lib/libc/citrus/modules/citrus_viqr.c

   1 /* $NetBSD: citrus_viqr.c,v 1.5 2011/11/19 18:20:13 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 1.5 2011/11/19 18:20:13 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 <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_wcrtomb;
 238                 _VIQRState      s_wcsrtombs;
 239                 _VIQRState      s_wctomb;
 240         } states;
 241 } _VIQRCTypeInfo;
 242
 243 #define _CEI_TO_EI(_cei_)               (&(_cei_)->ei)
 244 #define _CEI_TO_STATE(_cei_, _func_)    (_cei_)->states.s_##_func_
 245
 246 #define _FUNCNAME(m)                    _citrus_VIQR_##m
 247 #define _ENCODING_INFO                  _VIQREncodingInfo
 248 #define _CTYPE_INFO                     _VIQRCTypeInfo
 249 #define _ENCODING_STATE                 _VIQRState
 250 #define _ENCODING_MB_CUR_MAX(_ei_)      (_ei_)->mb_cur_max
 251 #define _ENCODING_IS_STATE_DEPENDENT            1
 252 #define _STATE_NEEDS_EXPLICIT_INIT(_ps_)        0
 253
 254 static __inline void
 255 /*ARGSUSED*/
 256 _citrus_VIQR_init_state(_VIQREncodingInfo * __restrict ei,
 257         _VIQRState * __restrict psenc)
 258 {
 259         /* ei may be unused */
 260         _DIAGASSERT(psenc != NULL);
 261
 262         psenc->chlen = 0;
 263 }
 264
 265 static __inline void
 266 /*ARGSUSED*/
 267 _citrus_VIQR_pack_state(_VIQREncodingInfo * __restrict ei,
 268         void *__restrict pspriv, const _VIQRState * __restrict psenc)
 269 {
 270         /* ei may be unused */
 271         _DIAGASSERT(pspriv != NULL);
 272         _DIAGASSERT(psenc != NULL);
 273
 274         memcpy(pspriv, (const void *)psenc, sizeof(*psenc));
 275 }
 276
 277 static __inline void
 278 /*ARGSUSED*/
 279 _citrus_VIQR_unpack_state(_VIQREncodingInfo * __restrict ei,
 280         _VIQRState * __restrict psenc, const void * __restrict pspriv)
 281 {
 282         /* ei may be unused */
 283         _DIAGASSERT(psenc != NULL);
 284         _DIAGASSERT(pspriv != NULL);
 285
 286         memcpy((void *)psenc, pspriv, sizeof(*psenc));
 287 }
 288
 289 static int
 290 _citrus_VIQR_mbrtowc_priv(_VIQREncodingInfo * __restrict ei,
 291         wchar_t * __restrict pwc, const char ** __restrict s, size_t n,
 292         _VIQRState * __restrict psenc, size_t * __restrict nresult)
 293 {
 294         const char *s0;
 295         wchar_t wc;
 296         mnemonic_t *m, *m0;
 297         size_t i;
 298         int ch, escape;
 299
 300         _DIAGASSERT(ei != NULL);
 301         /* pwc may be null */
 302         _DIAGASSERT(s != NULL);
 303         _DIAGASSERT(psenc != NULL);
 304         _DIAGASSERT(nresult != NULL);
 305
 306         if (*s == NULL) {
 307                 _citrus_VIQR_init_state(ei, psenc);
 308                 *nresult = (size_t)_ENCODING_IS_STATE_DEPENDENT;
 309                 return 0;
 310         }
 311         s0 = *s;
 312
 313         i = 0;
 314         m = ei->mroot;
 315         for (escape = 0;;) {
 316                 if (psenc->chlen == i) {
 317                         if (n-- < 1) {
 318                                 *s = s0;
 319                                 *nresult = (size_t)-2;
 320                                 return 0;
 321                         }
 322                         psenc->ch[psenc->chlen++] = *s0++;
 323                 }
 324                 ch = (unsigned char)psenc->ch[i++];
 325                 if (ch == ESCAPE) {
 326                         if (m != ei->mroot)
 327                                 break;
 328                         escape = 1;
 329                         continue;
 330                 }
 331                 if (escape != 0)
 332                         break;
 333                 m0 = mnemonic_list_find(&m->child, ch);
 334                 if (m0 == NULL)
 335                         break;
 336                 m = m0;
 337         }
 338         while (m != ei->mroot) {
