lib/libc/citrus/modules/citrus_viqr.c

   1 /* $NetBSD: citrus_viqr.c,v 1.6 2013/05/28 16:57:56 joerg 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 1.6 2013/05/28 16:57:56 joerg 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 <wchar.h>
  44 #include <limits.h>
  45
  46 #include "citrus_namespace.h"
  47 #include "citrus_types.h"
  48 #include "citrus_bcs.h"
  49 #include "citrus_module.h"
  50 #include "citrus_ctype.h"
  51 #include "citrus_stdenc.h"
  52 #include "citrus_viqr.h"
  53
  54 #define ESCAPE  '\\'
  55
  56 /*
  57  * this table generated from RFC 1456.
  58  */
  59 static const char *mnemonic_rfc1456[0x100] = {
  60   NULL , NULL , "A(?", NULL , NULL , "A(~", "A^~", NULL ,
  61   NULL , NULL , NULL , NULL , NULL , NULL , NULL , NULL ,
  62   NULL , NULL , NULL , NULL , "Y?" , NULL , NULL , NULL ,
  63   NULL , "Y~" , NULL , NULL , NULL , NULL , "Y." , 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   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   "A." , "A('", "A(`", "A(.", "A^'", "A^`", "A^?", "A^.",
  77   "E~" , "E." , "E^'", "E^`", "E^?", "E^~", "E^.", "O^'",
  78   "O^`", "O^?", "O^~", "O^.", "O+.", "O+'", "O+`", "O+?",
  79   "I." , "O?" , "O." , "I?" , "U?" , "U~" , "U." , "Y`" ,
  80   "O~" , "a('", "a(`", "a(.", "a^'", "a^`", "a^?", "a^.",
  81   "e~" , "e." , "e^'", "e^`", "e^?", "e^~", "e^.", "o^'",
  82   "o^`", "o^?", "o^~", "O+~", "O+" , "o^.", "o+`", "o+?",
  83   "i." , "U+.", "U+'", "U+`", "U+?", "o+" , "o+'", "U+" ,
  84   "A`" , "A'" , "A^" , "A~" , "A?" , "A(" , "a(?", "a(~",
  85   "E`" , "E'" , "E^" , "E?" , "I`" , "I'" , "I~" , "y`" ,
  86   "DD" , "u+'", "O`" , "O'" , "O^" , "a." , "y?" , "u+`",
  87   "u+?", "U`" , "U'" , "y~" , "y." , "Y'" , "o+~", "u+" ,
  88   "a`" , "a'" , "a^" , "a~" , "a?" , "a(" , "u+~", "a^~",
  89   "e`" , "e'" , "e^" , "e?" , "i`" , "i'" , "i~" , "i?" ,
  90   "dd" , "u+.", "o`" , "o'" , "o^" , "o~" , "o?" , "o." ,
  91   "u." , "u`" , "u'" , "u~" , "u?" , "y'" , "o+.", "U+~",
  92 };
  93
  94 typedef struct {
  95         const char *name;
  96         wchar_t value;
  97 } mnemonic_def_t;
  98
  99 static const mnemonic_def_t mnemonic_ext[] = {
 100 /* add extra mnemonic here (should be sorted by wchar_t order). */
 101 };
 102 static const size_t mnemonic_ext_size =
 103         sizeof(mnemonic_ext) / sizeof(mnemonic_def_t);
 104
 105 static const char *
 106 mnemonic_ext_find(wchar_t wc, const mnemonic_def_t *head, size_t n)
 107 {
 108         const mnemonic_def_t *mid;
 109
 110         _DIAGASSERT(head != NULL);
 111
 112         for (; n > 0; n >>= 1) {
 113                 mid = head + (n >> 1);
 114                 if (mid->value == wc) {
 115                         return mid->name;
 116                 } else if (mid->value < wc) {
 117                         head = mid + 1;
 118                         --n;
 119                 }
 120         }
 121         return NULL;
 122 }
 123
 124 struct mnemonic_t;
 125 typedef TAILQ_HEAD(mnemonic_list_t, mnemonic_t) mnemonic_list_t;
 126 typedef struct mnemonic_t {
 127         TAILQ_ENTRY(mnemonic_t) entry;
 128         int ascii;
 129         struct mnemonic_t *parent;
 130         mnemonic_list_t child;
 131         wchar_t value;
 132 } mnemonic_t;
 133
 134 static mnemonic_t *
 135 mnemonic_list_find(mnemonic_list_t *ml, int ch)
 136 {
 137         mnemonic_t *m;
 138
 139         _DIAGASSERT(ml != NULL);
 140
 141         TAILQ_FOREACH(m, ml, entry) {
 142                 if (m->ascii == ch)
 143                         return m;
