iconvdata/ibm943.c

   1 /* Conversion from and to IBM943.
   2    Copyright (C) 2000-2002, 2004 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 <dlfcn.h>
  22 #include <stdint.h>
  23 #include <stdbool.h>
  24 #include "ibm943.h"
  25
  26 #define FROM    0
  27 #define TO      1
  28
  29 /* Definitions used in the body of the `gconv' function.  */
  30 #define CHARSET_NAME    "IBM943//"
  31 #define FROM_LOOP       from_ibm943
  32 #define TO_LOOP         to_ibm943
  33
  34 /* Definitions of initialization and destructor function.  */
  35 #define DEFINE_INIT     1
  36 #define DEFINE_FINI     1
  37
  38 #define MIN_NEEDED_FROM 1
  39 #define MAX_NEEDED_FROM 2
  40 #define MIN_NEEDED_TO   4
  41
  42 /* First, define the conversion function from IBM-943 to UCS4.  */
  43 #define MIN_NEEDED_INPUT        MIN_NEEDED_FROM
  44 #define MAX_NEEDED_INPUT        MAX_NEEDED_FROM
  45 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_TO
  46 #define LOOPFCT                 FROM_LOOP
  47 #define BODY \
  48   {                                                                           \
  49     const struct gap *rp2 = __ibm943db_to_ucs4_idx;                           \
  50     uint32_t ch = *inptr;                                                     \
  51     uint32_t res;                                                             \
  52                                                                               \
  53     if (__builtin_expect (ch == 0x80, 0)                                      \
  54         || __builtin_expect (ch == 0xa0, 0)                                   \
  55         || __builtin_expect (ch == 0xfd, 0)                                   \
  56         || __builtin_expect (ch == 0xfe, 0)                                   \
  57         || __builtin_expect (ch == 0xff, 0))                                  \
  58       {                                                                       \
  59         /* This is an illegal character.  */                                  \
  60         STANDARD_FROM_LOOP_ERR_HANDLER (1);                                   \
  61       }                                                                       \
  62                                                                               \
  63     /* Use the IBM943 table for single byte.  */                              \
  64     if (__builtin_expect (ch > 0xdf, 0)                                       \
  65         || (res = __ibm943sb_to_ucs4[ch],                                     \
  66             __builtin_expect (res == 0, 0) && ch != 0))                       \
  67       {                                                                       \
  68         /* Use the IBM943 table for double byte.  */                          \
  69         if (__builtin_expect (inptr + 1 >= inend, 0))                         \
  70           {                                                                   \
  71             /* The second character is not available.                         \
  72                Store the intermediate result.  */                             \
  73             result = __GCONV_INCOMPLETE_INPUT;                                \
  74             break;                                                            \
  75           }                                                                   \
  76                                                                               \
  77         ch = (ch * 0x100) + inptr[1];                                         \
  78         while (ch > rp2->end)                                                 \
  79           ++rp2;                                                              \
  80                                                                               \
  81         if (__builtin_expect (rp2 == NULL, 0)                                 \
  82             || __builtin_expect (ch < rp2->start, 0)                          \
  83             || (res = __ibm943db_to_ucs4[ch + rp2->idx],                      \
  84             __builtin_expect (res, '\1') == 0 && ch !=0))                     \
  85           {                                                                   \
  86             /* This is an illegal character.  */                              \
  87             STANDARD_FROM_LOOP_ERR_HANDLER (2);                               \
  88           }                                                                   \
  89         else                                                                  \
  90           {                                                                   \
  91             put32 (outptr, res);                                              \
  92             outptr += 4;                                                      \
  93             inptr += 2;                                                       \
  94           }                                                                   \
  95       }                                                                       \
  96     else                                                                      \
  97       {                                                                       \
  98         if (res == 0x1c)                                                      \
  99           res = 0x1a;                                                         \
 100         else if (res == 0x7f)                                                 \
 101           res = 0x1c;                                                         \
 102         else if (res == 0xa5)                                                 \
 103           res = 0x5c;                                                         \
 104         else if (res == 0x203e)                                               \
 105           res = 0x7e;                                                         \
 106         else if (res == 0x1a)                                                 \
 107           res = 0x7f;                                                         \
 108         put32 (outptr, res);                                                  \
 109         outptr += 4;                                                          \
 110         inptr++;                                                              \
 111       }                                                                       \
 112   }
 113 #define LOOP_NEED_FLAGS
 114 #define ONEBYTE_BODY \
 115   {                                                                           \
 116     if (c == 0x80 || c == 0xa0 || c >= 0xe0)                                  \
 117       return WEOF;                                                            \
 118     uint32_t res = __ibm943sb_to_ucs4[c];                                     \
 119     if (res == 0 && c != 0)                                                   \
 120       return WEOF;                                                            \
 121     if (res == 0x1c)                                                          \
 122       res = 0x1a;                                                             \
 123     else if (res == 0x7f)                                                     \
 124       res = 0x1c;                                                             \
 125     else if (res == 0xa5)                                                     \
 126       res = 0x5c;                                                             \
 127     else if (res == 0x203e)                                                   \
