lib/libc/locale/rune.c

   1 /*      $NetBSD: rune.c,v 1.46 2013/04/13 10:21:20 joerg Exp $  */
   2 /*-
   3  * Copyright (c)2010 Citrus Project,
   4  * All rights reserved.
   5  *
   6  * Redistribution and use in source and binary forms, with or without
   7  * modification, are permitted provided that the following conditions
   8  * are met:
   9  * 1. Redistributions of source code must retain the above copyright
  10  *    notice, this list of conditions and the following disclaimer.
  11  * 2. Redistributions in binary form must reproduce the above copyright
  12  *    notice, this list of conditions and the following disclaimer in the
  13  *    documentation and/or other materials provided with the distribution.
  14  *
  15  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
  16  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  17  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  18  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
  19  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  20  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  21  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  22  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  23  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  24  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  25  * SUCH DAMAGE.
  26  */
  27
  28 #include <sys/endian.h>
  29 #include <sys/mman.h>
  30 #include <sys/stat.h>
  31 #include <assert.h>
  32 #include <errno.h>
  33 #include <fcntl.h>
  34 #define __SETLOCALE_SOURCE__
  35 #include <locale.h>
  36 #include <stddef.h>
  37 #include <stdio.h>
  38 #include <stdlib.h>
  39 #include <string.h>
  40 #include <unistd.h>
  41 #include <wchar.h>
  42
  43 #include "setlocale_local.h"
  44
  45 #include "citrus_module.h"
  46 #include "citrus_ctype.h"
  47
  48 #include "runetype_local.h"
  49
  50 #include "multibyte.h"
  51
  52 #include "_wctype_local.h"
  53 #include "_wctrans_local.h"
  54
  55 typedef struct {
  56         _RuneLocale rl;
  57         unsigned short  rlp_ctype_tab  [_CTYPE_NUM_CHARS + 1];
  58         short           rlp_tolower_tab[_CTYPE_NUM_CHARS + 1];
  59         short           rlp_toupper_tab[_CTYPE_NUM_CHARS + 1];
  60         char            rlp_codeset[33]; /* XXX */
  61
  62 #ifdef __BUILD_LEGACY
  63         unsigned char   rlp_compat_bsdctype[_CTYPE_NUM_CHARS + 1];
  64 #endif
  65 } _RuneLocalePriv;
  66
  67 static __inline void
  68 _rune_wctype_init(_RuneLocale *rl)
  69 {
  70         memcpy(&rl->rl_wctype, &_DefaultRuneLocale.rl_wctype,
  71             sizeof(rl->rl_wctype));
  72 }
  73
  74 static __inline void
  75 _rune_wctrans_init(_RuneLocale *rl)
  76 {
  77         rl->rl_wctrans[_WCTRANS_INDEX_LOWER].te_name   = "tolower";
  78         rl->rl_wctrans[_WCTRANS_INDEX_LOWER].te_cached = &rl->rl_maplower[0];
  79         rl->rl_wctrans[_WCTRANS_INDEX_LOWER].te_extmap = &rl->rl_maplower_ext;
  80         rl->rl_wctrans[_WCTRANS_INDEX_UPPER].te_name   = "toupper";
  81         rl->rl_wctrans[_WCTRANS_INDEX_UPPER].te_cached = &rl->rl_mapupper[0];
  82         rl->rl_wctrans[_WCTRANS_INDEX_UPPER].te_extmap = &rl->rl_mapupper_ext;
  83 }
  84
  85 static __inline void
  86 _rune_init_priv(_RuneLocalePriv *rlp)
  87 {
  88 #if _CTYPE_CACHE_SIZE != _CTYPE_NUM_CHARS
  89         int i;
  90
  91         for (i = _CTYPE_CACHE_SIZE; i < _CTYPE_NUM_CHARS; ++i) {
  92                 rlp->rlp_ctype_tab  [i + 1] = 0;
  93                 rlp->rlp_tolower_tab[i + 1] = i;
  94                 rlp->rlp_toupper_tab[i + 1] = i;
  95
  96 #ifdef __BUILD_LEGACY
  97                 rlp->rlp_compat_bsdctype[i + 1] = 0;
  98 #endif
  99         }
 100 #endif
 101         rlp->rlp_ctype_tab  [0] = 0;
 102         rlp->rlp_tolower_tab[0] = EOF;
 103         rlp->rlp_toupper_tab[0] = EOF;
 104
 105         rlp->rl.rl_ctype_tab   = (const unsigned short *)&rlp->rlp_ctype_tab[0];
 106         rlp->rl.rl_tolower_tab = (const short *)&rlp->rlp_tolower_tab[0];
 107         rlp->rl.rl_toupper_tab = (const short *)&rlp->rlp_toupper_tab[0];
 108         rlp->rl.rl_codeset     = (const char *)&rlp->rlp_codeset[0];
 109
 110         _rune_wctype_init(&rlp->rl);
 111         _rune_wctrans_init(&rlp->rl);
 112
 113 #ifdef __BUILD_LEGACY
 114         rlp->rlp_compat_bsdctype[0] = 0;
 115         rlp->rl.rl_compat_bsdctype = (const unsigned char *)
 116             &rlp->rlp_compat_bsdctype[0];
 117 #endif
 118 }
 119
 120 static __inline void
 121 _rune_find_codeset(char *s, size_t n,
 122     char *var, size_t *plenvar)
 123 {
 124         size_t lenvar;
 125         const char *endvar;
 126
 127 #define _RUNE_CODESET_LEN (sizeof(_RUNE_CODESET)-1)
 128
 129         lenvar = *plenvar;
 130         for (/**/; lenvar > _RUNE_CODESET_LEN; ++var, --lenvar) {
