iconvdata/euc-jp.c

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