iconv/gconv_trans.c

   1 /* Transliteration using the locale's data.
   2    Copyright (C) 2000-2014 Free Software Foundation, Inc.
   3    This file is part of the GNU C Library.
   4    Contributed by Ulrich Drepper <drepper@cygnus.com>, 2000.
   5
   6    The GNU C Library is free software; you can redistribute it and/or
   7    modify it under the terms of the GNU Lesser General Public
   8    License as published by the Free Software Foundation; either
   9    version 2.1 of the License, or (at your option) any later version.
  10
  11    The GNU C Library is distributed in the hope that it will be useful,
  12    but WITHOUT ANY WARRANTY; without even the implied warranty of
  13    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14    Lesser General Public License for more details.
  15
  16    You should have received a copy of the GNU Lesser General Public
  17    License along with the GNU C Library; if not, see
  18    <http://www.gnu.org/licenses/>.  */
  19
  20 #include <assert.h>
  21 #include <dlfcn.h>
  22 #include <search.h>
  23 #include <stdint.h>
  24 #include <string.h>
  25 #include <stdlib.h>
  26
  27 #include <bits/libc-lock.h>
  28 #include "gconv_int.h"
  29 #include "../locale/localeinfo.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   uint_fast32_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   uint_fast32_t low;
  49   uint_fast32_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 #ifdef PTR_DEMANGLE
  57   if (step->__shlib_handle != NULL)
  58     PTR_DEMANGLE (fct);
  59 #endif
  60
  61   /* If there is no transliteration information in the locale don't do
  62      anything and return the error.  */
  63   size = _NL_CURRENT_WORD (LC_CTYPE, _NL_CTYPE_TRANSLIT_TAB_SIZE);
  64   if (size == 0)
  65     goto no_rules;
  66
  67   /* Get the rest of the values.  */
  68   from_idx =
  69     (const uint32_t *) _NL_CURRENT (LC_CTYPE, _NL_CTYPE_TRANSLIT_FROM_IDX);
  70   from_tbl =
  71     (const uint32_t *) _NL_CURRENT (LC_CTYPE, _NL_CTYPE_TRANSLIT_FROM_TBL);
  72   to_idx =
  73     (const uint32_t *) _NL_CURRENT (LC_CTYPE, _NL_CTYPE_TRANSLIT_TO_IDX);
  74   to_tbl =
  75     (const uint32_t *) _NL_CURRENT (LC_CTYPE, _NL_CTYPE_TRANSLIT_TO_TBL);
  76
  77   /* Test whether there is enough input.  */
  78   if (winbuf + 1 > winbufend)
  79     return (winbuf == winbufend
  80             ? __GCONV_EMPTY_INPUT : __GCONV_INCOMPLETE_INPUT);
  81
  82   /* The array starting at FROM_IDX contains indeces to the string table
  83      in FROM_TBL.  The indeces are sorted wrt to the strings.  I.e., we
  84      are doing binary search.  */
  85   low = 0;
  86   high = size;
  87   while (low < high)
  88     {
  89       uint_fast32_t med = (low + high) / 2;
  90       uint32_t idx;
  91       int cnt;
  92
  93       /* Compare the string at this index with the string at the current
  94          position in the input buffer.  */
  95       idx = from_idx[med];
  96       cnt = 0;
  97       do
  98         {
  99           if (from_tbl[idx + cnt] != winbuf[cnt])
 100             /* Does not match.  */
 101             break;
 102           ++cnt;
 103         }
 104       while (from_tbl[idx + cnt] != L'\0' && winbuf + cnt < winbufend);
 105
 106       if (cnt > 0 && from_tbl[idx + cnt] == L'\0')
 107         {
 108           /* Found a matching input sequence.  Now try to convert the
 109              possible replacements.  */
 110           uint32_t idx2 = to_idx[med];
 111
 112           do
 113             {
 114               /* Determine length of replacement.  */
 115               uint_fast32_t len = 0;
 116               int res;
 117               const unsigned char *toinptr;
 118               unsigned char *outptr;
 119
 120               while (to_tbl[idx2 + len] != L'\0')
 121                 ++len;
 122
 123               /* Try this input text.  */
 124               toinptr = (const unsigned char *) &to_tbl[idx2];
 125               outptr = *outbufstart;
 126               res = DL_CALL_FCT (fct,
 127                                  (step, step_data, &toinptr,
 128                                   (const unsigned char *) &to_tbl[idx2 + len],
 129                                   &outptr, NULL, 0, 0));
 130               if (res != __GCONV_ILLEGAL_INPUT)
