1 /* $FreeBSD: head/lib/libiconv_modules/VIQR/citrus_viqr.c 281550 2015-04-15 09:09:20Z tijl $ */ 2 /* $NetBSD: citrus_viqr.c,v 1.5 2011/11/19 18:20:13 tnozaki Exp $ */ 3 4 /*- 5 * Copyright (c)2006 Citrus Project, 6 * All rights reserved. 7 * 8 * Redistribution and use in source and binary forms, with or without 9 * modification, are permitted provided that the following conditions 10 * are met: 11 * 1. Redistributions of source code must retain the above copyright 12 * notice, this list of conditions and the following disclaimer. 13 * 2. Redistributions in binary form must reproduce the above copyright 14 * notice, this list of conditions and the following disclaimer in the 15 * documentation and/or other materials provided with the distribution. 16 * 17 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND 18 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 20 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE 21 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 22 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 23 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 24 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 25 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 26 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 27 * SUCH DAMAGE. 28 * 29 */ 30 31 #include <sys/cdefs.h> 32 #include <sys/queue.h> 33 #include <sys/types.h> 34 35 #include <assert.h> 36 #include <errno.h> 37 #include <limits.h> 38 #include <stddef.h> 39 #include <stdint.h> 40 #include <stdlib.h> 41 #include <string.h> 42 #include <wchar.h> 43 44 #include "citrus_namespace.h" 45 #include "citrus_types.h" 46 #include "citrus_bcs.h" 47 #include "citrus_module.h" 48 #include "citrus_stdenc.h" 49 #include "citrus_viqr.h" 50 51 #define ESCAPE '\\' 52 53 /* 54 * this table generated from RFC 1456. 55 */ 56 static const char *mnemonic_rfc1456[0x100] = { 57 NULL , NULL , "A(?", NULL , NULL , "A(~", "A^~", NULL , 58 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 59 NULL , NULL , NULL , NULL , "Y?" , NULL , NULL , NULL , 60 NULL , "Y~" , NULL , NULL , NULL , NULL , "Y." , NULL , 61 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 62 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 63 NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL , 64 NULL , NULL , NULL , NULL , NULL , NULL , NULL , 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 "A." , "A('", "A(`", "A(.", "A^'", "A^`", "A^?", "A^.", 74 "E~" , "E." , "E^'", "E^`", "E^?", "E^~", "E^.", "O^'", 75 "O^`", "O^?", "O^~", "O^.", "O+.", "O+'", "O+`", "O+?", 76 "I." , "O?" , "O." , "I?" , "U?" , "U~" , "U." , "Y`" , 77 "O~" , "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." , "U+.", "U+'", "U+`", "U+?", "o+" , "o+'", "U+" , 81 "A`" , "A'" , "A^" , "A~" , "A?" , "A(" , "a(?", "a(~", 82 "E`" , "E'" , "E^" , "E?" , "I`" , "I'" , "I~" , "y`" , 83 "DD" , "u+'", "O`" , "O'" , "O^" , "a." , "y?" , "u+`", 84 "u+?", "U`" , "U'" , "y~" , "y." , "Y'" , "o+~", "u+" , 85 "a`" , "a'" , "a^" , "a~" , "a?" , "a(" , "u+~", "a^~", 86 "e`" , "e'" , "e^" , "e?" , "i`" , "i'" , "i~" , "i?" , 87 "dd" , "u+.", "o`" , "o'" , "o^" , "o~" , "o?" , "o." , 