lib/libc/citrus/modules/citrus_gbk2k.c

   1 /* $NetBSD: citrus_gbk2k.c,v 1.6 2006/02/15 19:50:27 tnozaki Exp $ */
   2 /* $DragonFly: src/lib/libc/citrus/modules/citrus_gbk2k.c,v 1.2 2008/04/10 10:21:01 hasso Exp $ */
   3
   4 /*-
   5  * Copyright (c)2003 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 #include <sys/types.h>
  31 #include <assert.h>
  32 #include <errno.h>
  33 #include <limits.h>
  34 #include <locale.h>
  35 #include <stddef.h>
  36 #include <stdio.h>
  37 #include <stdlib.h>
  38 #include <string.h>
  39 #include <wchar.h>
  40
  41 #include "citrus_namespace.h"
  42 #include "citrus_types.h"
  43 #include "citrus_bcs.h"
  44 #include "citrus_module.h"
  45 #include "citrus_ctype.h"
  46 #include "citrus_stdenc.h"
  47 #include "citrus_gbk2k.h"
  48
  49
  50 /* ----------------------------------------------------------------------
  51  * private stuffs used by templates
  52  */
  53
  54 typedef struct _GBK2KState {
  55         char ch[4];
  56         int chlen;
  57 } _GBK2KState;
  58
  59 typedef struct {
  60         int mb_cur_max;
  61 } _GBK2KEncodingInfo;
  62
  63 typedef struct {
  64         _GBK2KEncodingInfo      ei;
  65         struct {
  66                 /* for future multi-locale facility */
  67                 _GBK2KState     s_mblen;
  68                 _GBK2KState     s_mbrlen;
  69                 _GBK2KState     s_mbrtowc;
  70                 _GBK2KState     s_mbtowc;
  71                 _GBK2KState     s_mbsrtowcs;
  72                 _GBK2KState     s_wcrtomb;
  73                 _GBK2KState     s_wcsrtombs;
  74                 _GBK2KState     s_wctomb;
  75         } states;
  76 } _GBK2KCTypeInfo;
  77
  78 #define _CEI_TO_EI(_cei_)               (&(_cei_)->ei)
  79 #define _CEI_TO_STATE(_cei_, _func_)    (_cei_)->states.s_##_func_
  80
  81 #define _FUNCNAME(m)                    _citrus_GBK2K_##m
  82 #define _ENCODING_INFO                  _GBK2KEncodingInfo
  83 #define _CTYPE_INFO                     _GBK2KCTypeInfo
  84 #define _ENCODING_STATE                 _GBK2KState
  85 #define _ENCODING_MB_CUR_MAX(_ei_)      (_ei_)->mb_cur_max
  86 #define _ENCODING_IS_STATE_DEPENDENT    0
  87 #define _STATE_NEEDS_EXPLICIT_INIT(_ps_)        0
  88
  89 static __inline void
  90 /*ARGSUSED*/
  91 _citrus_GBK2K_init_state(_GBK2KEncodingInfo * __restrict ei,
  92                          _GBK2KState * __restrict s)
  93 {
  94         memset(s, 0, sizeof(*s));
  95 }
  96
  97 static __inline void
  98 /*ARGSUSED*/
  99 _citrus_GBK2K_pack_state(_GBK2KEncodingInfo * __restrict ei,
 100                          void * __restrict pspriv,
 101                          const _GBK2KState * __restrict s)
 102 {
 103         memcpy(pspriv, (const void *)s, sizeof(*s));
 104 }
 105
 106 static __inline void
 107 /*ARGSUSED*/
 108 _citrus_GBK2K_unpack_state(_GBK2KEncodingInfo * __restrict ei,
 109                            _GBK2KState * __restrict s,
 110                            const void * __restrict pspriv)
 111 {
 112         memcpy((void *)s, pspriv, sizeof(*s));
 113 }
 114
 115 static  __inline int
 116 _mb_singlebyte(int c)
 117 {
 118         c &= 0xff;
 119         return (c <= 0x7f);
 120 }
 121
 122 static __inline int
 123 _mb_leadbyte(int c)
 124 {
 125         c &= 0xff;
 126         return (0x81 <= c && c <= 0xfe);
 127 }
 128
 129 static __inline int
 130 _mb_trailbyte(int c)
 131 {
 132         c &= 0xff;
 133         return ((0x40 <= c && c <= 0x7e) || (0x80 <= c && c <= 0xfe));
 134 }
 135
 136 static __inline int
 137 _mb_surrogate(int c)
