iconvdata/utf-16.c

   1 /* Conversion module for UTF-16.
   2    Copyright (C) 1999, 2000 Free Software Foundation, Inc.
   3    This file is part of the GNU C Library.
   4    Contributed by Ulrich Drepper <drepper@cygnus.com>, 1999.
   5
   6    The GNU C Library is free software; you can redistribute it and/or
   7    modify it under the terms of the GNU Library General Public License as
   8    published by the Free Software Foundation; either version 2 of the
   9    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    Library General Public License for more details.
  15
  16    You should have received a copy of the GNU Library General Public
  17    License along with the GNU C Library; see the file COPYING.LIB.  If not,
  18    write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  19    Boston, MA 02111-1307, USA.  */
  20
  21 #include <byteswap.h>
  22 #include <gconv.h>
  23 #include <stddef.h>
  24 #include <stdint.h>
  25 #include <stdlib.h>
  26 #include <string.h>
  27
  28 /* This is the Byte Order Mark character (BOM).  */
  29 #define BOM     0xfeff
  30 /* And in the other byte order.  */
  31 #define BOM_OE  0xfffe
  32
  33
  34 /* Definitions used in the body of the `gconv' function.  */
  35 #define FROM_LOOP               from_utf16_loop
  36 #define TO_LOOP                 to_utf16_loop
  37 #define DEFINE_INIT             0
  38 #define DEFINE_FINI             0
  39 #define MIN_NEEDED_FROM         2
  40 #define MAX_NEEDED_FROM         4
  41 #define MIN_NEEDED_TO           4
  42 #define FROM_DIRECTION          (dir == from_utf16)
  43 #define PREPARE_LOOP \
  44   enum direction dir = ((struct utf16_data *) step->__data)->dir;             \
  45   enum variant var = ((struct utf16_data *) step->__data)->var;               \
  46   int swap = ((struct utf16_data *) step->__data)->swap;                      \
  47   if (FROM_DIRECTION || var == UTF_16)                                        \
  48     {                                                                         \
  49       if (data->__invocation_counter == 0)                                    \
  50         {                                                                     \
  51           /* We have to find out which byte order the file is encoded in.  */ \
  52           if (inptr + 2 > inend)                                              \
  53             return __GCONV_EMPTY_INPUT;                                       \
  54                                                                               \
  55           if (get16u (inptr) == BOM)                                          \
  56             /* Simply ignore the BOM character.  */                           \
  57             inptr += 2;                                                       \
  58           else if (get16u (inptr) == BOM_OE)                                  \
  59             {                                                                 \
  60               ((struct utf16_data *) step->__data)->swap = 1;                 \
  61               inptr += 2;                                                     \
  62             }                                                                 \
  63         }                                                                     \
  64     }                                                                         \
  65   else if (!FROM_DIRECTION && var == UTF_16 && !data->__internal_use          \
  66            && data->__invocation_counter == 0)                                \
  67     {                                                                         \
  68       /* Emit the Byte Order Mark.  */                                        \
  69       if (outbuf + 2 > outend)                                                \
  70         return __GCONV_FULL_OUTPUT;                                           \
  71                                                                               \
  72       put16u (outbuf, BOM);                                                   \
  73       outbuf += 2;                                                            \
  74     }
  75 #define EXTRA_LOOP_ARGS         , var, data, swap
  76
  77
  78 /* Direction of the transformation.  */
  79 enum direction
  80 {
  81   illegal_dir,
  82   to_utf16,
  83   from_utf16
  84 };
  85
  86 enum variant
  87 {
  88   illegal_var,
  89   UTF_16,
  90   UTF_16LE,
  91   UTF_16BE
  92 };
  93
  94 struct utf16_data
  95 {
  96   enum direction dir;
  97   enum variant var;
  98   int swap;
  99 };
 100
 101
 102 int
 103 gconv_init (struct __gconv_step *step)
 104 {
 105   /* Determine which direction.  */
 106   struct utf16_data *new_data;
 107   enum direction dir = illegal_dir;
 108   enum variant var = illegal_var;
 109   int result;
 110
 111   if (__strcasecmp (step->__from_name, "UTF-16") == 0)
 112     {
 113       dir = from_utf16;
 114       var = UTF_16;
 115     }
 116   else if (__strcasecmp (step->__to_name, "UTF-16") == 0)
 117     {
 118       dir = to_utf16;
 119       var = UTF_16;
 120     }
 121   else if (__strcasecmp (step->__from_name, "UTF-16BE") == 0)
 122     {
 123       dir = from_utf16;
 124       var = UTF_16BE;
 125     }
 126   else if (__strcasecmp (step->__to_name, "UTF-16BE") == 0)
 127     {
 128       dir = to_utf16;
 129       var = UTF_16BE;
 130     }
 131   else if (__strcasecmp (step->__from_name, "UTF-16LE") == 0)
 132     {
 133       dir = from_utf16;
 134       var = UTF_16LE;
 135     }
 136   else if (__strcasecmp (step->__to_name, "UTF-16LE") == 0)
 137     {
 138       dir = to_utf16;
 139       var = UTF_16LE;
 140     }
 141
 142   result = __GCONV_NOCONV;
 143   if (dir != illegal_dir)
 144     {
 145       new_data = (struct utf16_data *) malloc (sizeof (struct utf16_data));
 146
 147       result = __GCONV_NOMEM;
 148       if (new_data != NULL)
 149         {
 150           new_data->dir = dir;
 151           new_data->var = var;
 152           new_data->swap = ((var == UTF_16LE && BYTE_ORDER == BIG_ENDIAN)
 153                             || (var == UTF_16BE
 154                                 && BYTE_ORDER == LITTLE_ENDIAN));
 155           step->__data = new_data;
 156
 157           if (dir == from_utf16)
 158             {
 159               step->__min_needed_from = MIN_NEEDED_FROM;
 160               step->__max_needed_from = MIN_NEEDED_FROM;
