iconvdata/euc-tw.c

   1 /* Mapping tables for EUC-TW handling.
   2    Copyright (C) 1998-2017 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, see
  18    <http://www.gnu.org/licenses/>.  */
  19
  20 #include <dlfcn.h>
  21 #include <stdint.h>
  22 #include <cns11643l1.h>
  23 #include <cns11643.h>
  24
  25 /* Definitions used in the body of the `gconv' function.  */
  26 #define CHARSET_NAME            "EUC-TW//"
  27 #define FROM_LOOP               from_euc_tw
  28 #define TO_LOOP                 to_euc_tw
  29 #define DEFINE_INIT             1
  30 #define DEFINE_FINI             1
  31 #define MIN_NEEDED_FROM         1
  32 #define MAX_NEEDED_FROM         4
  33 #define MIN_NEEDED_TO           4
  34 #define ONE_DIRECTION           0
  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         STANDARD_FROM_LOOP_ERR_HANDLER (1);                                   \
  53       }                                                                       \
  54     else                                                                      \
  55       {                                                                       \
  56         /* Two or more byte character.  First test whether the next byte      \
  57            is also available.  */                                             \
  58         uint32_t ch2;                                                         \
  59                                                                               \
  60         if (inptr + 1 >= inend)                                               \
  61           {                                                                   \
  62             /* The second byte is not available.  Store the intermediate      \
  63                result.  */                                                    \
  64             result = __GCONV_INCOMPLETE_INPUT;                                \
  65             break;                                                            \
  66           }                                                                   \
  67                                                                               \
  68         ch2 = *(inptr + 1);                                                   \
  69                                                                               \
  70         /* All second bytes of a multibyte character must be >= 0xa1. */      \
  71         if (ch2 < 0xa1 || ch2 == 0xff)                                        \
  72           STANDARD_FROM_LOOP_ERR_HANDLER (1);                                 \
  73                                                                               \
  74         if (ch == 0x8e)                                                       \
  75           {                                                                   \
  76             /* This is code set 2: CNS 11643, planes 1 to 16.  */             \
  77             const unsigned char *endp = inptr + 1;                            \
  78                                                                               \
  79             ch = cns11643_to_ucs4 (&endp, inend - inptr - 1, 0x80);           \
  80                                                                               \
  81             if (ch == 0)                                                      \
  82               {                                                               \
  83                 /* The third or fourth byte is not available.  Store          \
  84                    the intermediate result.  */                               \
  85                 result = __GCONV_INCOMPLETE_INPUT;                            \
  86                 break;                                                        \
  87               }                                                               \
  88                                                                               \
  89             if (ch == __UNKNOWN_10646_CHAR)                                   \
  90               /* Illegal input.  */                                           \
  91               STANDARD_FROM_LOOP_ERR_HANDLER (1);                             \
  92                                                                               \
  93             inptr += 4;                                                       \
  94           }                                                                   \
  95         else                                                                  \
  96           {                                                                   \
  97             /* This is code set 1: CNS 11643, plane 1.  */                    \
  98             const unsigned char *endp = inptr;                                \
  99                                                                               \
 100             ch = cns11643l1_to_ucs4 (&endp, inend - inptr, 0x80);             \
 101             /* Please note that we need not test for the missing input        \
 102                characters here anymore.  */                                   \
 103             if (ch == __UNKNOWN_10646_CHAR)                                   \
 104               /* Illegal input.  */                                           \
 105               STANDARD_FROM_LOOP_ERR_HANDLER (2);                             \
 106                                                                               \
 107             inptr += 2;                                                       \
 108           }                                                                   \
 109       }                                                                       \
 110                                                                               \
 111     put32 (outptr, ch);                                                       \
 112     outptr += 4;                                                              \
 113   }
 114 #define LOOP_NEED_FLAGS
 115 #define ONEBYTE_BODY \
 116   {                                                                           \
 117     if (c < 0x80)                                                             \
 118       return c;                                                               \
 119     else                                                                      \
 120       return WEOF;                                                            \
 121   }
 122 #include <iconv/loop.c>
 123
 124
 125 /* Next, define the other direction.  */
 126 #define MIN_NEEDED_INPUT        MIN_NEEDED_TO
 127 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_FROM
 128 #define MAX_NEEDED_OUTPUT       MAX_NEEDED_FROM
 129 #define LOOPFCT                 TO_LOOP
 130 #define BODY \
 131   {                                                                           \
 132     uint32_t ch = get32 (inptr);                                              \
 133                                                                               \
 134     if (ch <= 0x7f)                                                           \
 135       /* It's plain ASCII.  */                                                \
 136       *outptr++ = ch;                                                         \
 137     else                                                                      \
 138       {                                                                       \
 139         /* Try the CNS 11643 planes.  */                                      \
 140         size_t found;                                                         \
 141                                                                               \
 142         found = ucs4_to_cns11643l1 (ch, outptr, outend - outptr);             \
 143         if (__builtin_expect (found, 1) == 0)                                 \
 144           {                                                                   \
 145             /* We ran out of space.  */                                       \
 146             result = __GCONV_FULL_OUTPUT;                                     \
 147             break;                                                            \
 148           }                                                                   \
 149         if (__builtin_expect (found, 1) != __UNKNOWN_10646_CHAR)              \
 150           {                                                                   \
 151             /* It's a CNS 11643, plane 1 character, adjust it for EUC-TW.  */ \
 152             *outptr++ += 0x80;                                                \
 153             *outptr++ += 0x80;                                                \
 154           }                                                                   \
 155         else                                                                  \
 156           {                                                                   \
 157             /* No CNS 11643, plane 1 character.  */                           \
 158                                                                               \
 159             found = ucs4_to_cns11643 (ch, outptr + 1, outend - outptr - 1);   \
 160             if (__builtin_expect (found, 1) == 0)                             \
 161               {                                                               \
 162                 /* We ran out of space.  */                                   \
 163                 result = __GCONV_FULL_OUTPUT;                                 \
 164                 break;                                                        \
 165               }                                                               \
 166             if (__builtin_expect (found, 0) == __UNKNOWN_10646_CHAR)          \
 167               {                                                               \
 168                 UNICODE_TAG_HANDLER (ch, 4);                                  \
 169                                                                               \
 170                 /* Illegal character.  */                                     \
 171                 STANDARD_TO_LOOP_ERR_HANDLER (4);                             \
 172               }                                                               \
 173                                                                               \
 174             /* It's a CNS 11643 character, adjust it for EUC-TW.  */          \
 175             *outptr++ = '\x8e';                                               \
 176             *outptr++ += 0xa0;                                                \
 177             *outptr++ += 0x80;                                                \
 178             *outptr++ += 0x80;                                                \
 179           }                                                                   \
 180       }                                                                       \
 181                                                                               \
 182     inptr += 4;                                                               \
 183   }
 184 #define LOOP_NEED_FLAGS
 185 #include <iconv/loop.c>
 186
 187
 188 /* Now define the toplevel functions.  */
 189 #include <iconv/skeleton.c>