 138 {
 139         c &= 0xff;
 140         return (0x30 <= c && c <= 0x39);
 141 }
 142
 143 static __inline int
 144 _mb_count(wchar_t v)
 145 {
 146         u_int32_t c;
 147
 148         c = (u_int32_t)v; /* XXX */
 149         if (!(c & 0xffffff00))
 150                 return (1);
 151         if (!(c & 0xffff0000))
 152                 return (2);
 153         return (4);
 154 }
 155
 156 #define _PSENC          (psenc->ch[psenc->chlen - 1])
 157 #define _PUSH_PSENC(c)  (psenc->ch[psenc->chlen++] = (c))
 158
 159 static int
 160 _citrus_GBK2K_mbrtowc_priv(_GBK2KEncodingInfo * __restrict ei,
 161                            wchar_t * __restrict pwc,
 162                            const char ** __restrict s, size_t n,
 163                            _GBK2KState * __restrict psenc,
 164                            size_t * __restrict nresult)
 165 {
 166         int chlenbak, len;
 167         const char *s0, *s1;
 168         wchar_t wc;
 169
 170         _DIAGASSERT(ei != NULL);
 171         /* pwc may be NULL */
 172         _DIAGASSERT(s != NULL);
 173         _DIAGASSERT(psenc != NULL);
 174
 175         s0 = *s;
 176
 177         if (s0 == NULL) {
 178                 /* _citrus_GBK2K_init_state(ei, psenc); */
 179                 psenc->chlen = 0;
 180                 *nresult = 0;
 181                 return (0);
 182         }
 183
 184         chlenbak = psenc->chlen;
 185
 186         switch (psenc->chlen) {
 187         case 3:
 188                 if (!_mb_leadbyte (_PSENC))
 189                         goto invalid;
 190         /* FALLTHROUGH */
 191         case 2:
 192                 if (!_mb_surrogate(_PSENC) || _mb_trailbyte(_PSENC))
 193                         goto invalid;
 194         /* FALLTHROUGH */
 195         case 1:
 196                 if (!_mb_leadbyte (_PSENC))
 197                         goto invalid;
 198         /* FALLTHOROUGH */
 199         case 0:
 200                 break;
 201         default:
 202                 goto invalid;
 203         }
 204
 205         for (;;) {
 206                 if (n-- < 1)
 207                         goto restart;
 208
 209                 _PUSH_PSENC(*s0++);
 210
 211                 switch (psenc->chlen) {
 212                 case 1:
 213                         if (_mb_singlebyte(_PSENC))
 214                                 goto convert;
 215                         if (_mb_leadbyte  (_PSENC))
 216                                 continue;
 217                         goto ilseq;
 218                 case 2:
 219                         if (_mb_trailbyte (_PSENC))
 220                                 goto convert;
 221                         if (ei->mb_cur_max == 4 &&
 222                             _mb_surrogate (_PSENC))
 223                                 continue;
 224                         goto ilseq;
 225                 case 3:
 226                         if (_mb_leadbyte  (_PSENC))
 227                                 continue;
 228                         goto ilseq;
 229                 case 4:
 230                         if (_mb_surrogate (_PSENC))
 231                                 goto convert;
 232                         goto ilseq;
 233                 }
 234         }
 235
 236 convert:
 237         len = psenc->chlen;
 238         s1  = &psenc->ch[0];
 239         wc  = 0;
 240         while (len-- > 0)
 241                 wc = (wc << 8) | (*s1++ & 0xff);
 242
 243         if (pwc != NULL)
 244                 *pwc = wc;
 245         *s = s0;
 246         *nresult = (wc == 0) ? 0 : psenc->chlen - chlenbak;