 161               step->__min_needed_to = MIN_NEEDED_TO;
 162               step->__max_needed_to = MIN_NEEDED_TO;
 163             }
 164           else
 165             {
 166               step->__min_needed_from = MIN_NEEDED_TO;
 167               step->__max_needed_from = MIN_NEEDED_TO;
 168               step->__min_needed_to = MIN_NEEDED_FROM;
 169               step->__max_needed_to = MIN_NEEDED_FROM;
 170             }
 171
 172           step->__stateful = 0;
 173
 174           result = __GCONV_OK;
 175         }
 176     }
 177
 178   return result;
 179 }
 180
 181
 182 void
 183 gconv_end (struct __gconv_step *data)
 184 {
 185   free (data->__data);
 186 }
 187
 188
 189 /* Convert from the internal (UCS4-like) format to UTF-16.  */
 190 #define MIN_NEEDED_INPUT        MIN_NEEDED_TO
 191 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_FROM
 192 #define MAX_NEEDED_OUTPUT       MAX_NEEDED_FROM
 193 #define LOOPFCT                 TO_LOOP
 194 #define BODY \
 195   {                                                                           \
 196     uint32_t c = get32 (inptr);                                               \
 197                                                                               \
 198     if (swap)                                                                 \
 199       {                                                                       \
 200         if (c >= 0x10000)                                                     \
 201           {                                                                   \
 202             if (c >= 0x110000)                                                \
 203               {                                                               \
 204                 result = __GCONV_ILLEGAL_INPUT;                               \
 205                 break;                                                        \
 206               }                                                               \
 207                                                                               \
 208             /* Generate a surrogate character.  */                            \
 209             if (NEED_LENGTH_TEST && outptr + 4 > outend)                      \
 210               {                                                               \
 211                 /* Overflow in the output buffer.  */                         \
 212                 result = __GCONV_FULL_OUTPUT;                                 \
 213                 break;                                                        \
 214               }                                                               \
 215                                                                               \
 216             put16 (outptr, bswap_16 (0xd7c0 + (c >> 10)));                    \
 217             outptr += 2;                                                      \
 218             put16 (outptr, bswap_16 (0xdc00 + (c & 0x3ff)));                  \
 219           }                                                                   \
 220         else                                                                  \
 221           put16 (outptr, bswap_16 (c));                                       \
 222       }                                                                       \
 223     else                                                                      \
 224       {                                                                       \
 225         if (c >= 0x10000)                                                     \
 226           {                                                                   \
 227             if (c >= 0x110000)                                                \
 228               {                                                               \
 229                 result = __GCONV_ILLEGAL_INPUT;                               \
 230                 break;                                                        \
 231               }                                                               \
 232                                                                               \
 233             /* Generate a surrogate character.  */                            \
 234             if (NEED_LENGTH_TEST && outptr + 4 > outend)                      \
 235               {                                                               \
 236                 /* Overflow in the output buffer.  */                         \
 237                 result = __GCONV_FULL_OUTPUT;                                 \
 238                 break;                                                        \
 239               }                                                               \
 240                                                                               \
 241             put16 (outptr, 0xd7c0 + (c >> 10));                               \
 242             outptr += 2;                                                      \
 243             put16 (outptr, 0xdc00 + (c & 0x3ff));                             \
 244           }                                                                   \
 245         else                                                                  \
 246           put16 (outptr, c);                                                  \
 247       }                                                                       \
 248     outptr += 2;                                                              \
 249     inptr += 4;                                                               \
 250   }
 251 #define EXTRA_LOOP_DECLS \
 252         , enum variant var, struct __gconv_step_data *step_data, int swap
 253 #include <iconv/loop.c>
 254
 255
 256 /* Convert from UTF-16 to the internal (UCS4-like) format.  */
 257 #define MIN_NEEDED_INPUT        MIN_NEEDED_FROM
 258 #define MAX_NEEDED_INPUT        MAX_NEEDED_FROM
 259 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_TO
 260 #define LOOPFCT                 FROM_LOOP
 261 #define BODY \
 262   {                                                                           \
 263     uint16_t u1 = get16 (inptr);                                              \
 264                                                                               \
 265     if (swap)                                                                 \
