iconvdata/ibm943.c

   1 /* Conversion from and to IBM943.
   2    Copyright (C) 2000-2018 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    <http://www.gnu.org/licenses/>.  */
  19
  20 #include <dlfcn.h>
  21 #include <stdint.h>
  22 #include <stdbool.h>
  23 #include "ibm943.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    "IBM943//"
  30 #define FROM_LOOP       from_ibm943
  31 #define TO_LOOP         to_ibm943
  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-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 (__glibc_unlikely (inptr + 1 >= inend))                            \
  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         /* ch was less than 0xfd.  */                                         \
  79         assert (ch < 0xfd00);                                                 \
  80         while (ch > rp2->end)                                                 \
  81           ++rp2;                                                              \
  82                                                                               \
  83         if (__builtin_expect (ch < rp2->start, 0)                             \
  84             || (res = __ibm943db_to_ucs4[ch + rp2->idx],                      \
  85             __builtin_expect (res, '\1') == 0 && ch !=0))                     \
  86           {                                                                   \
  87             /* This is an illegal character.  */                              \
  88             STANDARD_FROM_LOOP_ERR_HANDLER (2);                               \
  89           }                                                                   \
  90         else                                                                  \
  91           {                                                                   \
  92             put32 (outptr, res);                                              \
  93             outptr += 4;                                                      \
  94             inptr += 2;                                                       \
  95           }                                                                   \
  96       }                                                                       \
  97     else                                                                      \
  98       {                                                                       \
  99         if (res == 0xa5)                                                      \
 100           res = 0x5c;                                                         \
 101         else if (res == 0x203e)                                               \
 102           res = 0x7e;                                                         \
 103         put32 (outptr, res);                                                  \
 104         outptr += 4;                                                          \
 105         inptr++;                                                              \
 106       }                                                                       \
 107   }
 108 #define LOOP_NEED_FLAGS
 109 #define ONEBYTE_BODY \
 110   {                                                                           \
 111     if (c == 0x80 || c == 0xa0 || c >= 0xe0)                                  \
 112       return WEOF;                                                            \
 113     uint32_t res = __ibm943sb_to_ucs4[c];                                     \
 114     if (res == 0 && c != 0)                                                   \
 115       return WEOF;                                                            \
 116     if (res == 0xa5)                                                          \
 117       res = 0x5c;                                                             \
 118     else if (res == 0x203e)                                                   \
 119       res = 0x7e;                                                             \
 120     return res;                                                               \
 121   }
 122 #include <iconv/loop.c>
 123
 124 /* Next, define the other direction.  */
 125 #define MIN_NEEDED_INPUT        MIN_NEEDED_TO
 126 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_FROM
 127 #define MAX_NEEDED_OUTPUT       MAX_NEEDED_FROM
 128 #define LOOPFCT                 TO_LOOP
 129 #define BODY \
 130   {                                                                           \
 131     const struct gap *rp = __ucs4_to_ibm943sb_idx;                            \
 132     unsigned char sc;                                                         \
 133     uint32_t ch = get32(inptr);                                               \
 134     bool found = true;                                                        \
 135     uint32_t i;                                                               \
 136     uint32_t low;                                                             \
 137     uint32_t high;                                                            \
 138     uint16_t pccode;                                                          \
 139                                                                               \
 140     if (__glibc_unlikely (ch >= 0xffff))                                      \
 141       {                                                                       \
 142         UNICODE_TAG_HANDLER (ch, 4);                                          \
 143         rp = NULL;                                                            \
 144       }                                                                       \
 145     else                                                                      \
 146       while (ch > rp->end)                                                    \
 147         ++rp;                                                                 \
 148                                                                               \
 149     /* Use the UCS4 table for single byte.  */                                \
 150     if (__builtin_expect (rp == NULL, 0)                                      \
 151         || __builtin_expect (ch < rp->start, 0)                               \
 152         || (sc = __ucs4_to_ibm943sb[ch + rp->idx],                            \
 153         __builtin_expect (sc, '\1') == '\0' && ch != L'\0'))                  \
 154       {                                                                       \
 155                                                                               \
 156         /* Use the UCS4 table for double byte.  */                            \
 157         found = false;                                                        \
 158         low = 0;                                                              \
 159         high = (sizeof (__ucs4_to_ibm943db) >> 1)                             \
 160                 / sizeof (__ucs4_to_ibm943db[0][FROM]);                       \
 161         pccode = ch;                                                          \
 162         if (__glibc_likely (rp != NULL))                                      \
 163           while (low < high)                                                  \
 164             {                                                                 \
 165               i = (low + high) >> 1;                                          \
 166               if (pccode < __ucs4_to_ibm943db[i][FROM])                       \
 167                 high = i;                                                     \
 168               else if (pccode > __ucs4_to_ibm943db[i][FROM])                  \
 169                 low = i + 1;                                                  \
 170               else                                                            \
 171                 {                                                             \
 172                   pccode = __ucs4_to_ibm943db[i][TO];                         \
 173                   found = true;                                               \
 174                   break;                                                      \
 175                 }                                                             \
 176             }                                                                 \
 177         if (found)                                                            \
 178           {                                                                   \
 179             if (__glibc_unlikely (outptr + 2 > outend))                       \
 180               {                                                               \
 181                 result = __GCONV_FULL_OUTPUT;                                 \
 182                 break;                                                        \
 183               }                                                               \
 184             *outptr++ = pccode >> 8 & 0xff;                                   \
 185             *outptr++ = pccode & 0xff;                                        \
 186           }                                                                   \
 187         else                                                                  \
 188           {                                                                   \
 189             /* This is an illegal character.  */                              \
 190             STANDARD_TO_LOOP_ERR_HANDLER (4);                                 \
 191           }                                                                   \
 192       }                                                                       \
 193     else                                                                      \
 194       {                                                                       \
 195         if (__glibc_unlikely (outptr + 1 > outend))                           \
 196           {                                                                   \
 197             result = __GCONV_FULL_OUTPUT;                                     \
 198             break;                                                            \
 199           }                                                                   \
 200         if (ch == 0x5c)                                                       \
 201           *outptr++ = 0x5c;                                                   \
 202         else if (ch == 0x7e)                                                  \
 203           *outptr++ = 0x7e;                                                   \
 204         else                                                                  \
 205           *outptr++ = sc;                                                     \
 206       }                                                                       \
 207                                                                               \
 208     /* Now that we wrote the output increment the input pointer.  */          \
 209     inptr += 4;                                                               \
 210   }
 211 #define LOOP_NEED_FLAGS
 212 #include <iconv/loop.c>
 213
 214 /* Now define the toplevel functions.  */
 215 #include <iconv/skeleton.c>