lib/libc/citrus/modules/citrus_dechanyu.c

   1 /* $NetBSD: citrus_dechanyu.c,v 1.2 2007/04/24 15:42:08 tnozaki Exp $ */
   2 /* $DragonFly: src/lib/libc/citrus/modules/citrus_dechanyu.c,v 1.1 2008/04/10 10:21:01 hasso Exp $ */
   3
   4 /*-
   5  * Copyright (c)2007 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 <string.h>
  34 #include <stdint.h>
  35 #include <stdio.h>
  36 #include <stdlib.h>
  37 #include <stddef.h>
  38 #include <locale.h>
  39 #include <wchar.h>
  40 #include <limits.h>
  41
  42 #include "citrus_namespace.h"
  43 #include "citrus_types.h"
  44 #include "citrus_bcs.h"
  45 #include "citrus_module.h"
  46 #include "citrus_ctype.h"
  47 #include "citrus_stdenc.h"
  48 #include "citrus_dechanyu.h"
  49
  50 /* ----------------------------------------------------------------------
  51  * private stuffs used by templates
  52  */
  53
  54 typedef struct {
  55         int chlen;
  56         char ch[4];
  57 } _DECHanyuState;
  58
  59 typedef struct {
  60         int dummy;
  61 } _DECHanyuEncodingInfo;
  62
  63 typedef struct {
  64         _DECHanyuEncodingInfo   ei;
  65         struct {
  66                 /* for future multi-locale facility */
  67                 _DECHanyuState  s_mblen;
  68                 _DECHanyuState  s_mbrlen;
  69                 _DECHanyuState  s_mbrtowc;
  70                 _DECHanyuState  s_mbtowc;
  71                 _DECHanyuState  s_mbsrtowcs;
  72                 _DECHanyuState  s_wcrtomb;
  73                 _DECHanyuState  s_wcsrtombs;
  74                 _DECHanyuState  s_wctomb;
  75         } states;
  76 } _DECHanyuCTypeInfo;
  77
  78 #define _CEI_TO_EI(_cei_)               (&(_cei_)->ei)
  79 #define _CEI_TO_STATE(_cei_, _func_)    (_cei_)->states.__CONCAT(s_,_func_)
  80
  81 #define _FUNCNAME(m)                    __CONCAT(_citrus_DECHanyu_,m)
  82 #define _ENCODING_INFO                  _DECHanyuEncodingInfo
  83 #define _CTYPE_INFO                     _DECHanyuCTypeInfo
  84 #define _ENCODING_STATE                 _DECHanyuState
  85 #define _ENCODING_MB_CUR_MAX(_ei_)              4
  86 #define _ENCODING_IS_STATE_DEPENDENT            0
  87 #define _STATE_NEEDS_EXPLICIT_INIT(_ps_)        0
  88
  89 static __inline void
  90 /*ARGSUSED*/
  91 _citrus_DECHanyu_init_state(_DECHanyuEncodingInfo * __restrict ei,
  92         _DECHanyuState * __restrict psenc)
  93 {
  94         /* ei may be null */
  95         _DIAGASSERT(psenc != NULL);
  96
  97         psenc->chlen = 0;
  98 }
  99
 100 static __inline void
 101 /*ARGSUSED*/
 102 _citrus_DECHanyu_pack_state(_DECHanyuEncodingInfo * __restrict ei,
 103         void * __restrict pspriv,
 104         const _DECHanyuState * __restrict psenc)
 105 {
 106         /* ei may be null */
 107         _DIAGASSERT(pspriv != NULL);
 108         _DIAGASSERT(psenc != NULL);
 109
 110         memcpy(pspriv, (const void *)psenc, sizeof(*psenc));
 111 }
 112
 113 static __inline void
 114 /*ARGSUSED*/
 115 _citrus_DECHanyu_unpack_state(_DECHanyuEncodingInfo * __restrict ei,
 116         _DECHanyuState * __restrict psenc,
 117         const void * __restrict pspriv)
 118 {
 119         /* ei may be null */
 120         _DIAGASSERT(psenc != NULL);
 121         _DIAGASSERT(pspriv != NULL);
 122
 123         memcpy((void *)psenc, pspriv, sizeof(*psenc));
 124 }
