_ext[] = { 101 /* add extra mnemonic here (should be sorted by wchar_t order). */ 102 }; 103 static const size_t mnemonic_ext_size = 104 sizeof(mnemonic_ext) / sizeof(mnemonic_def_t); 105 106 static const char * 107 mnemonic_ext_find(wchar_t wc, const mnemonic_def_t *head, size_t n) 108 { 109 const mnemonic_def_t *mid; 110 111 _DIAGASSERT(head != NULL); 112 113 for (; n > 0; n >>= 1) { 114 mid = head + (n >> 1); 115 if (mid->value == wc) { 116 return mid->name; 117 } else if (mid->value < wc) { 118 head = mid + 1; 119 --n; 120 } 121 } 122 return NULL; 123 } 124 125 struct mnemonic_t; 126 typedef TAILQ_HEAD(mnemonic_list_t, mnemonic_t) mnemonic_list_t; 127 typedef struct mnemonic_t { 128 TAILQ_ENTRY(mnemonic_t) entry; 129 int ascii; 130 struct mnemonic_t *parent; 131 mnemonic_list_t child; 132 wchar_t value; 133 } mnemonic_t; 134 135 static mnemonic_t * 136 mnemonic_list_find(mnemonic_list_t *ml, int ch) 137 { 138 mnemonic_t *m; 139 140 _DIAGASSERT(ml != NULL); 141 142 TAILQ_FOREACH(m, ml, entry) { 143 if (m->ascii == ch) 144 return m; 145 } 146 147 return NULL; 148 } 149 150 static mnemonic_t * 151 mnemonic_create(mnemonic_t *parent, int ascii, wchar_t value) 152 { 153 mnemonic_t *m; 154 155 _DIAGASSERT(parent != NULL); 156 157 m = malloc(sizeof(*m)); 158 if (m != NULL) { 159 m->parent = parent; 160 m->ascii = ascii; 161 m->value = value; 162 TAILQ_INIT(&m->child); 163 } 164 165 return m; 166 } 167 168 static int 169 mnemonic_append_child(mnemonic_t *m, const char *s, 170 wchar_t value, wchar_t invalid) 171 { 172 int ch; 173 mnemonic_t *m0; 174 175 _DIAGASSERT(m != NULL); 176 _DIAGASSERT(s != NULL); 177 178 ch = (unsigned char)*s++; 179 if (ch == '\0') 180 return EINVAL; 181 m0 = mnemonic_list_find(&m->child, ch); 182 if (m0 == NULL) { 183 m0 = mnemonic_create(m, ch, (wchar_t)ch); 184 if (m0 == NULL) 185 return ENOMEM; 186 TAILQ_INSERT_TAIL(&m->child, m0, entry); 187 } 188 m = m0; 189 for (m0 = NULL; (ch = (unsigned char)*s) != '\0'; ++s) { 190 m0 = mnemonic_list_find(&m->child, ch); 191 if (m0 == NULL) { 192 m0 = mnemonic_create(m, ch, invalid); 193 if (m0 == NULL) 194 return ENOMEM; 195 TAILQ_INSERT_TAIL(&m->child, m0, entry); 196 } 197 m = m0; 198 } 199 if (m0 == NULL) 200 return EINVAL; 201 m0->value = value; 202 203 return 0; 204 } 205 206 static void 207 mnemonic_destroy(mnemonic_t *m) 208 { 209 mnemonic_t *m0; 210 211 _DIAGASSERT(m != NULL); 212 213 TAILQ_FOREACH(m0, &m->child, entry) 214 mnemonic_destroy(m0); 215 free(m); 216 } 217 218 typedef struct { 219 size_t mb_cur_max; 220 wchar_t invalid; 221 mnemonic_t *mroot; 222 } _VIQREncodingInfo; 223 224 typedef struct { 225 int chlen; 226 char ch[MB_LEN_MAX]; 227 } _VIQRState; 228 229 typedef struct { 230 _VIQREncodingInfo ei; 231 struct { 