iconv/gconv_trans.c

   1 /* Transliteration using the locale's data.
   2    Copyright (C) 2000-2024 Free Software Foundation, Inc.
   3    This file is part of the GNU C Library.
   4
   5    The GNU C Library is free software; you can redistribute it and/or
   6    modify it under the terms of the GNU Lesser General Public
   7    License as published by the Free Software Foundation; either
   8    version 2.1 of the License, or (at your option) any later version.
   9
  10    The GNU C Library is distributed in the hope that it will be useful,
  11    but WITHOUT ANY WARRANTY; without even the implied warranty of
  12    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  13    Lesser General Public License for more details.
  14
  15    You should have received a copy of the GNU Lesser General Public
  16    License along with the GNU C Library; if not, see
  17    <https://www.gnu.org/licenses/>.  */
  18
  19 #include <assert.h>
  20 #include <dlfcn.h>
  21 #include <search.h>
  22 #include <stdint.h>
  23 #include <string.h>
  24 #include <stdlib.h>
  25
  26 #include <libc-lock.h>
  27 #include "gconv_int.h"
  28 #include "../locale/localeinfo.h"
  29 #include <pointer_guard.h>
  30
  31
  32 int
  33 __gconv_transliterate (struct __gconv_step *step,
  34                        struct __gconv_step_data *step_data,
  35                        const unsigned char *inbufstart,
  36                        const unsigned char **inbufp,
  37                        const unsigned char *inbufend,
  38                        unsigned char **outbufstart, size_t *irreversible)
  39 {
  40   /* Find out about the locale's transliteration.  */
  41   uint32_t size;
  42   const uint32_t *from_idx;
  43   const uint32_t *from_tbl;
  44   const uint32_t *to_idx;
  45   const uint32_t *to_tbl;
  46   const uint32_t *winbuf;
  47   const uint32_t *winbufend;
  48   uint32_t low;
  49   uint32_t high;
  50
  51   /* The input buffer.  There are actually 4-byte values.  */
  52   winbuf = (const uint32_t *) *inbufp;
  53   winbufend = (const uint32_t *) inbufend;
  54
  55   __gconv_fct fct = step->__fct;
  56   if (step->__shlib_handle != NULL)
  57     PTR_DEMANGLE (fct);
  58
  59   /* If there is no transliteration information in the locale don't do
  60      anything and return the error.  */
  61   size = _NL_CURRENT_WORD (LC_CTYPE, _NL_CTYPE_TRANSLIT_TAB_SIZE);
  62   if (size == 0)
  63     goto no_rules;
  64
  65   /* Get the rest of the values.  */
  66   from_idx =
  67     (const uint32_t *) _NL_CURRENT (LC_CTYPE, _NL_CTYPE_TRANSLIT_FROM_IDX);
  68   from_tbl =
  69     (const uint32_t *) _NL_CURRENT (LC_CTYPE, _NL_CTYPE_TRANSLIT_FROM_TBL);
  70   to_idx =
  71     (const uint32_t *) _NL_CURRENT (LC_CTYPE, _NL_CTYPE_TRANSLIT_TO_IDX);
  72   to_tbl =
  73     (const uint32_t *) _NL_CURRENT (LC_CTYPE, _NL_CTYPE_TRANSLIT_TO_TBL);
  74
  75   /* Test whether there is enough input.  */
  76   if (winbuf + 1 > winbufend)
  77     return (winbuf == winbufend
  78             ? __GCONV_EMPTY_INPUT : __GCONV_INCOMPLETE_INPUT);
  79
  80   /* The array starting at FROM_IDX contains indices to the string table
  81      in FROM_TBL.  The indices are sorted wrt to the strings.  I.e., we
  82      are doing binary search.  */
  83   low = 0;
  84   high = size;
  85   while (low < high)
  86     {
  87       uint32_t med = (low + high) / 2;
  88       uint32_t idx;
  89       int cnt;
  90
  91       /* Compare the string at this index with the string at the current
  92          position in the input buffer.  */
  93       idx = from_idx[med];
  94       cnt = 0;
  95       do
  96         {
  97           if (from_tbl[idx + cnt] != winbuf[cnt])
  98             /* Does not match.  */
  99             break;
 100           ++cnt;
 101         }
 102       while (from_tbl[idx + cnt] != L'\0' && winbuf + cnt < winbufend);
 103
 104       if (cnt > 0 && from_tbl[idx + cnt] == L'\0')
 105         {
 106           /* Found a matching input sequence.  Now try to convert the
 107              possible replacements.  */
 108           uint32_t idx2 = to_idx[med];
 109
 110           do
 111             {
 112               /* Determine length of replacement.  */
 113               unsigned int len = 0;
 114               int res;
 115               const unsigned char *toinptr;
 116               unsigned char *outptr;
 117
 118               while (to_tbl[idx2 + len] != L'\0')
 119                 ++len;
 120
 121               /* Try this input text.  */
 122               toinptr = (const unsigned char *) &to_tbl[idx2];
