iconvdata/ibm937.c

   1 /* Conversion from and to IBM937.
   2    Copyright (C) 2000-2016 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, see
  18    <http://www.gnu.org/licenses/>.  */
  19
  20 /* IBM937 is designed for the representation of Traditional Chinese
  21    using a stateful EBCDIC encoding scheme.  It is also known as
  22    CCSID 937 or CP937. See:
  23    https://www-01.ibm.com/software/globalization/ccsid/ccsid937.html */
  24
  25 #include <dlfcn.h>
  26 #include <stdint.h>
  27 #include <wchar.h>
  28 #include <byteswap.h>
  29 #include "ibm937.h"
  30
  31 /* The shift sequences for this charset (it does not use ESC).  */
  32 #define SI              0x0F  /* Shift In, host code to turn DBCS off.  */
  33 #define SO              0x0E  /* Shift Out, host code to turn DBCS on.  */
  34
  35 /* Definitions used in the body of the `gconv' function.  */
  36 #define CHARSET_NAME    "IBM937//"
  37 #define FROM_LOOP       from_ibm937
  38 #define TO_LOOP         to_ibm937
  39 #define ONE_DIRECTION                   0
  40 #define FROM_LOOP_MIN_NEEDED_FROM       1
  41 #define FROM_LOOP_MAX_NEEDED_FROM       2
  42 #define FROM_LOOP_MIN_NEEDED_TO         4
  43 #define FROM_LOOP_MAX_NEEDED_TO         4
  44 #define TO_LOOP_MIN_NEEDED_FROM         4
  45 #define TO_LOOP_MAX_NEEDED_FROM         4
  46 #define TO_LOOP_MIN_NEEDED_TO           1
  47 #define TO_LOOP_MAX_NEEDED_TO           3
  48 #define PREPARE_LOOP \
  49   int save_curcs;                                                             \
  50   int *curcsp = &data->__statep->__count;
  51 #define EXTRA_LOOP_ARGS         , curcsp
  52
  53 /* Definitions of initialization and destructor function.  */
  54 #define DEFINE_INIT     1
  55 #define DEFINE_FINI     1
  56
  57
  58 /* Since this is a stateful encoding we have to provide code which resets
  59    the output state to the initial state.  This has to be done during the
  60    flushing.  */
  61 #define EMIT_SHIFT_TO_INIT \
  62   if ((data->__statep->__count & ~7) != sb)                                   \
  63     {                                                                         \
  64       if (FROM_DIRECTION)                                                     \
  65         data->__statep->__count &= 7;                                         \
  66       else                                                                    \
  67         {                                                                     \
  68           /* We are not in the initial state.  To switch back we have         \
  69              to emit `SI'.  */                                                \
  70           if (__glibc_unlikely (outbuf >= outend))                            \
  71             /* We don't have enough room in the output buffer.  */            \
  72             status = __GCONV_FULL_OUTPUT;                                     \
  73           else                                                                \
  74             {                                                                 \
  75               /* Write out the shift sequence.  */                            \
  76               *outbuf++ = SI;                                                 \
  77               data->__statep->__count &= 7;                                   \
  78             }                                                                 \
  79         }                                                                     \
  80     }
  81
  82
  83 /* Since we might have to reset input pointer we must be able to save
  84    and retore the state.  */
  85 #define SAVE_RESET_STATE(Save) \
  86   if (Save)                                                                   \
  87     save_curcs = *curcsp;                                                     \
  88   else                                                                        \
  89     *curcsp = save_curcs
  90
  91
  92 /* Current codeset type.  */
  93 enum
  94 {
  95   sb = 0,
  96   db = 64
  97 };
  98
  99 /* First, define the conversion function from IBM-937 to UCS4.  */
 100 #define MIN_NEEDED_INPUT        FROM_LOOP_MIN_NEEDED_FROM
 101 #define MAX_NEEDED_INPUT        FROM_LOOP_MAX_NEEDED_FROM
 102 #define MIN_NEEDED_OUTPUT       FROM_LOOP_MIN_NEEDED_TO
 103 #define MAX_NEEDED_OUTPUT       FROM_LOOP_MAX_NEEDED_TO
 104 #define LOOPFCT                 FROM_LOOP
 105 #define BODY \
 106   {                                                                           \
 107     uint32_t ch = *inptr;                                                     \
 108     uint32_t res;                                                             \
 109                                                                               \
 110     if (__builtin_expect (ch, 0) == SO)                                       \
 111       {                                                                       \
 112         /* Shift OUT, change to DBCS converter.  */                           \
 113         if (curcs == db)                                                      \
 114           {                                                                   \
 115             result = __GCONV_ILLEGAL_INPUT;                                   \
 116             break;                                                            \
 117           }                                                                   \
 118         curcs = db;                                                           \
 119         ++inptr;                                                              \
 120         continue;                                                             \
