iconvdata/ibm939.c

   1 /* Conversion to and from IBM939.
   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 "ibm939.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    "IBM939//"
  33 #define FROM_LOOP       from_ibm939
  34 #define TO_LOOP         to_ibm939
  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-939 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     uint32_t res;                                                             \
  99                                                                               \
 100     if (__builtin_expect (ch, 0) == SO)                                       \
 101       {                                                                       \
 102         /* Shift OUT, change to DBCS converter.  */                           \
 103         if (curcs == db)                                                      \
 104           {                                                                   \
 105             result = __GCONV_ILLEGAL_INPUT;                                   \
 106             break;                                                            \
 107           }                                                                   \
 108         curcs = db;                                                           \
 109         ++inptr;                                                              \
 110         continue;                                                             \
 111       }                                                                       \
 112     else if (__builtin_expect (ch, 0) == SI)                                  \
 113       {                                                                       \
 114         /* Shift IN, change to SBCS converter.  */                            \
 115         if (curcs == sb)                                                      \
 116           {                                                                   \
 117             result = __GCONV_ILLEGAL_INPUT;                                   \
 118             break;                                                            \
 119           }                                                                   \
 120         curcs = sb;                                                           \
 121         ++inptr;                                                              \
 122         continue;                                                             \
 123       }                                                                       \
 124                                                                               \
 125     if (curcs == sb)                                                          \
 126       {                                                                       \
 127         /* Use the IBM939 table for single byte.  */                          \
 128         res = __ibm939sb_to_ucs4[ch];                                         \
 129         if (__builtin_expect (res, L'\1') == L'\0' && ch != '\0')             \
 130           {                                                                   \
 131             /* This is an illegal character.  */                              \
 132             if (! ignore_errors_p ())                                         \
 133               {                                                               \
 134                 result = __GCONV_ILLEGAL_INPUT;                               \
 135                 break;                                                        \
 136               }                                                               \
 137             ++*irreversible;                                                  \
 138           }                                                                   \
 139         else                                                                  \
 140           {                                                                   \
 141             put32 (outptr, res);                                              \
 142             outptr += 4;                                                      \
 143           }                                                                   \
 144         ++inptr;                                                              \
 145       }                                                                       \
 146     else                                                                      \
 147       {                                                                       \
 148         /* Use the IBM939 table for double byte.  */                          \
 149         const struct gap *rp2 = __ibm939db_to_ucs4_idx;                       \
 150                                                                               \
 151         assert (curcs == db);                                                 \
 152                                                                               \
 153         if (__builtin_expect (inptr + 1 >= inend, 0))                         \
 154           {                                                                   \
 155             /* The second character is not available.  Store the              \
 156                intermediate result. */                                        \
 157             result = __GCONV_INCOMPLETE_INPUT;                                \
 158             break;                                                            \
 159           }                                                                   \
 160                                                                               \
 161         ch = (ch * 0x100) + inptr[1];                                         \
 162         while (ch > rp2->end)                                                 \
 163           ++rp2;                                                              \
 164                                                                               \
 165         if (__builtin_expect (rp2 == NULL, 0)                                 \
 166             || __builtin_expect (ch < rp2->start, 0)                          \
 167             || (res = __ibm939db_to_ucs4[ch + rp2->idx],                      \
 168                 __builtin_expect (res, L'\1') == L'\0' && ch != '\0'))        \
 169           {                                                                   \
 170             /* This is an illegal character.  */                              \
 171             if (! ignore_errors_p ())                                         \
 172               {                                                               \
 173                 result = __GCONV_ILLEGAL_INPUT;                               \
 174                 break;                                                        \
 175               }                                                               \
 176             ++*irreversible;                                                  \
 177           }                                                                   \
 178         else                                                                  \
 179           {                                                                   \
 180             put32 (outptr, res);                                              \
 181             outptr += 4;                                                      \
 182           }                                                                   \
 183         inptr += 2;                                                           \
 184       }                                                                       \
 185   }
 186 #define LOOP_NEED_FLAGS
 187 #define EXTRA_LOOP_DECLS        , int *curcsp
 188 #define INIT_PARAMS             int curcs = *curcsp & ~7
 189 #define UPDATE_PARAMS           *curcsp = curcs
 190 #include <iconv/loop.c>
 191
 192 /* Next, define the other direction */
 193 #define MIN_NEEDED_INPUT        MIN_NEEDED_TO