 131                 {
 132                   /* If the conversion succeeds we have to increment the
 133                      input buffer.  */
 134                   if (res == __GCONV_EMPTY_INPUT)
 135                     {
 136                       *inbufp += cnt * sizeof (uint32_t);
 137                       ++*irreversible;
 138                       res = __GCONV_OK;
 139                     }
 140                   /* Do not increment the output pointer if we could not
 141                      store the entire output. */
 142                   if (res != __GCONV_FULL_OUTPUT)
 143                     *outbufstart = outptr;
 144
 145                   return res;
 146                 }
 147
 148               /* Next replacement.  */
 149               idx2 += len + 1;
 150             }
 151           while (to_tbl[idx2] != L'\0');
 152
 153           /* Nothing found, continue searching.  */
 154         }
 155       else if (cnt > 0)
 156         /* This means that the input buffer contents matches a prefix of
 157            an entry.  Since we cannot match it unless we get more input,
 158            we will tell the caller about it.  */
 159         return __GCONV_INCOMPLETE_INPUT;
 160
 161       if (winbuf + cnt >= winbufend || from_tbl[idx + cnt] < winbuf[cnt])
 162         low = med + 1;
 163       else
 164         high = med;
 165     }
 166
 167  no_rules:
 168   /* Maybe the character is supposed to be ignored.  */
 169   if (_NL_CURRENT_WORD (LC_CTYPE, _NL_CTYPE_TRANSLIT_IGNORE_LEN) != 0)
 170     {
 171       int n = _NL_CURRENT_WORD (LC_CTYPE, _NL_CTYPE_TRANSLIT_IGNORE_LEN);
 172       const uint32_t *ranges =
 173         (const uint32_t *) _NL_CURRENT (LC_CTYPE, _NL_CTYPE_TRANSLIT_IGNORE);
 174       const uint32_t wc = *(const uint32_t *) (*inbufp);
 175       int i;
 176
 177       /* Test whether there is enough input.  */
 178       if (winbuf + 1 > winbufend)
 179         return (winbuf == winbufend
 180                 ? __GCONV_EMPTY_INPUT : __GCONV_INCOMPLETE_INPUT);
 181
 182       for (i = 0; i < n; ranges += 3, ++i)
 183         if (ranges[0] <= wc && wc <= ranges[1]
 184             && (wc - ranges[0]) % ranges[2] == 0)
 185           {
 186             /* Matches the range.  Ignore it.  */
 187             *inbufp += 4;
 188             ++*irreversible;
 189             return __GCONV_OK;
 190           }
 191         else if (wc < ranges[0])
 192           /* There cannot be any other matching range since they are
 193              sorted.  */
 194           break;
 195     }
 196
 197   /* One last chance: use the default replacement.  */
 198   if (_NL_CURRENT_WORD (LC_CTYPE, _NL_CTYPE_TRANSLIT_DEFAULT_MISSING_LEN) != 0)
 199     {
 200       const uint32_t *default_missing = (const uint32_t *)
 201         _NL_CURRENT (LC_CTYPE, _NL_CTYPE_TRANSLIT_DEFAULT_MISSING);
 202       const unsigned char *toinptr = (const unsigned char *) default_missing;
 203       uint32_t len = _NL_CURRENT_WORD (LC_CTYPE,
 204                                        _NL_CTYPE_TRANSLIT_DEFAULT_MISSING_LEN);
 205       unsigned char *outptr;
 206       int res;
 207
 208       /* Test whether there is enough input.  */
 209       if (winbuf + 1 > winbufend)
 210         return (winbuf == winbufend
 211                 ? __GCONV_EMPTY_INPUT : __GCONV_INCOMPLETE_INPUT);
 212
 213       outptr = *outbufstart;
 214       res = DL_CALL_FCT (fct,
 215                          (step, step_data, &toinptr,
 216                           (const unsigned char *) (default_missing + len),
 217                           &outptr, NULL, 0, 0));
 218
 219       if (res != __GCONV_ILLEGAL_INPUT)
 220         {
 221           /* If the conversion succeeds we have to increment the
 222              input buffer.  */
 223           if (res == __GCONV_EMPTY_INPUT)
 224             {
 225               /* This worked but is not reversible.  */
 226               ++*irreversible;
 227               *inbufp += 4;
 228               res = __GCONV_OK;
 229             }
 230           *outbufstart = outptr;
 231
 232           return res;
 233         }
 234     }
 235
 236   /* Haven't found a match.  */
 237   return __GCONV_ILLEGAL_INPUT;
 238 }
 239 libc_hidden_def (__gconv_transliterate)