iconvdata/utf-32.c

   1 /* Conversion module for UTF-32.
   2    Copyright (C) 1999-2013 Free Software Foundation, Inc.
   3    This file is part of the GNU C Library.
   4
   5    The GNU C Library is free software; you can redistribute it and/or
   6    modify it under the terms of the GNU Lesser General Public
   7    License as published by the Free Software Foundation; either
   8    version 2.1 of the License, or (at your option) any later version.
   9
  10    The GNU C Library is distributed in the hope that it will be useful,
  11    but WITHOUT ANY WARRANTY; without even the implied warranty of
  12    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  13    Lesser General Public License for more details.
  14
  15    You should have received a copy of the GNU Lesser General Public
  16    License along with the GNU C Library; if not, see
  17    <http://www.gnu.org/licenses/>.  */
  18
  19 #include <byteswap.h>
  20 #include <dlfcn.h>
  21 #include <gconv.h>
  22 #include <stddef.h>
  23 #include <stdint.h>
  24 #include <stdlib.h>
  25 #include <string.h>
  26
  27 /* This is the Byte Order Mark character (BOM).  */
  28 #define BOM     0x0000feffu
  29 /* And in the other byte order.  */
  30 #define BOM_OE  0xfffe0000u
  31
  32
  33 /* Definitions used in the body of the `gconv' function.  */
  34 #define FROM_LOOP               from_utf32_loop
  35 #define TO_LOOP                 to_utf32_loop
  36 #define DEFINE_INIT             0
  37 #define DEFINE_FINI             0
  38 #define MIN_NEEDED_FROM         4
  39 #define MIN_NEEDED_TO           4
  40 #define FROM_DIRECTION          (dir == from_utf32)
  41 #define PREPARE_LOOP \
  42   enum direction dir = ((struct utf32_data *) step->__data)->dir;             \
  43   enum variant var = ((struct utf32_data *) step->__data)->var;               \
  44   int swap;                                                                   \
  45   if (FROM_DIRECTION && var == UTF_32)                                        \
  46     {                                                                         \
  47       if (__builtin_expect (data->__invocation_counter == 0, 0))              \
  48         {                                                                     \
  49           /* We have to find out which byte order the file is encoded in.  */ \
  50           if (inptr + 4 > inend)                                              \
  51             return (inptr == inend                                            \
  52                     ? __GCONV_EMPTY_INPUT : __GCONV_INCOMPLETE_INPUT);        \
  53                                                                               \
  54           if (get32u (inptr) == BOM)                                          \
  55             /* Simply ignore the BOM character.  */                           \
  56             *inptrp = inptr += 4;                                             \
  57           else if (get32u (inptr) == BOM_OE)                                  \
  58             {                                                                 \
  59               data->__flags |= __GCONV_SWAP;                                  \
  60               *inptrp = inptr += 4;                                           \
  61             }                                                                 \
  62         }                                                                     \
  63     }                                                                         \
  64   else if (!FROM_DIRECTION && var == UTF_32 && !data->__internal_use          \
  65            && data->__invocation_counter == 0)                                \
  66     {                                                                         \
  67       /* Emit the Byte Order Mark.  */                                        \
  68       if (__builtin_expect (outbuf + 4 > outend, 0))                          \
  69         return __GCONV_FULL_OUTPUT;                                           \
  70                                                                               \
  71       put32u (outbuf, BOM);                                                   \
  72       outbuf += 4;                                                            \
  73     }                                                                         \
  74   else if (__builtin_expect (data->__invocation_counter == 0, 0)              \
  75            && ((var == UTF_32LE && BYTE_ORDER == BIG_ENDIAN)                  \
  76                || (var == UTF_32BE && BYTE_ORDER == LITTLE_ENDIAN)))          \
  77     data->__flags |= __GCONV_SWAP;                                            \
  78   swap = data->__flags & __GCONV_SWAP;
  79 #define EXTRA_LOOP_ARGS         , var, swap
  80
  81
  82 /* Direction of the transformation.  */
  83 enum direction
  84 {
  85   illegal_dir,
  86   to_utf32,
  87   from_utf32
  88 };
  89
  90 enum variant
  91 {
  92   illegal_var,
  93   UTF_32,
  94   UTF_32LE,
  95   UTF_32BE
  96 };
  97
  98 struct utf32_data
  99 {
 100   enum direction dir;
 101   enum variant var;
 102 };
 103
 104
 105 extern int gconv_init (struct __gconv_step *step);
 106 int
 107 gconv_init (struct __gconv_step *step)
 108 {
 109   /* Determine which direction.  */
 110   struct utf32_data *new_data;
 111   enum direction dir = illegal_dir;
 112   enum variant var = illegal_var;
 113   int result;
 114
 115   if (__strcasecmp (step->__from_name, "UTF-32//") == 0)
 116     {
 117       dir = from_utf32;
 118       var = UTF_32;
 119     }
 120   else if (__strcasecmp (step->__to_name, "UTF-32//") == 0)
 121     {
 122       dir = to_utf32;
 123       var = UTF_32;
 124     }
 125   else if (__strcasecmp (step->__from_name, "UTF-32BE//") == 0)
 126     {
 127       dir = from_utf32;
 128       var = UTF_32BE;
 129     }
 130   else if (__strcasecmp (step->__to_name, "UTF-32BE//") == 0)
 131     {
 132       dir = to_utf32;
 133       var = UTF_32BE;
 134     }