 194 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_FROM
 195 #define MAX_NEEDED_OUTPUT       MAX_NEEDED_FROM
 196 #define LOOPFCT                 TO_LOOP
 197 #define BODY \
 198   {                                                                           \
 199     uint32_t ch = get32 (inptr);                                              \
 200     const struct gap *rp1 = __ucs4_to_ibm939sb_idx;                           \
 201     const struct gap *rp2 = __ucs4_to_ibm939db_idx;                           \
 202     const char *cp;                                                           \
 203                                                                               \
 204     if (__builtin_expect (ch, 0) >= 0xffff)                                   \
 205       {                                                                       \
 206         UNICODE_TAG_HANDLER (ch, 4);                                          \
 207         rp1 = NULL;                                                           \
 208         rp2 = NULL;                                                           \
 209       }                                                                       \
 210     else                                                                      \
 211       while (ch > rp1->end)                                                   \
 212         ++rp1;                                                                \
 213                                                                               \
 214     /* Use the UCS4 table for single byte.  */                                \
 215     if (__builtin_expect (rp1 == NULL, 0)                                     \
 216         || __builtin_expect (ch < rp1->start, 0)                              \
 217         || (cp = __ucs4_to_ibm939sb[ch + rp1->idx],                           \
 218             __builtin_expect (cp[0], L'\1') == L'\0' && ch != '\0'))          \
 219       {                                                                       \
 220         /* Use the UCS4 table for double byte.  */                            \
 221         while (ch > rp2->end)                                                 \
 222           ++rp2;                                                              \
 223                                                                               \
 224         if (__builtin_expect (rp2 == NULL, 0)                                 \
 225             || __builtin_expect (ch < rp2->start, 0)                          \
 226             || (cp = __ucs4_to_ibm939db[ch + rp2->idx],                       \
 227                 __builtin_expect (cp[0], L'\1')==L'\0' && ch != '\0'))        \
 228           {                                                                   \
 229             /* This is an illegal character.  */                              \
 230             if (! ignore_errors_p ())                                         \
 231               {                                                               \
 232                 result = __GCONV_ILLEGAL_INPUT;                               \
 233                 break;                                                        \
 234               }                                                               \
 235             ++*irreversible;                                                  \
 236           }                                                                   \
 237         else                                                                  \
 238           {                                                                   \
 239             if (curcs == sb)                                                  \
 240               {                                                               \
 241                 if (__builtin_expect (outptr+1 > outend, 0))                  \
 242                   {                                                           \
 243                     result = __GCONV_FULL_OUTPUT;                             \
 244                     break;                                                    \
 245                   }                                                           \
 246                 *outptr++ = SO;                                               \
 247                 curcs = db;                                                   \
 248               }                                                               \
 249                                                                               \
 250             if (__builtin_expect (outptr + 2 > outend, 0))                    \
 251               {                                                               \
 252                 result = __GCONV_FULL_OUTPUT;                                 \
 253                 break;                                                        \
 254               }                                                               \
 255             *outptr++ = cp[0];                                                \
 256             *outptr++ = cp[1];                                                \
 257           }                                                                   \
 258       }                                                                       \
 259     else                                                                      \
 260       {                                                                       \
 261         if (curcs == db)                                                      \
 262           {                                                                   \
 263             if (__builtin_expect (outptr + 1 > outend, 0))                    \
 264               {                                                               \
 265                 result = __GCONV_FULL_OUTPUT;                                 \
 266                 break;                                                        \
 267               }                                                               \
 268             *outptr++ = SI;                                                   \
 269           }                                                                   \
 270                                                                               \
 271         if (__builtin_expect (outptr + 1 > outend, 0))                        \
 272           {                                                                   \
 273             result = __GCONV_FULL_OUTPUT;                                     \
 274             break;                                                            \
 275           }                                                                   \
 276         if (ch == 0x7e)                                                       \
 277           *outptr++ = 0xa0;                                                   \
 278         else if (ch == 0x5c)                                                  \
 279           *outptr++ = 0xb2;                                                   \
 280         else                                                                  \
 281           *outptr++ = cp[0];                                                  \
 282         curcs = sb;                                                           \
 283       }                                                                       \
 284                                                                               \
 285     /* Now that we wrote the output increment the input pointer.  */          \
 286     inptr += 4;                                                               \
 287   }
 288 #define LOOP_NEED_FLAGS
 289 #define EXTRA_LOOP_DECLS        , int *curcsp
 290 #define INIT_PARAMS             int curcs = *curcsp & ~7
 291 #define UPDATE_PARAMS           *curcsp = curcs
 292 #include <iconv/loop.c>
 293
 294 /* Now define the toplevel functions. */
 295 #include <iconv/skeleton.c>