iconvdata/ibm932.c

   1 /* Conversion from and to IBM932.
   2    Copyright (C) 2000-2021 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, see
  18    <https://www.gnu.org/licenses/>.  */
  19
  20 #include <dlfcn.h>
  21 #include <stdint.h>
  22 #include <stdbool.h>
  23 #include "ibm932.h"
  24
  25 #define FROM    0
  26 #define TO      1
  27
  28 /* Definitions used in the body of the `gconv' function.  */
  29 #define CHARSET_NAME    "IBM932//"
  30 #define FROM_LOOP       from_ibm932
  31 #define TO_LOOP         to_ibm932
  32 #define ONE_DIRECTION   0
  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-932 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 = __ibm932db_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 IBM932 table for single byte.  */                              \
  64     res = __ibm932sb_to_ucs4[ch];                                             \
  65     if (__builtin_expect (res == 0, 0) && ch != 0)                            \
  66       {                                                                       \
  67         /* Use the IBM932 table for double byte.  */                          \
  68         if (__glibc_unlikely (inptr + 1 >= inend))                            \
  69           {                                                                   \
  70             /* The second character is not available.                         \
  71                Store the intermediate result.  */                             \
  72             result = __GCONV_INCOMPLETE_INPUT;                                \
  73             break;                                                            \
  74           }                                                                   \
  75                                                                               \
  76         ch = (ch * 0x100) + inptr[1];                                         \
  77         /* ch was less than 0xfd.  */                                         \
  78         assert (ch < 0xfd00);                                                 \
  79         while (ch > rp2->end)                                                 \
  80           ++rp2;                                                              \
  81                                                                               \
  82         if (__builtin_expect (ch < rp2->start, 0)                             \
  83             || (res = __ibm932db_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 == 0xa5)                                                      \
  99           res = 0x5c;                                                         \
 100         else if (res == 0x203e)                                               \
 101           res = 0x7e;                                                         \
 102         put32 (outptr, res);                                                  \
 103         outptr += 4;                                                          \
 104         inptr++;                                                              \
 105       }                                                                       \
 106   }
 107 #define LOOP_NEED_FLAGS
 108 #define ONEBYTE_BODY \
 109   {                                                                           \
 110     if (c == 0x80 || c == 0xa0 || c >= 0xfd)                                  \
 111       return WEOF;                                                            \
 112     uint32_t res = __ibm932sb_to_ucs4[c];                                     \
 113     if (res == 0 && c != 0)                                                   \
 114       return WEOF;                                                            \
 115     if (res == 0xa5)                                                          \
 116       res = 0x5c;                                                             \
 117     else if (res == 0x203e)                                                   \
 118       res = 0x7e;                                                             \
 119     return res;                                                               \
 120   }
 121 #include <iconv/loop.c>
 122
 123 /* Next, define the other direction.  */
 124 #define MIN_NEEDED_INPUT        MIN_NEEDED_TO
 125 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_FROM
 126 #define MAX_NEEDED_OUTPUT       MAX_NEEDED_FROM
 127 #define LOOPFCT                 TO_LOOP
 128 #define BODY \
 129   {                                                                           \
 130     const struct gap *rp = __ucs4_to_ibm932sb_idx;                            \
 131     unsigned char sc;                                                         \
 132     uint32_t ch = get32 (inptr);                                              \
 133     bool found = true;                                                        \
 134     uint32_t i;                                                               \
 135     uint32_t low;                                                             \
 136     uint32_t high;                                                            \
 137     uint16_t pccode;                                                          \
 138                                                                               \
 139     if (__glibc_unlikely (ch >= 0xffff))                                      \
 140       {                                                                       \
 141         UNICODE_TAG_HANDLER (ch, 4);                                          \
 142         rp = NULL;                                                            \
 143       }                                                                       \
 144     else                                                                      \
 145       while (ch > rp->end)                                                    \
 146         ++rp;                                                                 \
 147                                                                               \
 148     /* Use the UCS4 table for single byte.  */                                \
 149     if (__builtin_expect (rp == NULL, 0)                                      \
 150         || __builtin_expect (ch < rp->start, 0)                               \
 151         || (sc = __ucs4_to_ibm932sb[ch + rp->idx],                            \
 152         __builtin_expect (sc, '\1') == '\0' && ch != L'\0'))                  \
 153       {                                                                       \
 154                                                                               \
 155         /* Use the UCS4 table for double byte.  */                            \
 156         found = false;                                                        \
 157         low = 0;                                                              \
 158         high = (sizeof (__ucs4_to_ibm932db) >> 1)                             \
 159                 / sizeof (__ucs4_to_ibm932db[0][FROM]);                       \
 160         pccode = ch;                                                          \
 161         if (__glibc_likely (rp != NULL))                                      \
 162           while (low < high)                                                  \
 163             {                                                                 \
 164               i = (low + high) >> 1;                                          \
 165               if (pccode < __ucs4_to_ibm932db[i][FROM])                       \
 166                 high = i;                                                     \
 167               else if (pccode > __ucs4_to_ibm932db[i][FROM])                  \
 168                 low = i + 1;                                                  \
 169               else                                                            \
 170                 {                                                             \
 171                   pccode = __ucs4_to_ibm932db[i][TO];                         \
 172                   found = true;                                               \
 173                   break;                                                      \
 174                 }                                                             \
 175             }                                                                 \
 176         if (found)                                                            \
 177           {                                                                   \
 178             if (__glibc_unlikely (outptr + 2 > outend))                       \
 179               {                                                               \
 180                 result = __GCONV_FULL_OUTPUT;                                 \
 181                 break;                                                        \
 182               }                                                               \
 183             *outptr++ = pccode >> 8 & 0xff;                                   \
 184             *outptr++ = pccode & 0xff;                                        \
 185           }                                                                   \
 186         else                                                                  \
 187           {                                                                   \
 188             /* This is an illegal character.  */                              \
 189             STANDARD_TO_LOOP_ERR_HANDLER (4);                                 \
 190           }                                                                   \
 191       }                                                                       \
 192     else                                                                      \
 193       {                                                                       \
 194         if (__glibc_unlikely (outptr + 1 > outend))                           \
 195           {                                                                   \
 196             result = __GCONV_FULL_OUTPUT;                                     \
 197             break;                                                            \
 198           }                                                                   \
 199         if (ch == 0x5c)                                                       \
 200           *outptr++ = 0x5c;                                                   \
 201         else if (ch == 0x7e)                                                  \
 202           *outptr++ = 0x7e;                                                   \
 203         else                                                                  \
 204           *outptr++ = sc;                                                     \
 205       }                                                                       \
 206                                                                               \
 207     /* Now that we wrote the output increment the input pointer.  */          \
 208     inptr += 4;                                                               \
 209   }
 210 #define LOOP_NEED_FLAGS
 211 #include <iconv/loop.c>
 212
 213 /* Now define the toplevel functions.  */
 214 #include <iconv/skeleton.c>