88 "u." , "u`" , "u'" , "u~" , "u?" , "y'" , "o+.", "U+~", 89 }; 90 91 typedef struct { 92 const char *name; 93 wchar_t value; 94 } mnemonic_def_t; 95 96 static const mnemonic_def_t mnemonic_ext[] = { 97 /* add extra mnemonic here (should be sorted by wchar_t order). */ 98 }; 99 static const size_t mnemonic_ext_size = 100 sizeof(mnemonic_ext) / sizeof(mnemonic_def_t); 101 102 static const char * 103 mnemonic_ext_find(wchar_t wc, const mnemonic_def_t *head, size_t n) 104 { 105 const mnemonic_def_t *mid; 106 107 for (; n > 0; n >>= 1) { 108 mid = head + (n >> 1); 109 if (mid->value == wc) 110 return (mid->name); 111 else if (mid->value < wc) { 112 head = mid + 1; 113 --n; 114 } 115 } 116 return (NULL); 117 } 118 119 struct mnemonic_t; 120 typedef TAILQ_HEAD(mnemonic_list_t, mnemonic_t) mnemonic_list_t; 121 typedef struct mnemonic_t { 122 TAILQ_ENTRY(mnemonic_t) entry; 123 struct mnemonic_t *parent; 124 mnemonic_list_t child; 125 wchar_t value; 126 int ascii; 127 } mnemonic_t; 128 129 static mnemonic_t * 130 mnemonic_list_find(mnemonic_list_t *ml, int ch) 131 { 132 mnemonic_t *m; 133 134 TAILQ_FOREACH(m, ml, entry) { 135 if (m->ascii == ch) 136 return (m); 137 } 138 139 return (NULL); 140 } 141 142 static mnemonic_t * 143 mnemonic_create(mnemonic_t *parent, int ascii, wchar_t value) 144 { 145 mnemonic_t *m; 146 147 m = malloc(sizeof(*m)); 148 if (m != NULL) { 149 m->parent = parent; 150 m->ascii = ascii; 151 m->value = value; 152 TAILQ_INIT(&m->child); 153 } 154 155 return (m); 156 } 157 158 static int 159 mnemonic_append_child(mnemonic_t *m, const char *s, 160 wchar_t value, wchar_t invalid) 161 { 162 mnemonic_t *m0; 163 int ch; 164 165 ch = (unsigned char)*s++; 166 if (ch == '\0') 167 return (EINVAL); 168 m0 = mnemonic_list_find(&m->child, ch); 169 if (m0 == NULL) { 170 m0 = mnemonic_create(m, ch, (wchar_t)ch); 171 if (m0 == NULL) 172 return (ENOMEM); 173 TAILQ_INSERT_TAIL(&m->child, m0, entry); 174 } 175 m = m0; 176 for (m0 = NULL; (ch = (unsigned char)*s) != '\0'; ++s) { 177 m0 = mnemonic_list_find(&m->child, ch); 178 if (m0 == NULL) { 179 m0 = mnemonic_create(m, ch, invalid); 180 if (m0 == NULL) 181 return (ENOMEM); 182 TAILQ_INSERT_TAIL(&m->child, m0, entry); 183 } 184 m = m0; 185 } 186 if (m0 == NULL) 187 return (EINVAL); 188 m0->value = value; 189 190 return (0); 191 } 192 193 static void 194 mnemonic_destroy(mnemonic_t *m) 195 { 196 mnemonic_t *m0; 197 198 TAILQ_FOREACH(m0, &m->child, entry) 199 mnemonic_destroy(m0); 200 free(m); 201 } 202 203 typedef struct { 204 mnemonic_t *mroot; 205 wchar_t invalid; 206 size_t mb_cur_max; 207 } _VIQREncodingInfo; 208 209 typedef struct { 210 int chlen; 211 char ch[MB_LEN_MAX]; 212 } _VIQRState; 213 214 #define _CEI_TO_EI(_cei_) (&(_cei_)->ei) 215 #define _CEI_TO_STATE(_cei_, _func_) (_cei_)->states.s_##_func_ 216 217 #define _FUNCNAME(m) _citrus_VIQR_##m 218 #define _ENCODING_INFO _VIQREncodingInfo 219 #define _ENCODING_STATE _VIQRState 220 #define _ENCODING_MB_CUR_MAX(_ei_) (_ei_)->mb_cur_max 221 #define _ENCODING_IS_STATE_DEPENDENT 1 222 #define _STATE_NEEDS_EXPLICIT_INIT(_ps_) 0 223 224 static __inline void 225 /*ARGSUSED*/ 226 _citrus_VIQR_init_state(_VIQREncodingInfo * __restrict ei __unused, 227 _VIQRState * __restrict psenc) 228 { 229 230 psenc->chlen = 0; 231 } 232 233 #if 0 234 static __inline void 235 /*ARGSUSED*/ 236 _citrus_VIQR_pack_state(_VIQREncodingInfo * __restrict ei __unused, 237 void *__restrict pspriv, const _VIQRState * __restrict psenc) 238 { 239 240 memcpy(pspriv, (const void *)psenc, sizeof(*psenc)); 241 } 242 243 static __inline void 244 /*ARGSUSED*/ 245 _citrus_VIQR_unpack_state(_VIQREncodingInfo * __restrict ei __unused, 246 _VIQRState * __restrict psenc, const void * __restrict pspriv) 247 { 248 249 memcpy((void *)psenc, pspriv, sizeof(*psenc)); 250 } 251 #endif 252 253 static int 254 _citrus_VIQR_mbrtowc_priv(_VIQREncodingInfo * __restrict ei, 255 wchar_t * __restrict pwc, char ** __restrict s, size_t n, 256 _VIQRState * __restrict psenc, size_t * __restrict nresult) 257 { 258 mnemonic_t *m, *m0; 259 char *s0; 260 wchar_t wc; 261 ssize_t i; 262 int ch, escape; 263 264 if (*s == NULL) { 265 _citrus_VIQR_init_state(ei, psenc); 266 *nresult = (size_t)_ENCODING_IS_STATE_DEPENDENT; 267 return (0); 268 } 269 s0 = *s; 270 271 i = 0; 272 m = ei->mroot; 273 for (escape = 0;;) { 274 if (psenc->chlen == i) { 275 if (n-- < 1) { 276 *s = s0; 277 *nresult = (size_t)-2; 278 return (0); 279 } 280 psenc->ch[psenc->chlen++] = *s0++; 281 } 282 ch = (unsigned char)psenc->ch[i++]; 283 if (ch == ESCAPE) { 284 if (m != ei->mroot) 285 break; 286 escape = 1; 287 continue; 288 } 289 if (escape != 0) 290 break; 291 m0 = mnemonic_list_find(&m->child, ch); 292 if (m0 == NULL) 293 break; 294 m = m0; 295 } 296 while (m != ei->mroot) { 297 --i; 298 if (m->value != ei->invalid) 299 break; 300 m = m->parent; 301 } 302 if (ch == ESCAPE && m != ei->mroot) 303 ++i; 304 psenc->chlen -= i; 305 memmove(&psenc->ch[0], &psenc->ch[i], psenc->chlen); 306 wc = (m == ei->mroot) ? (wchar_t)ch : m->value; 307 if (pwc != NULL) 308 *pwc = wc; 309 *nresult = (size_t)(wc == 0 ? 0 : s0 - *s); 310 *s = s0; 311 312 return (0); 313 } 314 315 static int 316 _citrus_VIQR_wcrtomb_priv(_VIQREncodingInfo * __restrict ei, 317 char * __restrict s, size_t n, wchar_t wc, 318 _VIQRState * __restrict psenc, size_t * __restrict nresult) 319 { 320 mnemonic_t *m; 321 const char *p; 322 int ch = 0; 323 324 switch (psenc->chlen) { 325 case 0: case 1: 326 break; 327 default: 328 return (EINVAL); 329 } 330 m = NULL; 331 if ((uint32_t)wc <= 0xFF) { 332 p = mnemonic_rfc1456[wc & 0xFF]; 333 if (p != NULL) 334 goto mnemonic_found; 335 if (n-- < 1) 336 goto e2big; 337 ch = (unsigned int)wc; 338 m = ei->mroot; 339 if (psenc->chlen > 0) { 340 m = mnemonic_list_find(&m->child, psenc->ch[0]); 341 if (m == NULL) 342 return (EINVAL); 343 psenc->ch[0] = ESCAPE; 344 } 345 if (mnemonic_list_find(&m->child, ch) == NULL) { 346 psenc->chlen = 0; 347 m = NULL; 348 } 349 psenc->ch[psenc->chlen++] = ch; 350 } else { 351 p = mnemonic_ext_find(wc, &mnemonic_ext[0], mnemonic_ext_size); 352 if (p == NULL) { 353 *nresult = (size_t)-1; 354 return (EILSEQ); 355 } else { 356 mnemonic_found: 357 psenc->chlen = 0; 358 while (*p != '\0') { 359 if (n-- < 1) 360 goto e2big; 361 