iconvdata/euc-jp.c

   1 /* Mapping tables for EUC-JP 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 <gconv.h>
  24 #include <jis0201.h>
  25 #include <jis0208.h>
  26 #include <jis0212.h>
  27
  28 /* Definitions used in the body of the `gconv' function.  */
  29 #define CHARSET_NAME            "EUC-JP//"
  30 #define FROM_LOOP               from_euc_jp
  31 #define TO_LOOP                 to_euc_jp
  32 #define DEFINE_INIT             1
  33 #define DEFINE_FINI             1
  34 #define MIN_NEEDED_FROM         1
  35 #define MAX_NEEDED_FROM         3
  36 #define MIN_NEEDED_TO           4
  37
  38
  39 /* First define the conversion function from EUC-JP to UCS4.  */
  40 #define MIN_NEEDED_INPUT        MIN_NEEDED_FROM
  41 #define MAX_NEEDED_INPUT        MAX_NEEDED_FROM
  42 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_TO
  43 #define LOOPFCT                 FROM_LOOP
  44 #define BODY \
  45   {                                                                           \
  46     uint32_t ch = *inptr;                                                     \
  47                                                                               \
  48     if (ch < 0x8e || (ch >= 0x90 && ch <= 0x9f))                              \
  49       ++inptr;                                                                \
  50     else if (ch == 0xff)                                                      \
  51       {                                                                       \
  52         /* This is illegal.  */                                               \
  53         if (! ignore_errors_p ())                                             \
  54           {                                                                   \
  55             result = __GCONV_ILLEGAL_INPUT;                                   \
  56             break;                                                            \
  57           }                                                                   \
  58                                                                               \
  59         ++inptr;                                                              \
  60         ++*irreversible;                                                      \
  61         continue;                                                             \
  62       }                                                                       \
  63     else                                                                      \
  64       {                                                                       \
  65         /* Two or more byte character.  First test whether the next           \
  66            character is also available.  */                                   \
  67         int ch2;                                                              \
  68                                                                               \
  69         if (__builtin_expect (inptr + 1 >= inend, 0))                         \
  70           {                                                                   \
  71             /* The second character is not available.  Store the              \
  72                intermediate result.  */                                       \
  73             result = __GCONV_INCOMPLETE_INPUT;                                \
  74             break;                                                            \
  75           }                                                                   \
  76                                                                               \
  77         ch2 = inptr[1];                                                       \
  78                                                                               \
  79         /* All second bytes of a multibyte character must be >= 0xa1. */      \
  80         if (__builtin_expect (ch2 < 0xa1, 0))                                 \
  81           {                                                                   \
  82             /* This is an illegal character.  */                              \
  83             if (! ignore_errors_p ())                                         \
  84               {                                                               \
  85                 result = __GCONV_ILLEGAL_INPUT;                               \
  86                 break;                                                        \
  87               }                                                               \
  88                                                                               \
  89             ++inptr;                                                          \
  90             ++*irreversible;                                                  \
  91             continue;                                                         \
  92           }                                                                   \
  93                                                                               \
  94         if (ch == 0x8e)                                                       \
  95           {                                                                   \
  96             /* This is code set 2: half-width katakana.  */                   \
  97             ch = jisx0201_to_ucs4 (ch2);                                      \
  98             if (__builtin_expect (ch, 0) == __UNKNOWN_10646_CHAR)             \
  99               {                                                               \
 100                 /* Illegal character.  */                                     \
 101                 if (! ignore_errors_p ())                                     \
 102                   {                                                           \
 103                     /* This is an illegal character.  */                      \
 104                     result = __GCONV_ILLEGAL_INPUT;                           \
 105                     break;                                                    \
 106                   }                                                           \
 107               }                                                               \
 108                                                                               \
 109             inptr += 2;                                                       \
 110           }                                                                   \
 111         else                                                                  \
 112           {                                                                   \
 113             const unsigned char *endp;                                        \
 114                                                                               \
 115             if (ch == 0x8f)                                                   \
 116               {                                                               \
 117                 /* This is code set 3: JIS X 0212-1990.  */                   \
 118                 endp = inptr + 1;                                             \
 119                                                                               \
 120                 ch = jisx0212_to_ucs4 (&endp, inend - endp, 0x80);            \
 121               }                                                               \
 122             else                                                              \
 123               {                                                               \
 124                 /* This is code set 1: JIS X 0208.  */                        \
 125                 endp = inptr;                                                 \
 126                                                                               \
 127                 ch = jisx0208_to_ucs4 (&endp, inend - inptr, 0x80);           \
 128               }                                                               \
 129                                                                               \
 130             if (__builtin_expect (ch, 1) == 0)                                \
 131               {                                                               \
 132                 /* Not enough input available.  */                            \
 133                 result = __GCONV_INCOMPLETE_INPUT;                            \
 134                 break;                                                        \