 247         /* _citrus_GBK2K_init_state(ei, psenc); */
 248         psenc->chlen = 0;
 249
 250         return (0);
 251
 252 restart:
 253         *s = s0;
 254         *nresult = (size_t)-2;
 255
 256         return (0);
 257
 258 invalid:
 259         return (EINVAL);
 260
 261 ilseq:
 262         *nresult = (size_t)-1;
 263         return (EILSEQ);
 264 }
 265
 266 static int
 267 _citrus_GBK2K_wcrtomb_priv(_GBK2KEncodingInfo * __restrict ei,
 268                            char * __restrict s, size_t n, wchar_t wc,
 269                            _GBK2KState * __restrict psenc,
 270                            size_t * __restrict nresult)
 271 {
 272         int len, ret;
 273
 274         _DIAGASSERT(ei != NULL);
 275         _DIAGASSERT(s != NULL);
 276         _DIAGASSERT(psenc != NULL);
 277
 278         if (psenc->chlen != 0) {
 279                 ret = EINVAL;
 280                 goto err;
 281         }
 282
 283         len = _mb_count(wc);
 284         if (n < len) {
 285                 ret = E2BIG;
 286                 goto err;
 287         }
 288
 289         switch (len) {
 290         case 1:
 291                 if (!_mb_singlebyte(_PUSH_PSENC(wc     ))) {
 292                         ret = EILSEQ;
 293                         goto err;
 294                 }
 295                 break;
 296         case 2:
 297                 if (!_mb_leadbyte  (_PUSH_PSENC(wc >> 8)) ||
 298                     !_mb_trailbyte (_PUSH_PSENC(wc     ))) {
 299                         ret = EILSEQ;
 300                         goto err;
 301                 }
 302                 break;
 303         case 4:
 304                 if (ei->mb_cur_max != 4 ||
 305                     !_mb_leadbyte  (_PUSH_PSENC(wc >> 24)) ||
 306                     !_mb_surrogate (_PUSH_PSENC(wc >> 16)) ||
 307                     !_mb_leadbyte  (_PUSH_PSENC(wc >>  8)) ||
 308                     !_mb_surrogate (_PUSH_PSENC(wc      ))) {
 309                         ret = EILSEQ;
 310                         goto err;
 311                 }
 312                 break;
 313         }
 314
 315         _DIAGASSERT(len == psenc->chlen);
 316
 317         memcpy(s, psenc->ch, psenc->chlen);
 318         *nresult = psenc->chlen;
 319         /* _citrus_GBK2K_init_state(ei, psenc); */
 320         psenc->chlen = 0;
 321
 322         return (0);
 323
 324 err:
 325         *nresult = (size_t)-1;
 326         return ret;
 327 }
 328
 329 static __inline int
 330 /*ARGSUSED*/
 331 _citrus_GBK2K_stdenc_wctocs(_GBK2KEncodingInfo * __restrict ei,
 332                             _csid_t * __restrict csid,
 333                             _index_t * __restrict idx, wchar_t wc)
 334 {
 335         u_int8_t ch, cl;
 336
 337         _DIAGASSERT(csid != NULL && idx != NULL);
 338
 339         if ((u_int32_t)wc<0x80) {
 340                 /* ISO646 */
 341                 *csid = 0;
 342                 *idx = (_index_t)wc;
 343         } else if ((u_int32_t)wc>=0x10000) {
 344                 /* GBKUCS : XXX */
 345                 *csid = 3;
 346                 *idx = (_index_t)wc;
 347         } else {
 348                 ch = (u_int8_t)(wc >> 8);
 349                 cl = (u_int8_t)wc;
 350                 if (ch>=0xA1 && cl>=0xA1) {
 351                         /* EUC G1 */
 352                         *csid = 1;
 353                         *idx = (_index_t)wc & 0x7F7FU;
 354                 } else {
 355                         /* extended area (0x8140-) */
 356                         *csid = 2;
 357                         *idx = (_index_t)wc;