 131                 if (!memcmp(var, _RUNE_CODESET, _RUNE_CODESET_LEN)) {
 132                         *var = '\0';
 133                         *plenvar -= lenvar;
 134                         endvar = &var[_RUNE_CODESET_LEN];
 135                         while (n-- > 1 && lenvar-- > _RUNE_CODESET_LEN) {
 136                                 if (*endvar == ' ' || *endvar == '\t')
 137                                         break;
 138                                 *s++ = *endvar++;
 139                         }
 140                         break;
 141                 }
 142         }
 143         *s = '\0';
 144 }
 145
 146 #ifdef __BUILD_LEGACY
 147 static __inline int
 148 _runetype_to_bsdctype(_RuneType bits)
 149 {
 150         int ret;
 151
 152         if (bits == (_RuneType)0)
 153                 return 0;
 154         ret = 0;
 155         if (bits & _RUNETYPE_U)
 156                 ret |= _COMPAT_U;
 157         if (bits & _RUNETYPE_L)
 158                 ret |= _COMPAT_L;
 159         if (bits & _RUNETYPE_D)
 160                 ret |= _COMPAT_N;
 161         if (bits & _RUNETYPE_S)
 162                 ret |= _COMPAT_S;
 163         if (bits & _RUNETYPE_P)
 164                 ret |= _COMPAT_P;
 165         if (bits & _RUNETYPE_C)
 166                 ret |= _COMPAT_C;
 167         if ((bits & (_RUNETYPE_X | _RUNETYPE_D)) == _RUNETYPE_X)
 168                 ret |= _COMPAT_X;
 169         if ((bits & (_RUNETYPE_R | _RUNETYPE_G)) == _RUNETYPE_R)
 170                 ret |= _COMPAT_B;
 171         return ret;
 172 }
 173 #endif /* __BUILD_LEGACY */
 174
 175 static __inline int
 176 _rune_read_file(const char * __restrict var, size_t lenvar,
 177     _RuneLocale ** __restrict prl)
 178 {
 179         int ret, i;
 180         const _FileRuneLocale *frl;
 181         const _FileRuneEntry *fre;
 182         const uint32_t *frune;
 183         _RuneLocalePriv *rlp;
 184         _RuneLocale *rl;
 185         _RuneEntry *re;
 186         uint32_t *rune;
 187         uint32_t runetype_len, maplower_len, mapupper_len, variable_len;
 188         size_t len, n;
 189
 190         if (lenvar < sizeof(*frl))
 191                 return EFTYPE;
 192         lenvar -= sizeof(*frl);
 193         frl = (const _FileRuneLocale *)(const void *)var;
 194         if (memcmp(_RUNECT10_MAGIC, &frl->frl_magic[0], sizeof(frl->frl_magic)))
 195                 return EFTYPE;
 196
 197         runetype_len = be32toh(frl->frl_runetype_ext.frr_nranges);
 198         maplower_len = be32toh(frl->frl_maplower_ext.frr_nranges);
 199         mapupper_len = be32toh(frl->frl_mapupper_ext.frr_nranges);
 200         len = runetype_len + maplower_len + mapupper_len;
 201
 202         fre = (const _FileRuneEntry *)(const void *)(frl + 1);
 203         frune = (const uint32_t *)(const void *)(fre + len);
 204
 205         variable_len = be32toh((uint32_t)frl->frl_variable_len);
 206
 207         n = len * sizeof(*fre);
 208         if (lenvar < n)
 209                 return EFTYPE;
 210         lenvar -= n;
 211
 212         n = sizeof(*rlp) + (len * sizeof(*re)) + lenvar;
 213         rlp = (_RuneLocalePriv *)malloc(n);
 214         if (rlp == NULL)
 215                 return ENOMEM;
 216         _rune_init_priv(rlp);
 217
 218         rl = &rlp->rl;
 219         re = (_RuneEntry *)(void *)(rlp + 1);
 220         rune = (uint32_t *)(void *)(re + len);
 221
 222         for (i = 0; i < _CTYPE_CACHE_SIZE; ++i) {
 223                 rl->rl_runetype[i] = be32toh(frl->frl_runetype[i]);
 224                 rl->rl_maplower[i] = be32toh((uint32_t)frl->frl_maplower[i]);
 225                 rl->rl_mapupper[i] = be32toh((uint32_t)frl->frl_mapupper[i]);
 226         }
 227
 228 #define READ_RANGE(name)                                                \
 229 do {                                                                    \
 230         const _FileRuneEntry *end_fre;                                  \
 231         const uint32_t *end_frune;                                      \
 232                                                                         \
 233         rl->rl_##name##_ext.rr_nranges = name##_len;                    \
 234         rl->rl_##name##_ext.rr_rune_ranges = re;                        \
 235                                                                         \
 236         end_fre = fre + name##_len;                                     \
 237         while (fre < end_fre) {                                         \
 238                 re->re_min = be32toh((uint32_t)fre->fre_min);           \
 239                 re->re_max = be32toh((uint32_t)fre->fre_max);           \
 240                 re->re_map = be32toh((uint32_t)fre->fre_map);           \
 241                 if (re->re_map != 0) {                                  \