psenc->ch[psenc->chlen++] = *p++; 362 } 363 } 364 } 365 memcpy(s, psenc->ch, psenc->chlen); 366 *nresult = psenc->chlen; 367 if (m == ei->mroot) { 368 psenc->ch[0] = ch; 369 psenc->chlen = 1; 370 } else 371 psenc->chlen = 0; 372 373 return (0); 374 375 e2big: 376 *nresult = (size_t)-1; 377 return (E2BIG); 378 } 379 380 static int 381 /* ARGSUSED */ 382 _citrus_VIQR_put_state_reset(_VIQREncodingInfo * __restrict ei __unused, 383 char * __restrict s __unused, size_t n __unused, 384 _VIQRState * __restrict psenc, size_t * __restrict nresult) 385 { 386 387 switch (psenc->chlen) { 388 case 0: case 1: 389 break; 390 default: 391 return (EINVAL); 392 } 393 *nresult = 0; 394 psenc->chlen = 0; 395 396 return (0); 397 } 398 399 static __inline int 400 /*ARGSUSED*/ 401 _citrus_VIQR_stdenc_wctocs(_VIQREncodingInfo * __restrict ei __unused, 402 _csid_t * __restrict csid, _index_t * __restrict idx, wchar_t wc) 403 { 404 405 *csid = 0; 406 *idx = (_index_t)wc; 407 408 return (0); 409 } 410 411 static __inline int 412 /*ARGSUSED*/ 413 _citrus_VIQR_stdenc_cstowc(_VIQREncodingInfo * __restrict ei __unused, 414 wchar_t * __restrict pwc, _csid_t csid, _index_t idx) 415 { 416 417 if (csid != 0) 418 return (EILSEQ); 419 *pwc = (wchar_t)idx; 420 421 return (0); 422 } 423 424 static void 425 _citrus_VIQR_encoding_module_uninit(_VIQREncodingInfo *ei) 426 { 427 428 mnemonic_destroy(ei->mroot); 429 } 430 431 static int 432 /*ARGSUSED*/ 433 _citrus_VIQR_encoding_module_init(_VIQREncodingInfo * __restrict ei, 434 const void * __restrict var __unused, size_t lenvar __unused) 435 { 436 const mnemonic_def_t *p; 437 const char *s; 438 size_t i, n; 439 int errnum; 440 441 ei->mb_cur_max = 1; 442 ei->invalid = (wchar_t)-1; 443 ei->mroot = mnemonic_create(NULL, '\0', ei->invalid); 444 if (ei->mroot == NULL) 445 return (ENOMEM); 446 for (i = 0; i < sizeof(mnemonic_rfc1456) / sizeof(const char *); ++i) { 447 s = mnemonic_rfc1456[i]; 448 if (s == NULL) 449 continue; 450 n = strlen(s); 451 if (ei->mb_cur_max < n) 452 ei->mb_cur_max = n; 453 errnum = mnemonic_append_child(ei->mroot, 454 s, (wchar_t)i, ei->invalid); 455 if (errnum != 0) { 456 _citrus_VIQR_encoding_module_uninit(ei); 457 return (errnum); 458 } 459 } 460 for (i = 0; i < mnemonic_ext_size; ++i) { 461 p = &mnemonic_ext[i]; 462 n = strlen(p->name); 463 if (ei->mb_cur_max < n) 464 ei->mb_cur_max = n; 465 errnum = mnemonic_append_child(ei->mroot, 466 p->name, p->value, ei->invalid); 467 if (errnum != 0) { 468 _citrus_VIQR_encoding_module_uninit(ei); 469 return (errnum); 470 } 471 } 472 473 return (0); 474 } 475 476 static __inline int 477 /*ARGSUSED*/ 478 _citrus_VIQR_stdenc_get_state_desc_generic(_VIQREncodingInfo * __restrict ei __unused, 479 _VIQRState * __restrict psenc, int * __restrict rstate) 480 { 481 482 *rstate = (psenc->chlen == 0) ? 483 _STDENC_SDGEN_INITIAL : 484 _STDENC_SDGEN_INCOMPLETE_CHAR; 485 486 return (0); 487 } 488 489 /* ---------------------------------------------------------------------- 490 * public interface for stdenc 491 */ 492 493 _CITRUS_STDENC_DECLS(VIQR); 494 _CITRUS_STDENC_DEF_OPS(VIQR); 495 496 #include "citrus_stdenc_template.h" 497