 135   else if (__strcasecmp (step->__from_name, "UTF-32LE//") == 0)
 136     {
 137       dir = from_utf32;
 138       var = UTF_32LE;
 139     }
 140   else if (__strcasecmp (step->__to_name, "UTF-32LE//") == 0)
 141     {
 142       dir = to_utf32;
 143       var = UTF_32LE;
 144     }
 145
 146   result = __GCONV_NOCONV;
 147   if (__builtin_expect (dir, to_utf32) != illegal_dir)
 148     {
 149       new_data = (struct utf32_data *) malloc (sizeof (struct utf32_data));
 150
 151       result = __GCONV_NOMEM;
 152       if (new_data != NULL)
 153         {
 154           new_data->dir = dir;
 155           new_data->var = var;
 156           step->__data = new_data;
 157
 158           if (dir == from_utf32)
 159             {
 160               step->__min_needed_from = MIN_NEEDED_FROM;
 161               step->__max_needed_from = MIN_NEEDED_FROM;
 162               step->__min_needed_to = MIN_NEEDED_TO;
 163               step->__max_needed_to = MIN_NEEDED_TO;
 164             }
 165           else
 166             {
 167               step->__min_needed_from = MIN_NEEDED_TO;
 168               step->__max_needed_from = MIN_NEEDED_TO;
 169               step->__min_needed_to = MIN_NEEDED_FROM;
 170               step->__max_needed_to = MIN_NEEDED_FROM;
 171             }
 172
 173           step->__stateful = 0;
 174
 175           result = __GCONV_OK;
 176         }
 177     }
 178
 179   return result;
 180 }
 181
 182
 183 extern void gconv_end (struct __gconv_step *data);
 184 void
 185 gconv_end (struct __gconv_step *data)
 186 {
 187   free (data->__data);
 188 }
 189
 190
 191 /* Convert from the internal (UCS4-like) format to UTF-32.  */
 192 #define MIN_NEEDED_INPUT        MIN_NEEDED_TO
 193 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_FROM
 194 #define LOOPFCT                 TO_LOOP
 195 #define BODY \
 196   {                                                                           \
 197     uint32_t c = get32 (inptr);                                               \
 198                                                                               \
 199     if (__builtin_expect (c >= 0x110000, 0))                                  \
 200       {                                                                       \
 201         STANDARD_TO_LOOP_ERR_HANDLER (4);                                     \
 202       }                                                                       \
 203     else if (__builtin_expect (c >= 0xd800 && c < 0xe000, 0))                 \
 204       {                                                                       \
 205         /* Surrogate characters in UCS-4 input are not valid.                 \
 206            We must catch this.  If we let surrogates pass through,            \
 207            attackers could make a security hole exploit by                    \
 208            generating "irregular UTF-32" sequences.  */                       \
 209         result = __GCONV_ILLEGAL_INPUT;                                       \
 210         if (! ignore_errors_p ())                                             \
 211           break;                                                              \
 212         inptr += 4;                                                           \
 213         ++*irreversible;                                                      \
 214         continue;                                                             \
 215       }                                                                       \
 216                                                                               \
 217     if (swap)                                                                 \
 218       c = bswap_32 (c);                                                       \
 219     put32 (outptr, c);                                                        \
 220                                                                               \
 221     outptr += 4;                                                              \
 222     inptr += 4;                                                               \
 223   }
 224 #define LOOP_NEED_FLAGS
 225 #define EXTRA_LOOP_DECLS \
 226         , enum variant var, int swap
 227 #include <iconv/loop.c>
 228
 229
 230 /* Convert from UTF-32 to the internal (UCS4-like) format.  */
 231 #define MIN_NEEDED_INPUT        MIN_NEEDED_FROM
 232 #define MIN_NEEDED_OUTPUT       MIN_NEEDED_TO
 233 #define LOOPFCT                 FROM_LOOP
 234 #define BODY \
 235   {                                                                           \
 236     uint32_t u1 = get32 (inptr);                                              \
 237                                                                               \
 238     if (swap)                                                                 \
 239       u1 = bswap_32 (u1);                                                     \
 240                                                                               \
 241     if (__builtin_expect (u1 >= 0x110000, 0))                                 \
 242       {                                                                       \
 243         /* This is illegal.  */                                               \
 244         STANDARD_FROM_LOOP_ERR_HANDLER (4);                                   \
 245       }                                                                       \
 246                                                                               \
 247     put32 (outptr, u1);                                                       \
 248     inptr += 4;                                                               \
 249     outptr += 4;                                                              \
 250   }
 251 #define LOOP_NEED_FLAGS
 252 #define EXTRA_LOOP_DECLS \
 253         , enum variant var, int swap
 254 #include <iconv/loop.c>
 255
 256
 257 /* Now define the toplevel functions.  */
 258 #include <iconv/skeleton.c>