 121       }                                                                       \
 122     else if (__builtin_expect (ch, 0) == SI)                                  \
 123       {                                                                       \
 124         /* Shift IN, change to SBCS converter.  */                            \
 125         if (curcs == sb)                                                      \
 126           {                                                                   \
 127             result = __GCONV_ILLEGAL_INPUT;                                   \
 128             break;                                                            \
 129           }                                                                   \
 130         curcs = sb;                                                           \
 131         ++inptr;                                                              \
 132         continue;                                                             \
 133       }                                                                       \
 134                                                                               \
 135     if (curcs == sb)                                                          \
 136       {                                                                       \
 137         /* Use the IBM937 table for single byte.  */                          \
 138         res = __ibm937sb_to_ucs4[ch];                                         \
 139         if (__builtin_expect (res, L'\1') == L'\0' && ch != '\0')             \
 140           {                                                                   \
 141             /* This is an illegal character.  */                              \
 142             STANDARD_FROM_LOOP_ERR_HANDLER (1);                               \
 143           }                                                                   \
 144         else                                                                  \
 145           {                                                                   \
 146             put32 (outptr, res);                                              \
 147             outptr += 4;                                                      \
 148           }                                                                   \
 149         ++inptr;                                                              \
 150       }                                                                       \
 151     else                                                                      \
 152       {                                                                       \
 153         const struct gap *rp2 = __ibm937db_to_ucs4_idx;                       \
 154                                                                               \
 155         assert (curcs == db);                                                 \
 156                                                                               \
 157         /* Use the IBM937 table for double byte.  */                          \
 158         if (__glibc_unlikely (inptr + 1 >= inend))                            \
 159           {                                                                   \
 160             /* The second character is not available.                         \
 161                Store the intermediate result. */                              \
 162             result = __GCONV_INCOMPLETE_INPUT;                                \
 163             break;                                                            \
 164           }                                                                   \
 165                                                                               \
 166         ch = (ch * 0x100) + inptr[1];                                         \
 167         while (ch > rp2->end)                                                 \
 168           ++rp2;                                                              \
 169                                                                               \
 170         if (__builtin_expect (rp2->start == 0xffff, 0)                        \
 171             || __builtin_expect (ch < rp2->start, 0)                          \
 172             || (res = __ibm937db_to_ucs4[ch + rp2->idx],                      \
 173                 __builtin_expect (res, L'\1') == L'\0' && ch != '\0'))        \
 174           {                                                                   \
 175             /* This is an illegal character.  */                              \
 176             STANDARD_FROM_LOOP_ERR_HANDLER (2);                               \
 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        TO_LOOP_MIN_NEEDED_FROM
 194 #define MAX_NEEDED_INPUT        TO_LOOP_MAX_NEEDED_FROM
 195 #define MIN_NEEDED_OUTPUT       TO_LOOP_MIN_NEEDED_TO
 196 #define MAX_NEEDED_OUTPUT       TO_LOOP_MAX_NEEDED_TO
 197 #define LOOPFCT                 TO_LOOP
 198 #define BODY \
 199   {                                                                           \
 200     uint32_t ch = get32 (inptr);                                              \
 201     const struct gap *rp1 = __ucs4_to_ibm937sb_idx;                           \
 202     const struct gap *rp2 = __ucs4_to_ibm937db_idx;                           \
 203     const char *cp;                                                           \
 204                                                                               \
 205     if (__glibc_unlikely (ch >= 0xffff))                                      \
 206       {                                                                       \