 144         }
 145
 146         return NULL;
 147 }
 148
 149 static mnemonic_t *
 150 mnemonic_create(mnemonic_t *parent, int ascii, wchar_t value)
 151 {
 152         mnemonic_t *m;
 153
 154         _DIAGASSERT(parent != NULL);
 155
 156         m = malloc(sizeof(*m));
 157         if (m != NULL) {
 158                 m->parent = parent;
 159                 m->ascii = ascii;
 160                 m->value = value;
 161                 TAILQ_INIT(&m->child);
 162         }
 163
 164         return m;
 165 }
 166
 167 static int
 168 mnemonic_append_child(mnemonic_t *m, const char *s,
 169         wchar_t value, wchar_t invalid)
 170 {
 171         int ch;
 172         mnemonic_t *m0;
 173
 174         _DIAGASSERT(m != NULL);
 175         _DIAGASSERT(s != NULL);
 176
 177         ch = (unsigned char)*s++;
 178         if (ch == '\0')
 179                 return EINVAL;
 180         m0 = mnemonic_list_find(&m->child, ch);
 181         if (m0 == NULL) {
 182                 m0 = mnemonic_create(m, ch, (wchar_t)ch);
 183                 if (m0 == NULL)
 184                         return ENOMEM;
 185                 TAILQ_INSERT_TAIL(&m->child, m0, entry);
 186         }
 187         m = m0;
 188         for (m0 = NULL; (ch = (unsigned char)*s) != '\0'; ++s) {
 189                 m0 = mnemonic_list_find(&m->child, ch);
 190                 if (m0 == NULL) {
 191                         m0 = mnemonic_create(m, ch, invalid);
 192                         if (m0 == NULL)
 193                                 return ENOMEM;
 194                         TAILQ_INSERT_TAIL(&m->child, m0, entry);
 195                 }
 196                 m = m0;
 197         }
 198         if (m0 == NULL)
 199                 return EINVAL;
 200         m0->value = value;
 201
 202         return 0;
 203 }
 204
 205 static void
 206 mnemonic_destroy(mnemonic_t *m)
 207 {
 208         mnemonic_t *m0;
 209
 210         _DIAGASSERT(m != NULL);
 211
 212         TAILQ_FOREACH(m0, &m->child, entry)
 213                 mnemonic_destroy(m0);
 214         free(m);
 215 }
 216
 217 typedef struct {
 218         size_t mb_cur_max;
 219         wchar_t invalid;
 220         mnemonic_t *mroot;
 221 } _VIQREncodingInfo;
 222
 223 typedef struct {
 224         int chlen;
 225         char ch[MB_LEN_MAX];
 226 } _VIQRState;
 227
 228 typedef struct {
 229         _VIQREncodingInfo       ei;
 230         struct {
 231                 /* for future multi-locale facility */
 232                 _VIQRState      s_mblen;
 233                 _VIQRState      s_mbrlen;
 234                 _VIQRState      s_mbrtowc;
 235                 _VIQRState      s_mbtowc;
 236                 _VIQRState      s_mbsrtowcs;
 237                 _VIQRState      s_mbsnrtowcs;
 238                 _VIQRState      s_wcrtomb;
 239                 _VIQRState      s_wcsrtombs;
 240                 _VIQRState      s_wcsnrtombs;
 241                 _VIQRState      s_wctomb;
 242         } states;
 243 } _VIQRCTypeInfo;
 244
 245 #define _CEI_TO_EI(_cei_)               (&(_cei_)->ei)
 246 #define _CEI_TO_STATE(_cei_, _func_)    (_cei_)->states.s_##_func_
 247
 248 #define _FUNCNAME(m)                    _citrus_VIQR_##m
 249 #define _ENCODING_INFO                  _VIQREncodingInfo
 250 #define _CTYPE_INFO                     _VIQRCTypeInfo
 251 #define _ENCODING_STATE                 _VIQRState
 252 #define _ENCODING_MB_CUR_MAX(_ei_)      (_ei_)->mb_cur_max
 253 #define _ENCODING_IS_STATE_DEPENDENT            1
 254 #define _STATE_NEEDS_EXPLICIT_INIT(_ps_)        0
