iconvdata/ibm937.c

   1 /* Conversion to and from IBM937.
   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@yamato.ibm.co.jp>, 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 <wchar.h>
  24 #include <byteswap.h>
  25 #include "ibm937.h"
  26
  27 /* The shift sequences for this charset (it does not use ESC).  */
  28 #define SI              0x0F  /* Shift In, host code to turn DBCS off.  */
  29 #define SO              0x0E  /* Shift Out, host code to turn DBCS on.  */
  30
  31 /* Definitions used in the body of the `gconv' function.  */
  32 #define CHARSET_NAME    "IBM937//"
  33 #define FROM_LOOP       from_ibm937
  34 #define TO_LOOP         to_ibm937
  35 #define MIN_NEEDED_FROM 1
  36 #define MAX_NEEDED_FROM 2
  37 #define MIN_NEEDED_TO   4
  38 #define MAX_NEEDED_TO   4
  39 #define PREPARE_LOOP \
  40   int save_curcs;                                                             \
  41   int *curcsp = &data->__statep->__count;
  42 #define EXTRA_LOOP_ARGS         , curcsp
  43
  44 /* Definitions of initialization and destructor function.  */
  45 #define DEFINE_INIT     1
  46 #define DEFINE_FINI     1
  47
  48
  49 /* Since this is a stateful encoding we have to provide code which resets
  50    the output state to the initial state.  This has to be done during the
  51    flushing.  */
  52 #define EMIT_SHIFT_TO_INIT \
  53   if ((data->__statep->__count & ~7) != sb)                                   \
  54     {                                                                         \
  55       if (FROM_DIRECTION)                                                     \
  56         data->__statep->__count &= 7;                                         \
  57       else                                                                    \
  58         {                                                                     \
  59           /* We are not in the initial state.  To switch back we have         \
  60              to emit `SI'.  */                                                \
  61           if (__builtin_expect (outbuf >= outend, 0))                         \
  62             /* We don't have enough room in the output buffer.  */            \
  63             status = __GCONV_FULL_OUTPUT;                                     \
  64           else                                                                \
  65             {                                                                 \
  66               /* Write out the shift sequence.  */                            \
  67               *outbuf++ = SI;                                                 \
  68               data->__statep->__count &= 7;                                   \
  69             }                                                                 \
  70         }                                                                     \
  71     }
  72
  73
  74 /* Since we might have to reset input pointer we must be able to save
  75    and retore the state.  */
  76 #define SAVE_RESET_STATE(Save) \
  77   if (Save)                                                                   \
  78     save_curcs = *curcsp;                                                     \
  79   else                                                                        \
  80     *curcsp = save_curcs
  81
  82
  83 /* Current codeset type.  */
  84 enum
  85 {
  86   sb = 0,
  87   db = 64
  88 };
  89
  90 /* First, define the conversion function from IBM-937 to UCS4.  */
  91 #define MIN_NEEDED_INPUT        MIN_NEEDED_FROM
  92 #define MAX_NEEDED_INPUT        MAX_NEEDED_FROM
  93 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_TO
  94 #define LOOPFCT                 FROM_LOOP
  95 #define BODY \
  96   {                                                                           \
  97     uint32_t ch = *inptr;                                                     \
  98                                                                               \
  99     if (__builtin_expect (ch, 0) == SO)                                       \
 100       {                                                                       \
 101         /* Shift OUT, change to DBCS converter.  */                           \
 102         if (curcs == db)                                                      \
 103           {                                                                   \
 104             result = __GCONV_ILLEGAL_INPUT;                                   \
 105             break;                                                            \
 106           }                                                                   \
 107         curcs = db;                                                           \
 108         ++inptr;                                                              \
 109         continue;                                                             \
 110       }                                                                       \
 111     else if (__builtin_expect (ch, 0) == SI)                                  \
 112       {                                                                       \
 113         /* Shift IN, change to SBCS converter.  */                            \
 114         if (curcs == sb)                                                      \
 115           {                                                                   \
 116             result = __GCONV_ILLEGAL_INPUT;                                   \
 117             break;                                                            \
 118           }                                                                   \
 119         curcs = sb;                                                           \
 120         ++inptr;                                                              \
 121         continue;                                                             \
 122       }                                                                       \
 123                                                                               \
 124     if (curcs == sb)                                                          \
 125       {                                                                       \
 126         /* Use the UCS4 table for single byte.  */                            \
 127         ch = __ibm937sb_to_ucs4[ch];                                          \
 128         if (__builtin_expect (ch, L'\1') == L'\0' && *inptr != '\0')          \
 129           {                                                                   \
 130             /* This is an illegal character.  */                              \
 131             if (! ignore_errors_p ())                                         \
 132               {                                                               \
 133                 result = __GCONV_ILLEGAL_INPUT;                               \
 134                 break;                                                        \
 135               }                                                               \
 136             ++*irreversible;                                                  \
 137           }                                                                   \
 138         else                                                                  \
 139           {                                                                   \
 140             put32 (outptr, ch);                                               \
 141             outptr += 4;                                                      \
 142           }                                                                   \
 143         ++inptr;                                                              \
 144       }                                                                       \