232 /* for future multi-locale facility */ 233 _VIQRState s_mblen; 234 _VIQRState s_mbrlen; 235 _VIQRState s_mbrtowc; 236 _VIQRState s_mbtowc; 237 _VIQRState s_mbsrtowcs; 238 _VIQRState s_wcrtomb; 239 _VIQRState s_wcsrtombs; 240 _VIQRState s_wctomb; 241 } states; 242 } _VIQRCTypeInfo; 243 244 #define _CEI_TO_EI(_cei_) (&(_cei_)->ei) 245 #define _CEI_TO_STATE(_cei_, _func_) (_cei_)->states.s_##_func_ 246 247 #define _FUNCNAME(m) _citrus_VIQR_##m 248 #define _ENCODING_INFO _VIQREncodingInfo 249 #define _CTYPE_INFO _VIQRCTypeInfo 250 #define _ENCODING_STATE _VIQRState 251 #define _ENCODING_MB_CUR_MAX(_ei_) (_ei_)->mb_cur_max 252 #define _ENCODING_IS_STATE_DEPENDENT 1 253 #define _STATE_NEEDS_EXPLICIT_INIT(_ps_) 0 254 255 static __inline void 256 /*ARGSUSED*/ 257 _citrus_VIQR_init_state(_VIQREncodingInfo * __restrict ei, 258 _VIQRState * __restrict psenc) 259 { 260 /* ei may be unused */ 261 _DIAGASSERT(psenc != NULL); 262 263 psenc->chlen = 0; 264 } 265 266 static __inline void 267 /*ARGSUSED*/ 268 _citrus_VIQR_pack_state(_VIQREncodingInfo * __restrict ei, 269 void *__restrict pspriv, const _VIQRState * __restrict psenc) 270 { 271 /* ei may be unused */ 272 _DIAGASSERT(pspriv != NULL); 273 _DIAGASSERT(psenc != NULL); 274 275 memcpy(pspriv, (const void *)psenc, sizeof(*psenc)); 276 } 277 278 static __inline void 279 /*ARGSUSED*/ 280 _citrus_VIQR_unpack_state(_VIQREncodingInfo * __restrict ei, 281 _VIQRState * __restrict psenc, const void * __restrict pspriv) 282 { 283 /* ei may be unused */ 284 _DIAGASSERT(psenc != NULL); 285 _DIAGASSERT(pspriv != NULL); 286 287 memcpy((void *)psenc, pspriv, sizeof(*psenc)); 288 } 289 290 static int 291 _citrus_VIQR_mbrtowc_priv(_VIQREncodingInfo * __restrict ei, 292 wchar_t * __restrict pwc, const char ** __restrict s, size_t n, 293 _VIQRState * __restrict psenc, size_t * __restrict nresult) 294 { 295 const char *s0; 296 wchar_t wc; 297 mnemonic_t *m, *m0; 298 size_t i; 299 int ch, escape; 300 301 _DIAGASSERT(ei != NULL); 302 /* pwc may be null */ 303 _DIAGASSERT(s != NULL); 304 _DIAGASSERT(psenc != NULL); 305 _DIAGASSERT(nresult != NULL); 306 307 if (*s == NULL) { 308 _citrus_VIQR_init_state(ei, psenc); 309 *nresult = (size_t)_ENCODING_IS_STATE_DEPENDENT; 310 return 0; 311 } 312 s0 = *s; 313 314 i = 0; 315 m = ei->mroot; 316 for (escape = 0;;) { 317 if (psenc->chlen == i) { 318 if (n-- < 1) { 319 *s = s0; 320 *nresult = (size_t)-2; 321 return 0; 322 } 323 psenc->ch[psenc->chlen++] = *s0++; 324 } 325 ch = (unsigned char)psenc->ch[i++]; 326 if (ch == ESCAPE) { 327 if (m != ei->mroot) 328 break; 329 escape = 1; 330 continue; 331 } 332 if (escape != 0) 333 break; 334 m0 = mnemonic_list_find(&m->child, ch); 335 if (m0 == NULL) 336 break; 337 m = m0; 338 } 339 while (m != ei->mroot) { 340 --i; 341 if (m->value != ei->invalid) 342 break; 343 m = m->parent; 344 } 345 if (ch == ESCAPE && m != ei->mroot) 346 ++i; 347 psenc->chlen -= i; 348 memmove(&psenc->ch[0], &psenc->ch[i], psenc->chlen); 349 wc = (m == ei->mroot) ? (wchar_t)ch : m->value; 350 if (pwc != NULL) 351 *pwc = wc; 352 *nresult = (size_t)(wc == 0 ? 