 255
 256 static __inline void
 257 /*ARGSUSED*/
 258 _citrus_VIQR_init_state(_VIQREncodingInfo * __restrict ei,
 259         _VIQRState * __restrict psenc)
 260 {
 261         /* ei may be unused */
 262         _DIAGASSERT(psenc != NULL);
 263
 264         psenc->chlen = 0;
 265 }
 266
 267 static __inline void
 268 /*ARGSUSED*/
 269 _citrus_VIQR_pack_state(_VIQREncodingInfo * __restrict ei,
 270         void *__restrict pspriv, const _VIQRState * __restrict psenc)
 271 {
 272         /* ei may be unused */
 273         _DIAGASSERT(pspriv != NULL);
 274         _DIAGASSERT(psenc != NULL);
 275
 276         memcpy(pspriv, (const void *)psenc, sizeof(*psenc));
 277 }
 278
 279 static __inline void
 280 /*ARGSUSED*/
 281 _citrus_VIQR_unpack_state(_VIQREncodingInfo * __restrict ei,
 282         _VIQRState * __restrict psenc, const void * __restrict pspriv)
 283 {
 284         /* ei may be unused */
 285         _DIAGASSERT(psenc != NULL);
 286         _DIAGASSERT(pspriv != NULL);
 287
 288         memcpy((void *)psenc, pspriv, sizeof(*psenc));
 289 }
 290
 291 static int
 292 _citrus_VIQR_mbrtowc_priv(_VIQREncodingInfo * __restrict ei,
 293         wchar_t * __restrict pwc, const char ** __restrict s, size_t n,
 294         _VIQRState * __restrict psenc, size_t * __restrict nresult)
 295 {
 296         const char *s0;
 297         wchar_t wc;
 298         mnemonic_t *m, *m0;
 299         size_t i;
 300         int ch, escape;
 301
 302         _DIAGASSERT(ei != NULL);
 303         /* pwc may be null */
 304         _DIAGASSERT(s != NULL);
 305         _DIAGASSERT(psenc != NULL);
 306         _DIAGASSERT(nresult != NULL);
 307
 308         if (*s == NULL) {
 309                 _citrus_VIQR_init_state(ei, psenc);
 310                 *nresult = (size_t)_ENCODING_IS_STATE_DEPENDENT;
 311                 return 0;
 312         }
 313         s0 = *s;
 314
 315         i = 0;
 316         m = ei->mroot;
 317         for (escape = 0;;) {
 318                 if (psenc->chlen == i) {
 319                         if (n-- < 1) {
 320                                 *s = s0;
 321                                 *nresult = (size_t)-2;
 322                                 return 0;
 323                         }
 324                         psenc->ch[psenc->chlen++] = *s0++;
 325                 }
 326                 ch = (unsigned char)psenc->ch[i++];
 327                 if (ch == ESCAPE) {
 328                         if (m != ei->mroot)
 329                                 break;
 330                         escape = 1;
 331                         continue;
 332                 }
 333                 if (escape != 0)
 334                         break;
 335                 m0 = mnemonic_list_find(&m->child, ch);
 336                 if (m0 == NULL)
 337                         break;
 338                 m = m0;
 339         }
 340         while (m != ei->mroot) {
 341                 --i;
 342                 if (m->value != ei->invalid)
 343                         break;
 344                 m = m->parent;
 345         }
 346         if (ch == ESCAPE && m != ei->mroot)
 347                 ++i;
 348         psenc->chlen -= i;
 349         memmove(&psenc->ch[0], &psenc->ch[i], psenc->chlen);
 350         wc = (m == ei->mroot) ? (wchar_t)ch : m->value;
 351         if (pwc != NULL)
 352                 *pwc = wc;
 353         *nresult = (size_t)(wc == 0 ? 0 : s0 - *s);
 354         *s = s0;
 355
 356         return 0;
 357 }
 358
 359 static int
 360 _citrus_VIQR_wcrtomb_priv(_VIQREncodingInfo * __restrict ei,