 266       {                                                                       \
 267         u1 = bswap_16 (u1);                                                   \
 268                                                                               \
 269         if (u1 < 0xd800 || u1 > 0xdfff)                                       \
 270           {                                                                   \
 271             /* No surrogate.  */                                              \
 272             put32 (outptr, u1);                               \
 273             inptr += 2;                                                       \
 274           }                                                                   \
 275         else                                                                  \
 276           {                                                                   \
 277             uint16_t u2;                                                      \
 278                                                                               \
 279             /* It's a surrogate character.  At least the first word says      \
 280                it is.  */                                                     \
 281             if (NEED_LENGTH_TEST && inptr + 4 > inend)                        \
 282               {                                                               \
 283                 /* We don't have enough input for another complete input      \
 284                    character.  */                                             \
 285                 result = __GCONV_INCOMPLETE_INPUT;                            \
 286                 break;                                                        \
 287               }                                                               \
 288                                                                               \
 289             inptr += 2;                                                       \
 290             u2 = bswap_16 (get16 (inptr));                                    \
 291             if (u2 < 0xdc00 || u2 >= 0xdfff)                                  \
 292               {                                                               \
 293                 /* This is no valid second word for a surrogate.  */          \
 294                 result = __GCONV_ILLEGAL_INPUT;                               \
 295                 inptr -= 2;                                                   \
 296                 break;                                                        \
 297               }                                                               \
 298                                                                               \
 299             put32 (outptr, ((u1 - 0xd7c0) << 10) + (u2 - 0xdc00));            \
 300             inptr += 2;                                                       \
 301           }                                                                   \
 302       }                                                                       \
 303     else                                                                      \
 304       {                                                                       \
 305         if (u1 < 0xd800 || u1 > 0xdfff)                                       \
 306           {                                                                   \
 307             /* No surrogate.  */                                              \
 308             put32 (outptr, u1);                                               \
 309             inptr += 2;                                                       \
 310           }                                                                   \
 311         else                                                                  \
 312           {                                                                   \
 313             uint16_t u2;                                                      \
 314                                                                               \
 315             /* It's a surrogate character.  At least the first word says      \
 316                it is.  */                                                     \
 317             if (NEED_LENGTH_TEST && inptr + 4 > inend)                        \
 318               {                                                               \
 319                 /* We don't have enough input for another complete input      \
 320                    character.  */                                             \
 321                 result = __GCONV_INCOMPLETE_INPUT;                            \
 322                 break;                                                        \
 323               }                                                               \
 324                                                                               \
 325             inptr += 2;                                                       \
 326             u2 = get16 (inptr);                                               \
 327             if (u2 < 0xdc00 || u2 >= 0xdfff)                                  \
 328               {                                                               \
 329                 /* This is no valid second word for a surrogate.  */          \
 330                 result = __GCONV_ILLEGAL_INPUT;                               \
 331                 inptr -= 2;                                                   \
 332                 break;                                                        \
 333               }                                                               \
 334                                                                               \
 335             put32 (outptr, ((u1 - 0xd7c0) << 10) + (u2 - 0xdc00));            \
 336             inptr += 2;                                                       \
 337           }                                                                   \
 338       }                                                                       \
 339     outptr += 4;                                                              \
 340   }
 341 #define EXTRA_LOOP_DECLS \
 342         , enum variant var, struct __gconv_step_data *step_data, int swap
 343 #include <iconv/loop.c>
 344
 345
 346 /* Now define the toplevel functions.  */
 347 #include <iconv/skeleton.c>