 358                 }
 359         }
 360
 361         return 0;
 362 }
 363
 364 static __inline int
 365 /*ARGSUSED*/
 366 _citrus_GBK2K_stdenc_cstowc(_GBK2KEncodingInfo * __restrict ei,
 367                             wchar_t * __restrict wc,
 368                             _csid_t csid, _index_t idx)
 369 {
 370
 371         _DIAGASSERT(wc != NULL);
 372
 373         switch (csid) {
 374         case 0:
 375                 /* ISO646 */
 376                 *wc = (wchar_t)idx;
 377                 break;
 378         case 1:
 379                 /* EUC G1 */
 380                 *wc = (wchar_t)idx | 0x8080U;
 381                 break;
 382         case 2:
 383                 /* extended area */
 384                 *wc = (wchar_t)idx;
 385                 break;
 386         case 3:
 387                 /* GBKUCS : XXX */
 388                 if (ei->mb_cur_max != 4)
 389                         return EINVAL;
 390                 *wc = (wchar_t)idx;
 391                 break;
 392         default:
 393                 return EILSEQ;
 394         }
 395
 396         return 0;
 397 }
 398
 399 static __inline int
 400 /*ARGSUSED*/
 401 _citrus_GBK2K_stdenc_get_state_desc_generic(_GBK2KEncodingInfo * __restrict ei,
 402                                             _GBK2KState * __restrict psenc,
 403                                             int * __restrict rstate)
 404 {
 405
 406         if (psenc->chlen == 0)
 407                 *rstate = _STDENC_SDGEN_INITIAL;
 408         else
 409                 *rstate = _STDENC_SDGEN_INCOMPLETE_CHAR;
 410
 411         return 0;
 412 }
 413
 414 static int
 415 /*ARGSUSED*/
 416 _citrus_GBK2K_encoding_module_init(_GBK2KEncodingInfo * __restrict ei,
 417                                    const void * __restrict var, size_t lenvar)
 418 {
 419         const char *p;
 420
 421         _DIAGASSERT(ei != NULL);
 422
 423         p = var;
 424 #define MATCH(x, act)                                           \
 425 do {                                                            \
 426         if (lenvar >= (sizeof(#x)-1) &&                         \
 427             _bcs_strncasecmp(p, #x, sizeof(#x)-1) == 0) {       \
 428                 act;                                            \
 429                 lenvar -= sizeof(#x)-1;                         \
 430                 p += sizeof(#x)-1;                              \
 431         }                                                       \
 432 } while (/*CONSTCOND*/0)
 433         memset((void *)ei, 0, sizeof(*ei));
 434         ei->mb_cur_max = 4;
 435         while (lenvar>0) {
 436                 switch (_bcs_tolower(*p)) {
 437                 case '2':
 438                         MATCH("2byte", ei->mb_cur_max = 2);
 439                         break;
 440                 }
 441                 p++;
 442                 lenvar--;
 443         }
 444
 445         return (0);
 446 }
 447
 448 static void
 449 /*ARGSUSED*/
 450 _citrus_GBK2K_encoding_module_uninit(_GBK2KEncodingInfo *ei)
 451 {
 452 }
 453
 454
 455 /* ----------------------------------------------------------------------
 456  * public interface for ctype
 457  */
 458
 459 _CITRUS_CTYPE_DECLS(GBK2K);
 460 _CITRUS_CTYPE_DEF_OPS(GBK2K);
 461
 462 #include "citrus_ctype_template.h"
 463
 464 /* ----------------------------------------------------------------------
 465  * public interface for stdenc
 466  */
 467
 468 _CITRUS_STDENC_DECLS(GBK2K);
 469 _CITRUS_STDENC_DEF_OPS(GBK2K);
 470
 471 #include "citrus_stdenc_template.h"