 339                 --i;
 340                 if (m->value != ei->invalid)
 341                         break;
 342                 m = m->parent;
 343         }
 344         if (ch == ESCAPE && m != ei->mroot)
 345                 ++i;
 346         psenc->chlen -= i;
 347         memmove(&psenc->ch[0], &psenc->ch[i], psenc->chlen);
 348         wc = (m == ei->mroot) ? (wchar_t)ch : m->value;
 349         if (pwc != NULL)
 350                 *pwc = wc;
 351         *nresult = (size_t)(wc == 0 ? 0 : s0 - *s);
 352         *s = s0;
 353
 354         return 0;
 355 }
 356
 357 static int
 358 _citrus_VIQR_wcrtomb_priv(_VIQREncodingInfo * __restrict ei,
 359         char * __restrict s, size_t n, wchar_t wc,
 360         _VIQRState * __restrict psenc, size_t * __restrict nresult)
 361 {
 362         mnemonic_t *m;
 363         int ch;
 364         const char *p;
 365
 366         _DIAGASSERT(ei != NULL);
 367         _DIAGASSERT(s != NULL);
 368         _DIAGASSERT(psenc != NULL);
 369         _DIAGASSERT(nresult != NULL);
 370
 371         switch (psenc->chlen) {
 372         case 0: case 1:
 373                 break;
 374         default:
 375                 return EINVAL;
 376         }
 377         m = NULL;
 378         if ((uint32_t)wc <= 0xFF) {
 379                 p = mnemonic_rfc1456[wc & 0xFF];
 380                 if (p != NULL)
 381                         goto mnemonic_found;
 382                 if (n-- < 1)
 383                         goto e2big;
 384                 ch = (unsigned int)wc;
 385                 m = ei->mroot;
 386                 if (psenc->chlen > 0) {
 387                         m = mnemonic_list_find(&m->child, psenc->ch[0]);
 388                         if (m == NULL)
 389                                 return EINVAL;
 390                         psenc->ch[0] = ESCAPE;
 391                 }
 392                 if (mnemonic_list_find(&m->child, ch) == NULL) {
 393                         psenc->chlen = 0;
 394                         m = NULL;
 395                 }
 396                 psenc->ch[psenc->chlen++] = ch;
 397         } else {
 398                 p = mnemonic_ext_find(wc, &mnemonic_ext[0], mnemonic_ext_size);
 399                 if (p == NULL) {
 400                         *nresult = (size_t)-1;
 401                         return EILSEQ;
 402                 } else {
 403 mnemonic_found:
 404                         psenc->chlen = 0;
 405                         while (*p != '\0') {
 406                                 if (n-- < 1)
 407                                         goto e2big;
 408                                 psenc->ch[psenc->chlen++] = *p++;
 409                         }
 410                 }
 411         }
 412         memcpy(s, psenc->ch, psenc->chlen);
 413         *nresult = psenc->chlen;
 414         if (m == ei->mroot) {
 415                 psenc->ch[0] = ch;
 416                 psenc->chlen = 1;
 417         } else {
 418                 psenc->chlen = 0;
 419         }
 420
 421         return 0;
 422
 423 e2big:
 424         *nresult = (size_t)-1;
 425         return E2BIG;
 426 }
 427
 428 static int
 429 /* ARGSUSED */
 430 _citrus_VIQR_put_state_reset(_VIQREncodingInfo * __restrict ei,
 431         char * __restrict s, size_t n, _VIQRState * __restrict psenc,
 432         size_t * __restrict nresult)
 433 {
 434         /* ei may be unused */
 435         _DIAGASSERT(s != NULL);
 436         _DIAGASSERT(psenc != NULL);
 437         _DIAGASSERT(nresult != NULL);
 438
 439         switch (psenc->chlen) {
 440         case 0: case 1:
 441                 break;
 442         default:
 443                 return EINVAL;
 444         }
 445         *nresult = 0;
 446         psenc->chlen = 0;
 447
 448         return 0;
 449 }
 450
 451 static __inline int
 452 /*ARGSUSED*/
 453 _citrus_VIQR_stdenc_wctocs(_VIQREncodingInfo * __restrict ei,
