iconvdata/euc-tw.c

   1 /* Mapping tables for EUC-TW handling.
   2    Copyright (C) 1998, 1999, 2000, 2001 Free Software Foundation, Inc.
   3    This file is part of the GNU C Library.
   4    Contributed by Ulrich Drepper <drepper@cygnus.com>, 1998.
   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, write to the Free
  18    Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
  19    02111-1307 USA.  */
  20
  21 #include <dlfcn.h>
  22 #include <stdint.h>
  23 #include <cns11643l1.h>
  24 #include <cns11643.h>
  25
  26 /* Definitions used in the body of the `gconv' function.  */
  27 #define CHARSET_NAME            "EUC-TW//"
  28 #define FROM_LOOP               from_euc_tw
  29 #define TO_LOOP                 to_euc_tw
  30 #define DEFINE_INIT             1
  31 #define DEFINE_FINI             1
  32 #define MIN_NEEDED_FROM         1
  33 #define MAX_NEEDED_FROM         4
  34 #define MIN_NEEDED_TO           4
  35
  36
  37 /* First define the conversion function from EUC-TW to UCS4.  */
  38 #define MIN_NEEDED_INPUT        MIN_NEEDED_FROM
  39 #define MAX_NEEDED_INPUT        MAX_NEEDED_FROM
  40 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_TO
  41 #define LOOPFCT                 FROM_LOOP
  42 #define BODY \
  43   {                                                                           \
  44     uint32_t ch = *inptr;                                                     \
  45                                                                               \
  46     if (ch <= 0x7f)                                                           \
  47       /* Plain ASCII.  */                                                     \
  48       ++inptr;                                                                \
  49     else if ((ch <= 0xa0 || ch > 0xfe) && ch != 0x8e)                         \
  50       {                                                                       \
  51         /* This is illegal.  */                                               \
  52         if (! ignore_errors_p ())                                             \
  53           {                                                                   \
  54             result = __GCONV_ILLEGAL_INPUT;                                   \
  55             break;                                                            \
  56           }                                                                   \
  57                                                                               \
  58         ++inptr;                                                              \
  59         ++*irreversible;                                                      \
  60         continue;                                                             \
  61       }                                                                       \
  62     else                                                                      \
  63       {                                                                       \
  64         /* Two or more byte character.  First test whether the next byte      \
  65            is also available.  */                                             \
  66         uint32_t ch2;                                                         \
  67                                                                               \
  68         if (inptr + 1 >= inend)                                               \
  69           {                                                                   \
  70             /* The second byte is not available.  Store the intermediate      \
  71                result.  */                                                    \
  72             result = __GCONV_INCOMPLETE_INPUT;                                \
  73             break;                                                            \
  74           }                                                                   \
  75                                                                               \
  76         ch2 = *(inptr + 1);                                                   \
  77                                                                               \
  78         /* All second bytes of a multibyte character must be >= 0xa1. */      \
  79         if (ch2 < 0xa1 || ch2 == 0xff)                                        \
  80           {                                                                   \
  81             /* This is illegal.  */                                           \
  82             if (! ignore_errors_p ())                                         \
  83               {                                                               \
  84                 result = __GCONV_ILLEGAL_INPUT;                               \
  85                 break;                                                        \
  86               }                                                               \
  87                                                                               \
  88             ++inptr;                                                          \
  89             ++*irreversible;                                                  \
  90             continue;                                                         \
  91           }                                                                   \
  92                                                                               \
  93         if (ch == 0x8e)                                                       \
  94           {                                                                   \
  95             /* This is code set 2: CNS 11643, planes 1 to 16.  */             \
  96             const char *endp = inptr + 1;                                     \
  97                                                                               \
  98             ch = cns11643_to_ucs4 (&endp, inend - inptr - 1, 0x80);           \
  99                                                                               \
 100             if (ch == 0)                                                      \
 101               {                                                               \
 102                 /* The third or fourth byte is not available.  Store          \
 103                    the intermediate result.  */                               \
 104                 result = __GCONV_INCOMPLETE_INPUT;                            \
 105                 break;                                                        \
 106               }                                                               \
 107                                                                               \
 108             if (ch == __UNKNOWN_10646_CHAR)                                   \
 109               {                                                               \
 110                 /* Illegal input.  */                                         \
 111                 if (! ignore_errors_p ())                                     \
 112                   {                                                           \
 113                     /* This is an illegal character.  */                      \
 114                     result = __GCONV_ILLEGAL_INPUT;                           \
 115                     break;                                                    \
 116                   }                                                           \
 117                                                                               \
 118                 ++inptr;                                                      \
 119                 ++*irreversible;                                              \