 125
 126 static void
 127 /*ARGSUSED*/
 128 _citrus_DECHanyu_encoding_module_uninit(_DECHanyuEncodingInfo *ei)
 129 {
 130         /* ei may be null */
 131 }
 132
 133 static int
 134 /*ARGSUSED*/
 135 _citrus_DECHanyu_encoding_module_init(_DECHanyuEncodingInfo * __restrict ei,
 136         const void * __restrict var, size_t lenvar)
 137 {
 138         /* ei may be null */
 139         return 0;
 140 }
 141
 142 static __inline int
 143 is_singlebyte(int c)
 144 {
 145         return c <= 0x7F;
 146 }
 147
 148 static __inline int
 149 is_leadbyte(int c)
 150 {
 151         return c >= 0xA1 && c <= 0xFE;
 152 }
 153
 154 static __inline int
 155 is_trailbyte(int c)
 156 {
 157         c &= ~0x80;
 158         return c >= 0x21 && c <= 0x7E;
 159 }
 160
 161 static __inline int
 162 is_hanyu1(int c)
 163 {
 164         return c == 0xC2;
 165 }
 166
 167 static __inline int
 168 is_hanyu2(int c)
 169 {
 170         return c == 0xCB;
 171 }
 172
 173 #define HANYUBIT        0xC2CB0000
 174
 175 static __inline int
 176 is_94charset(int c)
 177 {
 178         return c >= 0x21 && c <= 0x7E;
 179 }
 180
 181 static int
 182 /*ARGSUSED*/
 183 _citrus_DECHanyu_mbrtowc_priv(_DECHanyuEncodingInfo * __restrict ei,
 184         wchar_t * __restrict pwc, const char ** __restrict s, size_t n,
 185         _DECHanyuState * __restrict psenc, size_t * __restrict nresult)
 186 {
 187         const char *s0;
 188         int ch, i;
 189         wchar_t wc;
 190
 191         /* ei may be unused */
 192         _DIAGASSERT(s != NULL);
 193         _DIAGASSERT(psenc != NULL);
 194         _DIAGASSERT(nresult != NULL);
 195
 196         if (*s == NULL) {
 197                 _citrus_DECHanyu_init_state(ei, psenc);
 198                 *nresult = _ENCODING_IS_STATE_DEPENDENT;
 199                 return 0;
 200         }
 201         s0 = *s;
 202
 203         wc = (wchar_t)0;
 204         switch (psenc->chlen) {
 205         case 0:
 206                 if (n-- < 1)
 207                         goto restart;
 208                 ch = *s0++ & 0xFF;
 209                 if (is_singlebyte(ch) != 0) {
 210                         if (pwc != NULL)
 211                                 *pwc = (wchar_t)ch;
 212                         *nresult = (size_t)((ch == 0) ? 0 : 1);
 213                         *s = s0;
 214                         return 0;
 215                 }
 216                 if (is_leadbyte(ch) == 0)
 217                         goto ilseq;
 218                 psenc->ch[psenc->chlen++] = ch;
 219                 break;
 220         case 1:
 221                 ch = psenc->ch[0] & 0xFF;
 222                 if (is_leadbyte(ch) == 0)
 223                         return EINVAL;
 224                 break;
 225         case 2: case 3:
 226                 ch = psenc->ch[0] & 0xFF;
 227                 if (is_hanyu1(ch) != 0) {
 228                         ch = psenc->ch[1] & 0xFF;
 229                         if (is_hanyu2(ch) != 0) {
 230                                 wc |= (wchar_t)HANYUBIT;
 231                                 break;
 232                         }
 233                 }
 234         /*FALLTHROUGH*/
 235         default:
 236                 return EINVAL;
 237         }
 238
 239         switch (psenc->chlen) {
 240         case 1:
 241                 if (is_hanyu1(ch) != 0) {