 123               outptr = *outbufstart;
 124               res = DL_CALL_FCT (fct,
 125                                  (step, step_data, &toinptr,
 126                                   (const unsigned char *) &to_tbl[idx2 + len],
 127                                   &outptr, NULL, 0, 0));
 128               if (res != __GCONV_ILLEGAL_INPUT)
 129                 {
 130                   /* If the conversion succeeds we have to increment the
 131                      input buffer.  */
 132                   if (res == __GCONV_EMPTY_INPUT)
 133                     {
 134                       *inbufp += cnt * sizeof (uint32_t);
 135                       ++*irreversible;
 136                       res = __GCONV_OK;
 137                     }
 138                   /* Do not increment the output pointer if we could not
 139                      store the entire output. */
 140                   if (res != __GCONV_FULL_OUTPUT)
 141                     *outbufstart = outptr;
 142
 143                   return res;
 144                 }
 145
 146               /* Next replacement.  */
 147               idx2 += len + 1;
 148             }
 149           while (to_tbl[idx2] != L'\0');
 150
 151           /* Nothing found, continue searching.  */
 152         }
 153       else if (cnt > 0)
 154         /* This means that the input buffer contents matches a prefix of
 155            an entry.  Since we cannot match it unless we get more input,
 156            we will tell the caller about it.  */
 157         return __GCONV_INCOMPLETE_INPUT;
 158
 159       if (winbuf + cnt >= winbufend || from_tbl[idx + cnt] < winbuf[cnt])
 160         low = med + 1;
 161       else
 162         high = med;
 163     }
 164
 165  no_rules:
 166   /* Maybe the character is supposed to be ignored.  */
 167   if (_NL_CURRENT_WORD (LC_CTYPE, _NL_CTYPE_TRANSLIT_IGNORE_LEN) != 0)
 168     {
 169       int n = _NL_CURRENT_WORD (LC_CTYPE, _NL_CTYPE_TRANSLIT_IGNORE_LEN);
 170       const uint32_t *ranges =
 171         (const uint32_t *) _NL_CURRENT (LC_CTYPE, _NL_CTYPE_TRANSLIT_IGNORE);
 172       const uint32_t wc = *(const uint32_t *) (*inbufp);
 173       int i;
 174
 175       /* Test whether there is enough input.  */
 176       if (winbuf + 1 > winbufend)
 177         return (winbuf == winbufend
 178                 ? __GCONV_EMPTY_INPUT : __GCONV_INCOMPLETE_INPUT);
 179
 180       for (i = 0; i < n; ranges += 3, ++i)
 181         if (ranges[0] <= wc && wc <= ranges[1]
 182             && (wc - ranges[0]) % ranges[2] == 0)
 183           {
 184             /* Matches the range.  Ignore it.  */
 185             *inbufp += 4;
 186             ++*irreversible;
 187             return __GCONV_OK;
 188           }
 189         else if (wc < ranges[0])
 190           /* There cannot be any other matching range since they are
 191              sorted.  */
 192           break;
 193     }
 194
 195   /* One last chance: use the default replacement.  */
 196   if (_NL_CURRENT_WORD (LC_CTYPE, _NL_CTYPE_TRANSLIT_DEFAULT_MISSING_LEN) != 0)
 197     {
 198       const uint32_t *default_missing = (const uint32_t *)
 199         _NL_CURRENT (LC_CTYPE, _NL_CTYPE_TRANSLIT_DEFAULT_MISSING);
 200       const unsigned char *toinptr = (const unsigned char *) default_missing;
 201       uint32_t len = _NL_CURRENT_WORD (LC_CTYPE,
 202                                        _NL_CTYPE_TRANSLIT_DEFAULT_MISSING_LEN);
 203       unsigned char *outptr;
 204       int res;
 205
 206       /* Test whether there is enough input.  */
 207       if (winbuf + 1 > winbufend)
 208         return (winbuf == winbufend
 209                 ? __GCONV_EMPTY_INPUT : __GCONV_INCOMPLETE_INPUT);
 210
 211       outptr = *outbufstart;
 212       res = DL_CALL_FCT (fct,
 213                          (step, step_data, &toinptr,
 214                           (const unsigned char *) (default_missing + len),
 215                           &outptr, NULL, 0, 0));
 216
 217       if (res != __GCONV_ILLEGAL_INPUT)
 218         {
 219           /* If the conversion succeeds we have to increment the
 220              input buffer.  */
 221           if (res == __GCONV_EMPTY_INPUT)
 222             {
 223               /* This worked but is not reversible.  */
 224               ++*irreversible;
 225               *inbufp += 4;
 226               res = __GCONV_OK;
 227             }
 228           *outbufstart = outptr;
 229
 230           return res;
 231         }
 232     }
 233
 234   /* Haven't found a match.  */
 235   return __GCONV_ILLEGAL_INPUT;
 236 }
 237 libc_hidden_def (__gconv_transliterate)