iconvdata/ibm932.c

   1 /* Conversion from and to IBM932.
   2    Copyright (C) 2000, 2001 Free Software Foundation, Inc.
   3    This file is part of the GNU C Library.
   4    Contributed by Masahide Washizawa <washi@jp.ibm.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, write to the Free
  18    Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
  19    02111-1307 USA.  */
  20
  21 #include "ibm932.h"
  22
  23 #ifndef TRUE
  24 #define TRUE  1
  25 #define FALSE 0
  26 #endif
  27
  28 #define FROM    0
  29 #define TO      1
  30
  31 /* Definitions used in the body of the `gconv' function.  */
  32 #define CHARSET_NAME    "IBM932//"
  33 #define FROM_LOOP       from_ibm932
  34 #define TO_LOOP         to_ibm932
  35
  36 /* Definitions of initialization and destructor function.  */
  37 #define DEFINE_INIT     1
  38 #define DEFINE_FINI     1
  39
  40 #define MIN_NEEDED_FROM 1
  41 #define MAX_NEEDED_FROM 2
  42 #define MIN_NEEDED_TO   4
  43
  44 /* First, define the conversion function from IBM-932 to UCS4.  */
  45 #define MIN_NEEDED_INPUT        MIN_NEEDED_FROM
  46 #define MAX_NEEDED_INPUT        MAX_NEEDED_FROM
  47 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_TO
  48 #define LOOPFCT                 FROM_LOOP
  49 #define BODY \
  50   {                                                                           \
  51     const struct gap *rp1 = __ibm932sb_to_ucs4_idx;                           \
  52     const struct gap *rp2 = __ibm932db_to_ucs4_idx;                           \
  53     uint32_t ch = *inptr;                                                     \
  54     uint32_t res;                                                             \
  55                                                                               \
  56     if (__builtin_expect (ch, 0) >= 0xffff)                                   \
  57       {                                                                       \
  58         rp1 = NULL;                                                           \
  59         rp2 = NULL;                                                           \
  60       }                                                                       \
  61     else if (__builtin_expect (ch, 0) == 0x80                                 \
  62              || __builtin_expect (ch, 0) == 0xa0                              \
  63              || __builtin_expect (ch, 0) == 0xfd                              \
  64              || __builtin_expect (ch, 0) == 0xfe                              \
  65              || __builtin_expect (ch, 0) == 0xff)                             \
  66       {                                                                       \
  67         /* This is an illegal character.  */                                  \
  68         if (! ignore_errors_p ())                                             \
  69           {                                                                   \
  70             result = __GCONV_ILLEGAL_INPUT;                                   \
  71             break;                                                            \
  72           }                                                                   \
  73       }                                                                       \
  74     else                                                                      \
  75       {                                                                       \
  76         while (ch > rp1->end)                                                 \
  77           ++rp1;                                                              \
  78       }                                                                       \
  79                                                                               \
  80     /* Use the IBM932 table for single byte.  */                              \
  81     if (__builtin_expect (rp1 == NULL, 0)                                     \
  82         || __builtin_expect (ch < rp1->start, 0)                              \
  83         || (res = __ibm932sb_to_ucs4[ch + rp1->idx],                          \
  84         __builtin_expect (res, '\1') == 0 && ch != 0))                        \
  85       {                                                                       \
  86                                                                               \
  87         /* Use the IBM932 table for double byte.  */                          \
  88         if (__builtin_expect (inptr + 1 >= inend, 0))                         \
  89           {                                                                   \
  90             /* The second character is not available.                         \
  91                Store the intermediate result.  */                             \
  92             result = __GCONV_INCOMPLETE_INPUT;                                \
  93             break;                                                            \
  94           }                                                                   \
  95                                                                               \
  96         ch = (ch * 0x100) + inptr[1];                                         \
  97         while (ch > rp2->end)                                                 \
  98           ++rp2;                                                              \
  99                                                                               \
 100         if (__builtin_expect (rp2 == NULL, 0)                                 \
 101             || __builtin_expect (ch < rp2->start, 0)                          \
 102             || (res = __ibm932db_to_ucs4[ch + rp2->idx],                      \
 103             __builtin_expect (res, '\1') == 0 && ch !=0))                     \
 104           {                                                                   \
 105             /* This is an illegal character.  */                              \
 106             if (! ignore_errors_p ())                                         \
 107               {                                                               \
 108                 result = __GCONV_ILLEGAL_INPUT;                               \
 109                 break;                                                        \
 110               }                                                               \
 111               ++*irreversible;                                                \
 112               inptr += 2;                                                     \
 113               continue;                                                       \
 114           }                                                                   \
 115         else                                                                  \
 116           {                                                                   \
 117             put32 (outptr, res);                                              \
 118             outptr += 4;                                                      \
 119             inptr += 2;                                                       \
 120           }                                                                   \
 121       }                                                                       \
 122     else                                                                      \
 123       {                                                                       \
 124         if (res == 0x1c)                                                      \
 125           res = 0x1a;                                                         \
 126         else if (res == 0x7f)                                                 \
 127           res = 0x1c;                                                         \
 128         else if (res == 0xa5)                                                 \
 129           res = 0x5c;                                                         \
 130         else if (res == 0x203e)                                               \