 120                 continue;                                                     \
 121               }                                                               \
 122                                                                               \
 123             inptr += 4;                                                       \
 124           }                                                                   \
 125         else                                                                  \
 126           {                                                                   \
 127             /* This is code set 1: CNS 11643, plane 1.  */                    \
 128             const unsigned char *endp = inptr;                                \
 129                                                                               \
 130             ch = cns11643l1_to_ucs4 (&endp, inend - inptr, 0x80);             \
 131             /* Please note that we need not test for the missing input        \
 132                characters here anymore.  */                                   \
 133             if (ch == __UNKNOWN_10646_CHAR)                                   \
 134               {                                                               \
 135                 /* Illegal input.  */                                         \
 136                 if (! ignore_errors_p ())                                     \
 137                   {                                                           \
 138                     /* This is an illegal character.  */                      \
 139                     result = __GCONV_ILLEGAL_INPUT;                           \
 140                     break;                                                    \
 141                   }                                                           \
 142                                                                               \
 143                 inptr += 2;                                                   \
 144                 ++*irreversible;                                              \
 145                 continue;                                                     \
 146               }                                                               \
 147                                                                               \
 148             inptr += 2;                                                       \
 149           }                                                                   \
 150       }                                                                       \
 151                                                                               \
 152     put32 (outptr, ch);                                                       \
 153     outptr += 4;                                                              \
 154   }
 155 #define LOOP_NEED_FLAGS
 156 #include <iconv/loop.c>
 157
 158
 159 /* Next, define the other direction.  */
 160 #define MIN_NEEDED_INPUT        MIN_NEEDED_TO
 161 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_FROM
 162 #define MAX_NEEDED_OUTPUT       MAX_NEEDED_FROM
 163 #define LOOPFCT                 TO_LOOP
 164 #define BODY \
 165   {                                                                           \
 166     uint32_t ch = get32 (inptr);                                              \
 167                                                                               \
 168     if (ch <= 0x7f)                                                           \
 169       /* It's plain ASCII.  */                                                \
 170       *outptr++ = ch;                                                         \
 171     else                                                                      \
 172       {                                                                       \
 173         /* Try the CNS 11643 planes.  */                                      \
 174         size_t found;                                                         \
 175                                                                               \
 176         found = ucs4_to_cns11643l1 (ch, outptr, outend - outptr);             \
 177         if (__builtin_expect (found, 1) == 0)                                 \
 178           {                                                                   \
 179             /* We ran out of space.  */                                       \
 180             result = __GCONV_INCOMPLETE_INPUT;                                \
 181             break;                                                            \
 182           }                                                                   \
 183         if (__builtin_expect (found, 1) != __UNKNOWN_10646_CHAR)              \
 184           {                                                                   \
 185             /* It's a CNS 11643, plane 1 character, adjust it for EUC-TW.  */ \
 186             *outptr++ += 0x80;                                                \
 187             *outptr++ += 0x80;                                                \
 188           }                                                                   \
 189         else                                                                  \
 190           {                                                                   \
 191             /* No CNS 11643, plane 1 character.  */                           \
 192                                                                               \
 193             found = ucs4_to_cns11643 (ch, outptr + 1, outend - outptr - 1);   \
 194             if (__builtin_expect (found, 1) == 0)                             \
 195               {                                                               \
 196                 /* We ran out of space.  */                                   \
 197                 result = __GCONV_INCOMPLETE_INPUT;                            \
 198                 break;                                                        \
 199               }                                                               \
 200             if (__builtin_expect (found, 0) == __UNKNOWN_10646_CHAR)          \
 201               {                                                               \
 202                 UNICODE_TAG_HANDLER (ch, 4);                                  \
 203                                                                               \
 204                 /* Illegal character.  */                                     \
 205                 STANDARD_ERR_HANDLER (4);                                     \
 206               }                                                               \
 207                                                                               \
 208             /* It's a CNS 11643 character, adjust it for EUC-TW.  */          \
 209             *outptr++ = '\x8e';                                               \
 210             *outptr++ += 0xa0;                                                \
 211             *outptr++ += 0x80;                                                \
 212             *outptr++ += 0x80;                                                \
 213           }                                                                   \
 214       }                                                                       \
 215                                                                               \
 216     inptr += 4;                                                               \
 217   }
 218 #define LOOP_NEED_FLAGS
 219 #include <iconv/loop.c>
 220
 221
 222 /* Now define the toplevel functions.  */
 223 #include <iconv/skeleton.c>