 135               }                                                               \
 136             if (__builtin_expect (ch, 0) == __UNKNOWN_10646_CHAR)             \
 137               {                                                               \
 138                 /* Illegal character.  */                                     \
 139                 if (! ignore_errors_p ())                                     \
 140                   {                                                           \
 141                     /* This is an illegal character.  */                      \
 142                     result = __GCONV_ILLEGAL_INPUT;                           \
 143                     break;                                                    \
 144                   }                                                           \
 145                                                                               \
 146                 inptr += 2;                                                   \
 147                 ++*irreversible;                                              \
 148                 continue;                                                     \
 149               }                                                               \
 150             inptr = endp;                                                     \
 151           }                                                                   \
 152       }                                                                       \
 153                                                                               \
 154     put32 (outptr, ch);                                                       \
 155     outptr += 4;                                                              \
 156   }
 157 #define LOOP_NEED_FLAGS
 158 #include <iconv/loop.c>
 159
 160
 161 /* Next, define the other direction.  */
 162 #define MIN_NEEDED_INPUT        MIN_NEEDED_TO
 163 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_FROM
 164 #define MAX_NEEDED_OUTPUT       MAX_NEEDED_FROM
 165 #define LOOPFCT                 TO_LOOP
 166 #define BODY \
 167   {                                                                           \
 168     uint32_t ch = get32 (inptr);                                              \
 169                                                                               \
 170     if (ch < 0x8e || (ch >= 0x90 && ch <= 0x9f))                              \
 171       /* It's plain ASCII or C1.  */                                          \
 172       *outptr++ = ch;                                                         \
 173     else if (ch == 0xa5)                                                      \
 174       /* YEN sign => backslash  */                                            \
 175       *outptr++ = 0x5c;                                                       \
 176     else if (ch == 0x203e)                                                    \
 177       /* overscore => asciitilde */                                           \
 178       *outptr++ = 0x7e;                                                       \
 179     else                                                                      \
 180       {                                                                       \
 181         /* Try the JIS character sets.  */                                    \
 182         size_t found;                                                         \
 183                                                                               \
 184         /* See whether we have room for at least two characters.  */          \
 185         if (__builtin_expect (outptr + 1 >= outend, 0))                       \
 186           {                                                                   \
 187             result = __GCONV_FULL_OUTPUT;                                     \
 188             break;                                                            \
 189           }                                                                   \
 190                                                                               \
 191         found = ucs4_to_jisx0201 (ch, outptr + 1);                            \
 192         if (found != __UNKNOWN_10646_CHAR)                                    \
 193           {                                                                   \
 194             /* Yes, it's a JIS 0201 character.  Store the shift byte.  */     \
 195             *outptr = 0x8e;                                                   \
 196             outptr += 2;                                                      \
 197           }                                                                   \
 198         else                                                                  \
 199           {                                                                   \
 200             /* No JIS 0201 character.  */                                     \
 201             found = ucs4_to_jisx0208 (ch, outptr, 2);                         \
 202             /* Please note that we always have enough room for the output. */ \
 203             if (found != __UNKNOWN_10646_CHAR)                                \
 204               {                                                               \
 205                 /* It's a JIS 0208 character, adjust it for EUC-JP.  */       \
 206                 *outptr++ += 0x80;                                            \
 207                 *outptr++ += 0x80;                                            \
 208               }                                                               \
 209             else                                                              \
 210               {                                                               \
 211                 /* No JIS 0208 character.  */                                 \
 212                 found = ucs4_to_jisx0212 (ch, outptr + 1,                     \
 213                                           outend - outptr - 1);               \
 214                                                                               \
 215                 if (__builtin_expect (found, 1) == 0)                         \
 216                   {                                                           \
 217                     /* We ran out of space.  */                               \
 218                     result = __GCONV_FULL_OUTPUT;                             \
 219                     break;                                                    \
 220                   }                                                           \
 221                 else if (__builtin_expect (found, 0) != __UNKNOWN_10646_CHAR) \
 222                   {                                                           \
 223                     /* It's a JIS 0212 character, adjust it for EUC-JP.  */   \
 224                     *outptr++ = 0x8f;                                         \
 225                     *outptr++ += 0x80;                                        \
 226                     *outptr++ += 0x80;                                        \
 227                   }                                                           \
 228                 else                                                          \
 229                   {                                                           \
 230                     UNICODE_TAG_HANDLER (ch, 4);                              \
 231                                                                               \
 232                     /* Illegal character.  */                                 \
 233                     STANDARD_ERR_HANDLER (4);                                 \
 234                   }                                                           \
 235               }                                                               \
 236           }                                                                   \
 237       }                                                                       \
 238                                                                               \
 239     inptr += 4;                                                               \
 240   }
 241 #define LOOP_NEED_FLAGS
 242 #include <iconv/loop.c>
 243
 244
 245 /* Now define the toplevel functions.  */
 246 #include <iconv/skeleton.c>