 131           res = 0x7e;                                                         \
 132         else if (res == 0x1a)                                                 \
 133           res = 0x7f;                                                         \
 134         put32 (outptr, res);                                                  \
 135         outptr += 4;                                                          \
 136         inptr++;                                                              \
 137       }                                                                       \
 138   }
 139 #define LOOP_NEED_FLAGS
 140 #include <iconv/loop.c>
 141
 142 /* Next, define the other direction.  */
 143 #define MIN_NEEDED_INPUT        MIN_NEEDED_TO
 144 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_FROM
 145 #define MAX_NEEDED_OUTPUT       MAX_NEEDED_FROM
 146 #define LOOPFCT                 TO_LOOP
 147 #define BODY \
 148   {                                                                           \
 149     const struct gap *rp = __ucs4_to_ibm932sb_idx;                            \
 150     unsigned char sc;                                                         \
 151     uint32_t ch = get32 (inptr);                                              \
 152     uint16_t found = TRUE;                                                    \
 153     uint32_t i;                                                               \
 154     uint32_t low;                                                             \
 155     uint32_t high;                                                            \
 156     uint16_t pccode;                                                          \
 157                                                                               \
 158     if (__builtin_expect (ch, 0) >= 0xffff)                                   \
 159       {                                                                       \
 160         UNICODE_TAG_HANDLER (ch, 4);                                          \
 161         rp = NULL;                                                            \
 162       }                                                                       \
 163     else                                                                      \
 164       while (ch > rp->end)                                                    \
 165         ++rp;                                                                 \
 166                                                                               \
 167     /* Use the UCS4 table for single byte.  */                                \
 168     if (__builtin_expect (rp == NULL, 0)                                      \
 169         || __builtin_expect (ch < rp->start, 0)                               \
 170         || (sc = __ucs4_to_ibm932sb[ch + rp->idx],                            \
 171         __builtin_expect (sc, '\1') == '\0' && ch != L'\0'))                  \
 172       {                                                                       \
 173                                                                               \
 174         /* Use the UCS4 table for double byte.  */                            \
 175         found = FALSE;                                                        \
 176         low = 0;                                                              \
 177         high = (sizeof (__ucs4_to_ibm932db) >> 1)                             \
 178                 / sizeof (__ucs4_to_ibm932db[0][FROM]);                       \
 179         pccode = ch;                                                          \
 180         while (low <= high)                                                   \
 181           {                                                                   \
 182             i = (low + high) >> 1;                                            \
 183             if (pccode < __ucs4_to_ibm932db[i][FROM])                         \
 184               high = i - 1;                                                   \
 185             else if (pccode > __ucs4_to_ibm932db[i][FROM])                    \
 186               low = i + 1;                                                    \
 187             else                                                              \
 188               {                                                               \
 189                 pccode = __ucs4_to_ibm932db[i][TO];                           \
 190                 found = TRUE;                                                 \
 191                 break;                                                        \
 192               }                                                               \
 193           }                                                                   \
 194         if (found)                                                            \
 195           {                                                                   \
 196             if (__builtin_expect (outptr + 2 > outend, 0))                    \
 197               {                                                               \
 198                 result = __GCONV_FULL_OUTPUT;                                 \
 199                 break;                                                        \
 200               }                                                               \
 201             *outptr++ = pccode >> 8 & 0xff;                                   \
 202             *outptr++ = pccode & 0xff;                                        \
 203           }                                                                   \
 204         else                                                                  \
 205           {                                                                   \
 206             /* This is an illegal character.  */                              \
 207             if (! ignore_errors_p ())                                         \
 208               {                                                               \
 209                 result = __GCONV_ILLEGAL_INPUT;                               \
 210                 break;                                                        \
 211               }                                                               \
 212             ++*irreversible;                                                  \
 213           }                                                                   \
 214       }                                                                       \
 215     else                                                                      \
 216       {                                                                       \
 217         if (__builtin_expect (outptr + 1 > outend, 0))                        \
 218           {                                                                   \
 219             result = __GCONV_FULL_OUTPUT;                                     \
 220             break;                                                            \
 221           }                                                                   \
 222         if (ch == 0x5c)                                                       \
 223           *outptr++ = 0x5c;                                                   \
 224         else if (ch == 0x7e)                                                  \
 225           *outptr++ = 0x7e;                                                   \
 226         else                                                                  \
 227           *outptr++ = sc;                                                     \
 228       }                                                                       \
 229                                                                               \
 230     /* Now that we wrote the output increment the input pointer.  */          \
 231     inptr += 4;                                                               \
 232   }
 233 #define LOOP_NEED_FLAGS
 234 #include <iconv/loop.c>
 235
 236 /* Now define the toplevel functions.  */
 237 #include <iconv/skeleton.c>