 454         _csid_t * __restrict csid, _index_t * __restrict idx, wchar_t wc)
 455 {
 456         /* ei may be unused */
 457         _DIAGASSERT(csid != NULL);
 458         _DIAGASSERT(idx != NULL);
 459
 460         *csid = 0;
 461         *idx = (_index_t)wc;
 462
 463         return 0;
 464 }
 465
 466 static __inline int
 467 /*ARGSUSED*/
 468 _citrus_VIQR_stdenc_cstowc(_VIQREncodingInfo * __restrict ei,
 469         wchar_t * __restrict pwc, _csid_t csid, _index_t idx)
 470 {
 471         /* ei may be unused */
 472         _DIAGASSERT(pwc != NULL);
 473
 474         if (csid != 0)
 475                 return EILSEQ;
 476         *pwc = (wchar_t)idx;
 477
 478         return 0;
 479 }
 480
 481 static void
 482 _citrus_VIQR_encoding_module_uninit(_VIQREncodingInfo *ei)
 483 {
 484         _DIAGASSERT(ei != NULL);
 485
 486         mnemonic_destroy(ei->mroot);
 487 }
 488
 489 static int
 490 /*ARGSUSED*/
 491 _citrus_VIQR_encoding_module_init(_VIQREncodingInfo * __restrict ei,
 492         const void * __restrict var, size_t lenvar)
 493 {
 494         int errnum;
 495         const char *s;
 496         size_t i, n;
 497         const mnemonic_def_t *p;
 498
 499         _DIAGASSERT(ei != NULL);
 500         /* var may be unused */
 501
 502         ei->mb_cur_max = 1;
 503         ei->invalid = (wchar_t)-1;
 504         ei->mroot = mnemonic_create(NULL, '\0', ei->invalid);
 505         if (ei->mroot == NULL)
 506                 return ENOMEM;
 507         for (i = 0; i < sizeof(mnemonic_rfc1456) / sizeof(const char *); ++i) {
 508                 s = mnemonic_rfc1456[i];
 509                 if (s == NULL)
 510                         continue;
 511                 n = strlen(s);
 512                 _DIAGASSERT(n <= MB_LEN_MAX);
 513                 if (ei->mb_cur_max < n)
 514                         ei->mb_cur_max = n;
 515                 errnum = mnemonic_append_child(ei->mroot,
 516                     s, (wchar_t)i, ei->invalid);
 517                 if (errnum != 0) {
 518                         _citrus_VIQR_encoding_module_uninit(ei);
 519                         return errnum;
 520                 }
 521         }
 522         for (i = 0; i < mnemonic_ext_size; ++i) {
 523                 p = &mnemonic_ext[i];
 524                 _DIAGASSERT(p != NULL && p->name != NULL);
 525                 n = strlen(p->name);
 526                 _DIAGASSERT(n <= MB_LEN_MAX);
 527                 if (ei->mb_cur_max < n)
 528                         ei->mb_cur_max = n;
 529                 errnum = mnemonic_append_child(ei->mroot,
 530                     p->name, p->value, ei->invalid);
 531                 if (errnum != 0) {
 532                         _citrus_VIQR_encoding_module_uninit(ei);
 533                         return errnum;
 534                 }
 535         }
 536
 537         return 0;
 538 }
 539
 540 static __inline int
 541 /*ARGSUSED*/
 542 _citrus_VIQR_stdenc_get_state_desc_generic(_VIQREncodingInfo * __restrict ei,
 543         _VIQRState * __restrict psenc, int * __restrict rstate)
 544 {
 545         /* ei may be unused */
 546         _DIAGASSERT(psenc != NULL);
 547         _DIAGASSERT(rstate != NULL);
 548
 549         *rstate = (psenc->chlen == 0)
 550             ? _STDENC_SDGEN_INITIAL
 551             : _STDENC_SDGEN_INCOMPLETE_CHAR;
 552
 553         return 0;
 554 }
 555
 556 /* ----------------------------------------------------------------------
 557  * public interface for ctype
 558  */
 559
 560 _CITRUS_CTYPE_DECLS(VIQR);
 561 _CITRUS_CTYPE_DEF_OPS(VIQR);
 562
 563 #include "citrus_ctype_template.h"
 564
 565 /* ----------------------------------------------------------------------
 566  * public interface for stdenc
 567  */
 568
 569 _CITRUS_STDENC_DECLS(VIQR);
 570 _CITRUS_STDENC_DEF_OPS(VIQR);
 571
 572 #include "citrus_stdenc_template.h"