 128       res = 0x7e;                                                             \
 129     else if (res == 0x1a)                                                     \
 130       res = 0x7f;                                                             \
 131     return res;                                                               \
 132   }
 133 #include <iconv/loop.c>
 134
 135 /* Next, define the other direction.  */
 136 #define MIN_NEEDED_INPUT        MIN_NEEDED_TO
 137 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_FROM
 138 #define MAX_NEEDED_OUTPUT       MAX_NEEDED_FROM
 139 #define LOOPFCT                 TO_LOOP
 140 #define BODY \
 141   {                                                                           \
 142     const struct gap *rp = __ucs4_to_ibm943sb_idx;                            \
 143     unsigned char sc;                                                         \
 144     uint32_t ch = get32(inptr);                                               \
 145     bool found = true;                                                        \
 146     uint32_t i;                                                               \
 147     uint32_t low;                                                             \
 148     uint32_t high;                                                            \
 149     uint16_t pccode;                                                          \
 150                                                                               \
 151     if (__builtin_expect (ch >= 0xffff, 0))                                   \
 152       {                                                                       \
 153         UNICODE_TAG_HANDLER (ch, 4);                                          \
 154         rp = NULL;                                                            \
 155       }                                                                       \
 156     else                                                                      \
 157       while (ch > rp->end)                                                    \
 158         ++rp;                                                                 \
 159                                                                               \
 160     /* Use the UCS4 table for single byte.  */                                \
 161     if (__builtin_expect (rp == NULL, 0)                                      \
 162         || __builtin_expect (ch < rp->start, 0)                               \
 163         || (sc = __ucs4_to_ibm943sb[ch + rp->idx],                            \
 164         __builtin_expect (sc, '\1') == '\0' && ch != L'\0'))                  \
 165       {                                                                       \
 166                                                                               \
 167         /* Use the UCS4 table for double byte.  */                            \
 168         found = false;                                                        \
 169         low = 0;                                                              \
 170         high = (sizeof (__ucs4_to_ibm943db) >> 1)                             \
 171                 / sizeof (__ucs4_to_ibm943db[0][FROM]);                       \
 172         pccode = ch;                                                          \
 173         if (__builtin_expect (rp != NULL, 1))                                 \
 174           while (low < high)                                                  \
 175             {                                                                 \
 176               i = (low + high) >> 1;                                          \
 177               if (pccode < __ucs4_to_ibm943db[i][FROM])                       \
 178                 high = i;                                                     \
 179               else if (pccode > __ucs4_to_ibm943db[i][FROM])                  \
 180                 low = i + 1;                                                  \
 181               else                                                            \
 182                 {                                                             \
 183                   pccode = __ucs4_to_ibm943db[i][TO];                         \
 184                   found = true;                                               \
 185                   break;                                                      \
 186                 }                                                             \
 187             }                                                                 \
 188         if (found)                                                            \
 189           {                                                                   \
 190             if (__builtin_expect (outptr + 2 > outend, 0))                    \
 191               {                                                               \
 192                 result = __GCONV_FULL_OUTPUT;                                 \
 193                 break;                                                        \
 194               }                                                               \
 195             *outptr++ = pccode >> 8 & 0xff;                                   \
 196             *outptr++ = pccode & 0xff;                                        \
 197           }                                                                   \
 198         else                                                                  \
 199           {                                                                   \
 200             /* This is an illegal character.  */                              \
 201             STANDARD_TO_LOOP_ERR_HANDLER (4);                                 \
 202           }                                                                   \
 203       }                                                                       \
 204     else                                                                      \
 205       {                                                                       \
 206         if (__builtin_expect (outptr + 1 > outend, 0))                        \
 207           {                                                                   \
 208             result = __GCONV_FULL_OUTPUT;                                     \
 209             break;                                                            \
 210           }                                                                   \
 211         if (ch == 0x5c)                                                       \
 212           *outptr++ = 0x5c;                                                   \
 213         else if (ch == 0x7e)                                                  \
 214           *outptr++ = 0x7e;                                                   \
 215         else                                                                  \
 216           *outptr++ = sc;                                                     \
 217       }                                                                       \
 218                                                                               \
 219     /* Now that we wrote the output increment the input pointer.  */          \
 220     inptr += 4;                                                               \
 221   }
 222 #define LOOP_NEED_FLAGS
 223 #include <iconv/loop.c>
 224
 225 /* Now define the toplevel functions.  */
 226 #include <iconv/skeleton.c>