 242                         re->re_rune_types = NULL;                       \
 243                 } else {                                                \
 244                         re->re_rune_types = rune;                       \
 245                         len = re->re_max - re->re_min + 1;              \
 246                         n = len * sizeof(*frune);                       \
 247                         if (lenvar < n) {                               \
 248                                 ret = EFTYPE;                           \
 249                                 goto err;                               \
 250                         }                                               \
 251                         lenvar -= n;                                    \
 252                         end_frune = frune + len;                        \
 253                         while (frune < end_frune)                       \
 254                                 *rune++ = be32toh(*frune++);            \
 255                 }                                                       \
 256                 ++fre, ++re;                                            \
 257         }                                                               \
 258 } while (/*CONSTCOND*/0)
 259
 260         READ_RANGE(runetype);
 261         READ_RANGE(maplower);
 262         READ_RANGE(mapupper);
 263
 264         if (lenvar < variable_len) {
 265                 ret = EFTYPE;
 266                 goto err;
 267         }
 268
 269         memcpy((void *)rune, (void const *)frune, variable_len);
 270         rl->rl_variable_len = variable_len;
 271         rl->rl_variable = (void *)rune;
 272
 273         _rune_find_codeset(rlp->rlp_codeset, sizeof(rlp->rlp_codeset),
 274             (char *)rl->rl_variable, &rl->rl_variable_len);
 275
 276         ret = _citrus_ctype_open(&rl->rl_citrus_ctype, frl->frl_encoding,
 277             rl->rl_variable, rl->rl_variable_len, _PRIVSIZE);
 278         if (ret)
 279                 goto err;
 280         if (__mb_len_max_runtime <
 281             _citrus_ctype_get_mb_cur_max(rl->rl_citrus_ctype)) {
 282                 ret = EINVAL;
 283                 goto err;
 284         }
 285
 286         for (i = 0; i < _CTYPE_CACHE_SIZE; ++i) {
 287                 wint_t wc;
 288                 _RuneType rc;
 289
 290                 ret = _citrus_ctype_btowc(rl->rl_citrus_ctype, i, &wc);
 291                 if (ret)
 292                         goto err;
 293                 if (wc == WEOF) {
 294                         rlp->rlp_ctype_tab[i + 1] = 0;
 295                         rlp->rlp_tolower_tab[i + 1] = i;
 296                         rlp->rlp_toupper_tab[i + 1] = i;
 297                 } else {
 298                         rc = _runetype_priv(rl, wc);
 299                         rlp->rlp_ctype_tab[i + 1] = (unsigned short)
 300                             ((rc & ~_RUNETYPE_SWM) >> 8);
 301
 302 #ifdef __BUILD_LEGACY
 303                         rlp->rlp_compat_bsdctype[i + 1]
 304                           = _runetype_to_bsdctype(rc);
 305 #endif
 306
 307 #define CONVERT_MAP(name)                                               \
 308 do {                                                                    \
 309         wint_t map;                                                     \
 310         int c;                                                          \
 311                                                                         \
 312         map = _towctrans_priv(wc, _wctrans_##name(rl));                 \
 313         if (map == wc || (_citrus_ctype_wctob(rl->rl_citrus_ctype,      \
 314             map, &c)  || c == EOF))                                     \
 315                 c = i;                                                  \
 316         rlp->rlp_to##name##_tab[i + 1] = (short)c;                      \
 317 } while (/*CONSTCOND*/0)
 318
 319                         CONVERT_MAP(lower);
 320                         CONVERT_MAP(upper);
 321                 }
 322         }
 323         *prl = rl;
 324         return 0;
 325
 326 err:
 327         free(rlp);
 328         return ret;
 329 }
 330
 331 int
 332 _rune_load(const char * __restrict var, size_t lenvar,
 333     _RuneLocale ** __restrict prl)
 334 {
 335         int ret;
 336
 337         _DIAGASSERT(var != NULL || lenvar < 1);
 338         _DIAGASSERT(prl != NULL);
 339
 340         if (lenvar < 1)
 341                 return EFTYPE;
 342         switch (*var) {
 343         case 'R':
 344                 ret = _rune_read_file(var, lenvar, prl);
 345                 break;
 346         default:
 347                 ret = EFTYPE;
 348         }
 349         return ret;
 350 }