0 : s0 - *s); 353 *s = s0; 354 355 return 0; 356 } 357 358 static int 359 _citrus_VIQR_wcrtomb_priv(_VIQREncodingInfo * __restrict ei, 360 char * __restrict s, size_t n, wchar_t wc, 361 _VIQRState * __restrict psenc, size_t * __restrict nresult) 362 { 363 mnemonic_t *m; 364 int ch, escape; 365 const char *p; 366 367 _DIAGASSERT(ei != NULL); 368 _DIAGASSERT(s != NULL); 369 _DIAGASSERT(psenc != NULL); 370 _DIAGASSERT(nresult != NULL); 371 372 switch (psenc->chlen) { 373 case 0: case 1: 374 break; 375 default: 376 return EINVAL; 377 } 378 m = NULL; 379 if ((uint32_t)wc <= 0xFF) { 380 p = mnemonic_rfc1456[wc & 0xFF]; 381 if (p != NULL) 382 goto mnemonic_found; 383 if (n-- < 1) 384 goto e2big; 385 ch = (unsigned int)wc; 386 m = ei->mroot; 387 if (psenc->chlen > 0) { 388 m = mnemonic_list_find(&m->child, psenc->ch[0]); 389 if (m == NULL) 390 return EINVAL; 391 psenc->ch[0] = ESCAPE; 392 } 393 if (mnemonic_list_find(&m->child, ch) == NULL) { 394 psenc->chlen = 0; 395 m = NULL; 396 } 397 psenc->ch[psenc->chlen++] = ch; 398 } else { 399 p = mnemonic_ext_find(wc, &mnemonic_ext[0], mnemonic_ext_size); 400 if (p == NULL) { 401 *nresult = (size_t)-1; 402 return EILSEQ; 403 } else { 404 mnemonic_found: 405 psenc->chlen = 0; 406 while (*p != '\0') { 407 if (n-- < 1) 408 goto e2big; 409 psenc->ch[psenc->chlen++] = *p++; 410 } 411 } 412 } 413 memcpy(s, psenc->ch, psenc->chlen); 414 *nresult = psenc->chlen; 415 if (m == ei->mroot) { 416 psenc->ch[0] = ch; 417 psenc->chlen = 1; 418 } else { 419 psenc->chlen = 0; 420 } 421 422 return 0; 423 424 e2big: 425 *nresult = (size_t)-1; 426 return E2BIG; 427 } 428 429 static int 430 /* ARGSUSED */ 431 _citrus_VIQR_put_state_reset(_VIQREncodingInfo * __restrict ei, 432 char * __restrict s, size_t n, _VIQRState * __restrict psenc, 433 size_t * __restrict nresult) 434 { 435 /* ei may be unused */ 436 _DIAGASSERT(s != NULL); 437 _DIAGASSERT(psenc != NULL); 438 _DIAGASSERT(nresult != NULL); 439 440 switch (psenc->chlen) { 441 case 0: case 1: 442 break; 443 default: 444 return EINVAL; 445 } 446 *nresult = 0; 447 psenc->chlen = 0; 448 449 return 0; 450 } 451 452 static __inline int 453 /*ARGSUSED*/ 454 _citrus_VIQR_stdenc_wctocs(_VIQREncodingInfo * __restrict ei, 455 _csid_t * __restrict csid, _index_t * __restrict idx, wchar_t wc) 456 { 457 /* ei may be unused */ 458 _DIAGASSERT(csid != NULL); 459 _DIAGASSERT(idx != NULL); 460 461 *csid = 0; 