 145     else                                                                      \
 146       {                                                                       \
 147         /* Use the IBM937 table for double byte.  */                          \
 148                                                                               \
 149         assert (curcs == db);                                                 \
 150                                                                               \
 151         ch = ibm937db_to_ucs4(inptr[0], inptr[1]);                            \
 152         if (__builtin_expect (ch, L'\1') == L'\0' && *inptr != '\0')          \
 153           {                                                                   \
 154             /* This is an illegal character.  */                              \
 155             if (! ignore_errors_p ())                                         \
 156               {                                                               \
 157                 result = __GCONV_ILLEGAL_INPUT;                               \
 158                 break;                                                        \
 159               }                                                               \
 160             ++*irreversible;                                                  \
 161           }                                                                   \
 162         else                                                                  \
 163           {                                                                   \
 164             put32 (outptr, ch);                                               \
 165             outptr += 4;                                                      \
 166           }                                                                   \
 167         inptr += 2;                                                           \
 168       }                                                                       \
 169   }
 170 #define LOOP_NEED_FLAGS
 171 #define EXTRA_LOOP_DECLS        , int *curcsp
 172 #define INIT_PARAMS             int curcs = *curcsp & ~7
 173 #define UPDATE_PARAMS           *curcsp = curcs
 174 #include <iconv/loop.c>
 175
 176 /* Next, define the other direction.  */
 177 #define MIN_NEEDED_INPUT        MIN_NEEDED_TO
 178 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_FROM
 179 #define MAX_NEEDED_OUTPUT       MAX_NEEDED_FROM
 180 #define LOOPFCT                 TO_LOOP
 181 #define BODY \
 182   {                                                                           \
 183     uint32_t ch = get32 (inptr);                                              \
 184     const char *cp;                                                           \
 185                                                                               \
 186     /* Use the UCS4 table for single byte.  */                                \
 187     cp = __ucs4_to_ibm937sb[ch];                                              \
 188     if (__builtin_expect (ch >= (sizeof (__ucs4_to_ibm937sb)                  \
 189                                  / sizeof (__ucs4_to_ibm937sb[0])), 0)        \
 190         || (__builtin_expect (cp[0], '\1') == '\0' && ch != 0))               \
 191       {                                                                       \
 192         /* Use the UCS4 table for double byte.  */                            \
 193         cp = __ucs4_to_ibm937db[ch];                                          \
 194         if (__builtin_expect (ch >= (sizeof (__ucs4_to_ibm937db)              \
 195                                      / sizeof (__ucs4_to_ibm937db[0])), 0)    \
 196             || __builtin_expect (cp[0], '\1') == '\0')                        \
 197           {                                                                   \
 198             UNICODE_TAG_HANDLER (ch, 4);                                      \
 199                                                                               \
 200             /* This is an illegal character.  */                              \
 201             if (! ignore_errors_p ())                                         \
 202               {                                                               \
 203                 result = __GCONV_ILLEGAL_INPUT;                               \
 204                 break;                                                        \
 205               }                                                               \
 206             ++*irreversible;                                                  \
 207           }                                                                   \
 208         else                                                                  \
 209           {                                                                   \
 210             if (curcs == sb)                                                  \
 211               {                                                               \
 212                 *outptr++ = SO;                                               \
 213                 curcs = db;                                                   \
 214               }                                                               \
 215             if (__builtin_expect (outptr + 1 >= outend, 0))                   \
 216               {                                                               \
 217                 result = __GCONV_FULL_OUTPUT;                                 \
 218                 break;                                                        \
 219               }                                                               \
 220             *outptr++ = cp[0];                                                \
 221             *outptr++ = cp[1];                                                \
 222           }                                                                   \
 223       }                                                                       \
 224     else                                                                      \
 225       {                                                                       \
 226         if (curcs == db)                                                      \
 227           {                                                                   \
 228             *outptr++ = SI;                                                   \
 229             curcs = sb;                                                       \
 230             if (__builtin_expect (outptr == outend, 0))                       \
 231               {                                                               \
 232                 result = __GCONV_FULL_OUTPUT;                                 \
 233                 break;                                                        \
 234               }                                                               \
 235           }                                                                   \
 236         *outptr++ = cp[0];                                                    \
 237       }                                                                       \
 238                                                                               \
 239     /* Now that we wrote the output increment the input pointer.  */          \
 240     inptr += 4;                                                               \
 241   }
 242 #define LOOP_NEED_FLAGS
 243 #define EXTRA_LOOP_DECLS        , int *curcsp
 244 #define INIT_PARAMS             int curcs = *curcsp & ~7
 245 #define UPDATE_PARAMS           *curcsp = curcs
 246 #include <iconv/loop.c>
 247
 248 /* Now define the toplevel functions.  */
 249 #include <iconv/skeleton.c>