iconvdata/ibm937.c

   1 /* Conversion from and to IBM937.
   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@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    <https://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 (redundant escape okay).  */   \
 113         curcs = db;                                                           \
 114         ++inptr;                                                              \
 115         continue;                                                             \
 116       }                                                                       \
 117     else if (__builtin_expect (ch, 0) == SI)                                  \
 118       {                                                                       \
 119         /* Shift IN, change to SBCS converter (redundant escape okay).  */    \
 120         curcs = sb;                                                           \
 121         ++inptr;                                                              \
 122         continue;                                                             \
 123       }                                                                       \
 124                                                                               \
 125     if (curcs == sb)                                                          \
 126       {                                                                       \
 127         /* Use the IBM937 table for single byte.  */                          \
 128         res = __ibm937sb_to_ucs4[ch];                                         \
 129         if (__builtin_expect (res, L'\1') == L'\0' && ch != '\0')             \
 130           {                                                                   \
 131             /* This is an illegal character.  */                              \
 132             STANDARD_FROM_LOOP_ERR_HANDLER (1);                               \
 133           }                                                                   \
 134         else                                                                  \
 135           {                                                                   \
 136             put32 (outptr, res);                                              \
 137             outptr += 4;                                                      \
 138           }                                                                   \
 139         ++inptr;                                                              \
 140       }                                                                       \
 141     else                                                                      \
 142       {                                                                       \
 143         const struct gap *rp2 = __ibm937db_to_ucs4_idx;                       \
 144                                                                               \
 145         assert (curcs == db);                                                 \
 146                                                                               \
 147         /* Use the IBM937 table for double byte.  */                          \
 148         if (__glibc_unlikely (inptr + 1 >= inend))                            \
 149           {                                                                   \
 150             /* The second character is not available.                         \
 151                Store the intermediate result. */                              \
 152             result = __GCONV_INCOMPLETE_INPUT;                                \
 153             break;                                                            \
 154           }                                                                   \
 155                                                                               \
 156         ch = (ch * 0x100) + inptr[1];                                         \
 157         while (ch > rp2->end)                                                 \
 158           ++rp2;                                                              \
 159                                                                               \
 160         if (__builtin_expect (rp2->start == 0xffff, 0)                        \
 161             || __builtin_expect (ch < rp2->start, 0)                          \
 162             || (res = __ibm937db_to_ucs4[ch + rp2->idx],                      \
 163                 __builtin_expect (res, L'\1') == L'\0' && ch != '\0'))        \
 164           {                                                                   \
 165             /* This is an illegal character.  */                              \
 166             STANDARD_FROM_LOOP_ERR_HANDLER (2);                               \
 167           }                                                                   \
 168         else                                                                  \
 169           {                                                                   \
 170             put32 (outptr, res);                                              \
 171             outptr += 4;                                                      \
 172           }                                                                   \
 173         inptr += 2;                                                           \
 174       }                                                                       \
 175   }
 176 #define LOOP_NEED_FLAGS
 177 #define EXTRA_LOOP_DECLS        , int *curcsp
 178 #define INIT_PARAMS             int curcs = *curcsp & ~7
 179 #define UPDATE_PARAMS           *curcsp = curcs
 180 #include <iconv/loop.c>
 181
 182 /* Next, define the other direction.  */
 183 #define MIN_NEEDED_INPUT        TO_LOOP_MIN_NEEDED_FROM
 184 #define MAX_NEEDED_INPUT        TO_LOOP_MAX_NEEDED_FROM
 185 #define MIN_NEEDED_OUTPUT       TO_LOOP_MIN_NEEDED_TO
 186 #define MAX_NEEDED_OUTPUT       TO_LOOP_MAX_NEEDED_TO
 187 #define LOOPFCT                 TO_LOOP
 188 #define BODY \
 189   {                                                                           \
 190     uint32_t ch = get32 (inptr);                                              \
 191     const struct gap *rp1 = __ucs4_to_ibm937sb_idx;                           \
 192     const struct gap *rp2 = __ucs4_to_ibm937db_idx;                           \
 193     const char *cp;                                                           \