 361         char * __restrict s, size_t n, wchar_t wc,
 362         _VIQRState * __restrict psenc, size_t * __restrict nresult)
 363 {
 364         mnemonic_t *m;
 365         int ch;
 366         const char *p;
 367
 368         _DIAGASSERT(ei != NULL);
 369         _DIAGASSERT(s != NULL);
 370         _DIAGASSERT(psenc != NULL);
 371         _DIAGASSERT(nresult != NULL);
 372
 373         switch (psenc->chlen) {
 374         case 0: case 1:
 375                 break;
 376         default:
 377                 return EINVAL;
 378         }
 379         m = NULL;
 380         if ((uint32_t)wc <= 0xFF) {
 381                 p = mnemonic_rfc1456[wc & 0xFF];
 382                 if (p != NULL)
 383                         goto mnemonic_found;
 384                 if (n-- < 1)
 385                         goto e2big;
 386                 ch = (unsigned int)wc;
 387                 m = ei->mroot;
 388                 if (psenc->chlen > 0) {
 389                         m = mnemonic_list_find(&m->child, psenc->ch[0]);
 390                         if (m == NULL)
 391                                 return EINVAL;
 392                         psenc->ch[0] = ESCAPE;
 393                 }
 394                 if (mnemonic_list_find(&m->child, ch) == NULL) {
 395                         psenc->chlen = 0;
 396                         m = NULL;
 397                 }
 398                 psenc->ch[psenc->chlen++] = ch;
 399         } else {
 400                 p = mnemonic_ext_find(wc, &mnemonic_ext[0], mnemonic_ext_size);
 401                 if (p == NULL) {
 402                         *nresult = (size_t)-1;
 403                         return EILSEQ;
 404                 } else {
 405 mnemonic_found:
 406                         psenc->chlen = 0;
 407                         while (*p != '\0') {
 408                                 if (n-- < 1)
 409                                         goto e2big;
 410                                 psenc->ch[psenc->chlen++] = *p++;
 411                         }
 412                 }
 413         }
 414         memcpy(s, psenc->ch, psenc->chlen);
 415         *nresult = psenc->chlen;
 416         if (m == ei->mroot) {
 417                 psenc->ch[0] = ch;
 418                 psenc->chlen = 1;
 419         } else {
 420                 psenc->chlen = 0;
 421         }
 422
 423         return 0;
 424
 425 e2big:
 426         *nresult = (size_t)-1;
 427         return E2BIG;
 428 }
 429
 430 static int
 431 /* ARGSUSED */
 432 _citrus_VIQR_put_state_reset(_VIQREncodingInfo * __restrict ei,
 433         char * __restrict s, size_t n, _VIQRState * __restrict psenc,
 434         size_t * __restrict nresult)
 435 {
 436         /* ei may be unused */
 437         _DIAGASSERT(s != NULL);
 438         _DIAGASSERT(psenc != NULL);
 439         _DIAGASSERT(nresult != NULL);
 440
 441         switch (psenc->chlen) {
 442         case 0: case 1:
 443                 break;
 444         default:
 445                 return EINVAL;
 446         }
 447         *nresult = 0;
 448         psenc->chlen = 0;
 449
 450         return 0;
 451 }
 452
 453 static __inline int
 454 /*ARGSUSED*/
 455 _citrus_VIQR_stdenc_wctocs(_VIQREncodingInfo * __restrict ei,
 456         _csid_t * __restrict csid, _index_t * __restrict idx, wchar_t wc)
 457 {
 458         /* ei may be unused */
 459         _DIAGASSERT(csid != NULL);
 460         _DIAGASSERT(idx != NULL);
 461
 462         *csid = 0;
 463         *idx = (_index_t)wc;
 464
 465         return 0;
 466 }
 467