 242                         if (n-- < 1)
 243                                 goto restart;
 244                         ch = *s0++ & 0xFF;
 245                         if (is_hanyu2(ch) == 0)
 246                                 goto ilseq;
 247                         psenc->ch[psenc->chlen++] = ch;
 248                         wc |= (wchar_t)HANYUBIT;
 249                         if (n-- < 1)
 250                                 goto restart;
 251                         ch = *s0++ & 0xFF;
 252                         if (is_leadbyte(ch) == 0)
 253                                 goto ilseq;
 254                         psenc->ch[psenc->chlen++] = ch;
 255                 }
 256                 break;
 257         case 2:
 258                 if (n-- < 1)
 259                         goto restart;
 260                 ch = *s0++ & 0xFF;
 261                 if (is_leadbyte(ch) == 0)
 262                         goto ilseq;
 263                 psenc->ch[psenc->chlen++] = ch;
 264                 break;
 265         case 3:
 266                 ch = psenc->ch[2] & 0xFF;
 267                 if (is_leadbyte(ch) == 0)
 268                         return EINVAL;
 269         }
 270         if (n-- < 1)
 271                 goto restart;
 272         wc |= (wchar_t)(ch << 8);
 273         ch = *s0++ & 0xFF;
 274         if (is_trailbyte(ch) == 0)
 275                 goto ilseq;
 276         wc |= (wchar_t)ch;
 277         if (pwc != NULL)
 278                 *pwc = wc;
 279         *nresult = (size_t)(s0 - *s);
 280         *s = s0;
 281         psenc->chlen = 0;
 282
 283         return 0;
 284
 285 restart:
 286         *nresult = (size_t)-2;
 287         *s = s0;
 288         return 0;
 289
 290 ilseq:
 291         *nresult = (size_t)-1;
 292         return EILSEQ;
 293 }
 294
 295 static int
 296 /*ARGSUSED*/
 297 _citrus_DECHanyu_wcrtomb_priv(_DECHanyuEncodingInfo * __restrict ei,
 298         char * __restrict s, size_t n, wchar_t wc,
 299         _DECHanyuState * __restrict psenc, size_t * __restrict nresult)
 300 {
 301         int ch;
 302
 303         /* ei may be unused */
 304         _DIAGASSERT(s != NULL);
 305         _DIAGASSERT(psenc != NULL);
 306         _DIAGASSERT(nresult != NULL);
 307
 308         if (psenc->chlen != 0)
 309                 return EINVAL;
 310
 311         /* XXX: assume wchar_t as int */
 312         if ((uint32_t)wc <= 0x7F) {
 313                 ch = wc & 0xFF;
 314         } else {
 315                 if ((uint32_t)wc > 0xFFFF) {
 316                         if ((wc & ~0xFFFF) != HANYUBIT)
 317                                 goto ilseq;
 318                         psenc->ch[psenc->chlen++] = (wc >> 24) & 0xFF;
 319                         psenc->ch[psenc->chlen++] = (wc >> 16) & 0xFF;
 320                         wc &= 0xFFFF;
 321                 }
 322                 ch = (wc >> 8) & 0xFF;
 323                 if (!is_leadbyte(ch))
 324                         goto ilseq;
 325                 psenc->ch[psenc->chlen++] = ch;
 326                 ch = wc & 0xFF;
 327                 if (is_trailbyte(ch) == 0)
 328                         goto ilseq;
 329         }
 330         psenc->ch[psenc->chlen++] = ch;
 331         if (n < psenc->chlen) {
 332                 *nresult = (size_t)-1;
 333                 return E2BIG;