 207         UNICODE_TAG_HANDLER (ch, 4);                                          \
 208                                                                               \
 209         STANDARD_TO_LOOP_ERR_HANDLER (4);                                     \
 210       }                                                                       \
 211                                                                               \
 212     while (ch > rp1->end)                                                     \
 213       ++rp1;                                                                  \
 214                                                                               \
 215     /* Use the UCS4 table for single byte.  */                                \
 216     if (__builtin_expect (ch < rp1->start, 0)                                 \
 217         || (cp = __ucs4_to_ibm937sb[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 (ch < rp2->start, 0)                             \
 225             || (cp = __ucs4_to_ibm937db[ch + rp2->idx],                       \
 226                 __builtin_expect (cp[0], L'\1')==L'\0' && ch != '\0'))        \
 227           {                                                                   \
 228             /* This is an illegal character.  */                              \
 229             STANDARD_TO_LOOP_ERR_HANDLER (4);                                 \
 230           }                                                                   \
 231         else                                                                  \
 232           {                                                                   \
 233             if (curcs == sb)                                                  \
 234               {                                                               \
 235                 if (__glibc_unlikely (outptr + 1 > outend))                   \
 236                   {                                                           \
 237                     result = __GCONV_FULL_OUTPUT;                             \
 238                     break;                                                    \
 239                   }                                                           \
 240                 *outptr++ = SO;                                               \
 241                 curcs = db;                                                   \
 242               }                                                               \
 243                                                                               \
 244             if (__glibc_unlikely (outptr + 2 > outend))                       \
 245               {                                                               \
 246                 result = __GCONV_FULL_OUTPUT;                                 \
 247                 break;                                                        \
 248               }                                                               \
 249             *outptr++ = cp[0];                                                \
 250             *outptr++ = cp[1];                                                \
 251           }                                                                   \
 252       }                                                                       \
 253     else                                                                      \
 254       {                                                                       \
 255         if (curcs == db)                                                      \
 256           {                                                                   \
 257             if (__glibc_unlikely (outptr + 1 > outend))                       \
 258               {                                                               \
 259                 result = __GCONV_FULL_OUTPUT;                                 \
 260                 break;                                                        \
 261               }                                                               \
 262             *outptr++ = SI;                                                   \
 263             curcs = sb;                                                       \
 264           }                                                                   \
 265                                                                               \
 266         if (__glibc_unlikely (outptr + 1 > outend))                           \
 267           {                                                                   \
 268             result = __GCONV_FULL_OUTPUT;                                     \
 269             break;                                                            \
 270           }                                                                   \
 271         *outptr++ = cp[0];                                                    \
 272       }                                                                       \
 273                                                                               \
 274     /* Now that we wrote the output increment the input pointer.  */          \
 275     inptr += 4;                                                               \
 276   }
 277 #define LOOP_NEED_FLAGS
 278 #define EXTRA_LOOP_DECLS        , int *curcsp
 279 #define INIT_PARAMS             int curcs = *curcsp & ~7
 280 #define REINIT_PARAMS           curcs = *curcsp & ~7
 281 #define UPDATE_PARAMS           *curcsp = curcs
 282 #include <iconv/loop.c>
 283
 284 /* Now define the toplevel functions.  */
 285 #include <iconv/skeleton.c>