462 *idx = (_index_t)wc; 463 464 return 0; 465 } 466 467 static __inline int 468 /*ARGSUSED*/ 469 _citrus_VIQR_stdenc_cstowc(_VIQREncodingInfo * __restrict ei, 470 wchar_t * __restrict pwc, _csid_t csid, _index_t idx) 471 { 472 /* ei may be unused */ 473 _DIAGASSERT(pwc != NULL); 474 475 if (csid != 0) 476 return EILSEQ; 477 *pwc = (wchar_t)idx; 478 479 return 0; 480 } 481 482 static void 483 _citrus_VIQR_encoding_module_uninit(_VIQREncodingInfo *ei) 484 { 485 _DIAGASSERT(ei != NULL); 486 487 mnemonic_destroy(ei->mroot); 488 } 489 490 static int 491 /*ARGSUSED*/ 492 _citrus_VIQR_encoding_module_init(_VIQREncodingInfo * __restrict ei, 493 const void * __restrict var, size_t lenvar) 494 { 495 int errnum; 496 const char *s; 497 size_t i, n; 498 const mnemonic_def_t *p; 499 500 _DIAGASSERT(ei != NULL); 501 /* var may be unused */ 502 503 ei->mb_cur_max = 1; 504 ei->invalid = (wchar_t)-1; 505 ei->mroot = mnemonic_create(NULL, '\0', ei->invalid); 506 if (ei->mroot == NULL) 507 return ENOMEM; 508 for (i = 0; i < sizeof(mnemonic_rfc1456) / sizeof(const char *); ++i) { 509 s = mnemonic_rfc1456[i]; 510 if (s == NULL) 511 continue; 512 n = strlen(s); 513 _DIAGASSERT(n <= MB_LEN_MAX); 514 if (ei->mb_cur_max < n) 515 ei->mb_cur_max = n; 516 errnum = mnemonic_append_child(ei->mroot, 517 s, (wchar_t)i, ei->invalid); 518 if (errnum != 0) { 519 _citrus_VIQR_encoding_module_uninit(ei); 520 return errnum; 521 } 522 } 523 for (i = 0; i < mnemonic_ext_size; ++i) { 524 p = &mnemonic_ext[i]; 525 _DIAGASSERT(p != NULL && p->name != NULL); 526 n = strlen(p->name); 527 _DIAGASSERT(n <= MB_LEN_MAX); 528 if (ei->mb_cur_max < n) 529 ei->mb_cur_max = n; 530 errnum = mnemonic_append_child(ei->mroot, 531 p->name, p->value, ei->invalid); 532 if (errnum != 0) { 533 _citrus_VIQR_encoding_module_uninit(ei); 534 return errnum; 535 } 536 } 537 538 return 0; 539 } 540 541 static __inline int 542 /*ARGSUSED*/ 543 _citrus_VIQR_stdenc_get_state_desc_generic(_VIQREncodingInfo * __restrict ei, 544 _VIQRState * __restrict psenc, int * __restrict rstate) 545 { 546 /* ei may be unused */ 547 _DIAGASSERT(psenc != NULL); 548 _DIAGASSERT(rstate != NULL); 549 550 *rstate = (psenc->chlen == 0) 551 ? _STDENC_SDGEN_INITIAL 552 : _STDENC_SDGEN_INCOMPLETE_CHAR; 553 554 return 0; 555 } 556 557 /* ---------------------------------------------------------------------- 558 * public interface for ctype 559 */ 560 561 _CITRUS_CTYPE_DECLS(VIQR); 562 _CITRUS_CTYPE_DEF_OPS(VIQR); 563 564 #include "citrus_ctype_template.h" 565 566 /* ---------------------------------------------------------------------- 567 * public interface for stdenc 568 */ 569 570 _CITRUS_STDENC_DECLS(VIQR); 571 _CITRUS_STDENC_DEF_OPS(VIQR); 572 573 #include "citrus_stdenc_template.h" 574