 194                                                                               \
 195     if (__glibc_unlikely (ch >= 0xffff))                                      \
 196       {                                                                       \
 197         UNICODE_TAG_HANDLER (ch, 4);                                          \
 198                                                                               \
 199         STANDARD_TO_LOOP_ERR_HANDLER (4);                                     \
 200       }                                                                       \
 201                                                                               \
 202     while (ch > rp1->end)                                                     \
 203       ++rp1;                                                                  \
 204                                                                               \
 205     /* Use the UCS4 table for single byte.  */                                \
 206     if (__builtin_expect (ch < rp1->start, 0)                                 \
 207         || (cp = __ucs4_to_ibm937sb[ch + rp1->idx],                           \
 208             __builtin_expect (cp[0], L'\1') == L'\0' && ch != '\0'))          \
 209       {                                                                       \
 210         /* Use the UCS4 table for double byte. */                             \
 211         while (ch > rp2->end)                                                 \
 212           ++rp2;                                                              \
 213                                                                               \
 214         if (__builtin_expect (ch < rp2->start, 0)                             \
 215             || (cp = __ucs4_to_ibm937db[ch + rp2->idx],                       \
 216                 __builtin_expect (cp[0], L'\1')==L'\0' && ch != '\0'))        \
 217           {                                                                   \
 218             /* This is an illegal character.  */                              \
 219             STANDARD_TO_LOOP_ERR_HANDLER (4);                                 \
 220           }                                                                   \
 221         else                                                                  \
 222           {                                                                   \
 223             if (curcs == sb)                                                  \
 224               {                                                               \
 225                 if (__glibc_unlikely (outptr + 1 > outend))                   \
 226                   {                                                           \
 227                     result = __GCONV_FULL_OUTPUT;                             \
 228                     break;                                                    \
 229                   }                                                           \
 230                 *outptr++ = SO;                                               \
 231                 curcs = db;                                                   \
 232               }                                                               \
 233                                                                               \
 234             if (__glibc_unlikely (outptr + 2 > outend))                       \
 235               {                                                               \
 236                 result = __GCONV_FULL_OUTPUT;                                 \
 237                 break;                                                        \
 238               }                                                               \
 239             *outptr++ = cp[0];                                                \
 240             *outptr++ = cp[1];                                                \
 241           }                                                                   \
 242       }                                                                       \
 243     else                                                                      \
 244       {                                                                       \
 245         if (curcs == db)                                                      \
 246           {                                                                   \
 247             if (__glibc_unlikely (outptr + 1 > outend))                       \
 248               {                                                               \
 249                 result = __GCONV_FULL_OUTPUT;                                 \
 250                 break;                                                        \
 251               }                                                               \
 252             *outptr++ = SI;                                                   \
 253             curcs = sb;                                                       \
 254           }                                                                   \
 255                                                                               \
 256         if (__glibc_unlikely (outptr + 1 > outend))                           \
 257           {                                                                   \
 258             result = __GCONV_FULL_OUTPUT;                                     \
 259             break;                                                            \
 260           }                                                                   \
 261         *outptr++ = cp[0];                                                    \
 262       }                                                                       \
 263                                                                               \
 264     /* Now that we wrote the output increment the input pointer.  */          \
 265     inptr += 4;                                                               \
 266   }
 267 #define LOOP_NEED_FLAGS
 268 #define EXTRA_LOOP_DECLS        , int *curcsp
 269 #define INIT_PARAMS             int curcs = *curcsp & ~7
 270 #define REINIT_PARAMS           curcs = *curcsp & ~7
 271 #define UPDATE_PARAMS           *curcsp = curcs
 272 #include <iconv/loop.c>
 273
 274 /* Now define the toplevel functions.  */
 275 #include <iconv/skeleton.c>