 334         }
 335         memcpy(s, psenc->ch, psenc->chlen);
 336         *nresult = psenc->chlen;
 337         psenc->chlen = 0;
 338
 339         return 0;
 340
 341 ilseq:
 342         *nresult = (size_t)-1;
 343         return EILSEQ;
 344 }
 345
 346 static __inline int
 347 /*ARGSUSED*/
 348 _citrus_DECHanyu_stdenc_wctocs(_DECHanyuEncodingInfo * __restrict ei,
 349         _csid_t * __restrict csid, _index_t * __restrict idx, wchar_t wc)
 350 {
 351         int plane;
 352         wchar_t mask;
 353
 354         /* ei may be unused */
 355         _DIAGASSERT(csid != NULL);
 356         _DIAGASSERT(idx != NULL);
 357
 358         plane = 0;
 359         mask = 0x7F;
 360         /* XXX: assume wchar_t as int */
 361         if ((uint32_t)wc > 0x7F) {
 362                 if ((uint32_t)wc > 0xFFFF) {
 363                         if ((wc & ~0xFFFF) != HANYUBIT)
 364                                 return EILSEQ;
 365                         plane += 2;
 366                 }
 367                 if (is_leadbyte((wc >> 8) & 0xFF) == 0 ||
 368                     is_trailbyte(wc & 0xFF) == 0)
 369                         return EILSEQ;
 370                 plane += (wc & 0x80) ? 1 : 2;
 371                 mask |= 0x7F00;
 372         }
 373         *csid = plane;
 374         *idx = (_index_t)(wc & mask);
 375
 376         return 0;
 377 }
 378
 379 static __inline int
 380 /*ARGSUSED*/
 381 _citrus_DECHanyu_stdenc_cstowc(_DECHanyuEncodingInfo * __restrict ei,
 382         wchar_t * __restrict wc, _csid_t csid, _index_t idx)
 383 {
 384         /* ei may be unused */
 385         _DIAGASSERT(wc != NULL);
 386
 387         if (csid == 0) {
 388                 if (idx > 0x7F)
 389                         return EILSEQ;
 390         } else if (csid <= 4) {
 391                 if (is_94charset(idx >> 8) == 0)
 392                         return EILSEQ;
 393                 if (is_94charset(idx & 0xFF) == 0)
 394                         return EILSEQ;
 395                 if (csid % 2)
 396                         idx |= 0x80;
 397                 idx |= 0x8000;
 398                 if (csid > 2)
 399                         idx |= HANYUBIT;
 400         } else
 401                 return EILSEQ;
 402         *wc = (wchar_t)idx;
 403         return 0;
 404 }
 405
 406 static __inline int
 407 /*ARGSUSED*/
 408 _citrus_DECHanyu_stdenc_get_state_desc_generic(
 409         _DECHanyuEncodingInfo * __restrict ei,
 410         _DECHanyuState * __restrict psenc, int * __restrict rstate)
 411 {
 412         /* ei may be unused */
 413         _DIAGASSERT(psenc != NULL);
 414         _DIAGASSERT(rstate != NULL);
 415
 416         *rstate = (psenc->chlen == 0)
 417             ? _STDENC_SDGEN_INITIAL
 418             : _STDENC_SDGEN_INCOMPLETE_CHAR;
 419         return 0;
 420 }
 421
 422 /* ----------------------------------------------------------------------
 423  * public interface for ctype
 424  */
 425
 426 _CITRUS_CTYPE_DECLS(DECHanyu);
 427 _CITRUS_CTYPE_DEF_OPS(DECHanyu);
 428
 429 #include "citrus_ctype_template.h"
 430
 431
 432 /* ----------------------------------------------------------------------
 433  * public interface for stdenc
 434  */
 435
 436 _CITRUS_STDENC_DECLS(DECHanyu);
 437 _CITRUS_STDENC_DEF_OPS(DECHanyu);
 438
 439 #include "citrus_stdenc_template.h"