 468 static __inline int
 469 /*ARGSUSED*/
 470 _citrus_VIQR_stdenc_cstowc(_VIQREncodingInfo * __restrict ei,
 471         wchar_t * __restrict pwc, _csid_t csid, _index_t idx)
 472 {
 473         /* ei may be unused */
 474         _DIAGASSERT(pwc != NULL);
 475
 476         if (csid != 0)
 477                 return EILSEQ;
 478         *pwc = (wchar_t)idx;
 479
 480         return 0;
 481 }
 482
 483 static void
 484 _citrus_VIQR_encoding_module_uninit(_VIQREncodingInfo *ei)
 485 {
 486         _DIAGASSERT(ei != NULL);
 487
 488         mnemonic_destroy(ei->mroot);
 489 }
 490
 491 static int
 492 /*ARGSUSED*/
 493 _citrus_VIQR_encoding_module_init(_VIQREncodingInfo * __restrict ei,
 494         const void * __restrict var, size_t lenvar)
 495 {
 496         int errnum;
 497         const char *s;
 498         size_t i, n;
 499         const mnemonic_def_t *p;
 500
 501         _DIAGASSERT(ei != NULL);
 502         /* var may be unused */
 503
 504         ei->mb_cur_max = 1;
 505         ei->invalid = (wchar_t)-1;
 506         ei->mroot = mnemonic_create(NULL, '\0', ei->invalid);
 507         if (ei->mroot == NULL)
 508                 return ENOMEM;
 509         for (i = 0; i < sizeof(mnemonic_rfc1456) / sizeof(const char *); ++i) {
 510                 s = mnemonic_rfc1456[i];
 511                 if (s == NULL)
 512                         continue;
 513                 n = strlen(s);
 514                 _DIAGASSERT(n <= MB_LEN_MAX);
 515                 if (ei->mb_cur_max < n)
 516                         ei->mb_cur_max = n;
 517                 errnum = mnemonic_append_child(ei->mroot,
 518                     s, (wchar_t)i, ei->invalid);
 519                 if (errnum != 0) {
 520                         _citrus_VIQR_encoding_module_uninit(ei);
 521                         return errnum;
 522                 }
 523         }
 524         for (i = 0; i < mnemonic_ext_size; ++i) {
 525                 p = &mnemonic_ext[i];
 526                 _DIAGASSERT(p != NULL && p->name != NULL);
 527                 n = strlen(p->name);
 528                 _DIAGASSERT(n <= MB_LEN_MAX);
 529                 if (ei->mb_cur_max < n)
 530                         ei->mb_cur_max = n;
 531                 errnum = mnemonic_append_child(ei->mroot,
 532                     p->name, p->value, ei->invalid);
 533                 if (errnum != 0) {
 534                         _citrus_VIQR_encoding_module_uninit(ei);
 535                         return errnum;
 536                 }
 537         }
 538
 539         return 0;
 540 }
 541
 542 static __inline int
 543 /*ARGSUSED*/
 544 _citrus_VIQR_stdenc_get_state_desc_generic(_VIQREncodingInfo * __restrict ei,
 545         _VIQRState * __restrict psenc, int * __restrict rstate)
 546 {
 547         /* ei may be unused */
 548         _DIAGASSERT(psenc != NULL);
 549         _DIAGASSERT(rstate != NULL);
 550
 551         *rstate = (psenc->chlen == 0)
 552             ? _STDENC_SDGEN_INITIAL
 553             : _STDENC_SDGEN_INCOMPLETE_CHAR;
 554
 555         return 0;
 556 }
 557
 558 /* ----------------------------------------------------------------------
 559  * public interface for ctype
 560  */
 561
 562 _CITRUS_CTYPE_DECLS(VIQR);
 563 _CITRUS_CTYPE_DEF_OPS(VIQR);
 564
 565 #include "citrus_ctype_template.h"
 566
 567 /* ----------------------------------------------------------------------
 568  * public interface for stdenc
 569  */
 570
 571 _CITRUS_STDENC_DECLS(VIQR);
 572 _CITRUS_STDENC_DEF_OPS(VIQR);
 573
 574 #include "citrus_stdenc_template.h"