* locales/iso14651_t1_common: Add support for Gurumukhi script.
[glibc.git] / iconvdata / unicode.c
blobdf452d2a1d27b18da061b484a154e6a3abcdc13a
1 /* Conversion module for Unicode
2 Copyright (C) 1999, 2000-2002 Free Software Foundation, Inc.
3 This file is part of the GNU C Library.
4 Contributed by Ulrich Drepper <drepper@cygnus.com>, 1999.
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.
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.
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. */
21 #include <byteswap.h>
22 #include <dlfcn.h>
23 #include <gconv.h>
24 #include <stddef.h>
25 #include <stdint.h>
26 #include <stdlib.h>
27 #include <string.h>
29 /* This is the Byte Order Mark character (BOM). */
30 #define BOM 0xfeff
31 /* And in the other endian format. */
32 #define BOM_OE 0xfffe
35 /* Definitions used in the body of the `gconv' function. */
36 #define FROM_LOOP from_unicode_loop
37 #define TO_LOOP to_unicode_loop
38 #define DEFINE_INIT 0
39 #define DEFINE_FINI 0
40 #define MIN_NEEDED_FROM 2
41 #define MIN_NEEDED_TO 4
42 #define FROM_DIRECTION (dir == from_unicode)
43 #define PREPARE_LOOP \
44 enum direction dir = ((struct unicode_data *) step->__data)->dir; \
45 int swap; \
46 if (FROM_DIRECTION) \
47 { \
48 if (data->__invocation_counter == 0) \
49 { \
50 /* We have to find out which byte order the file is encoded in. */ \
51 if (inptr + 2 > inend) \
52 return (inptr == inend \
53 ? __GCONV_EMPTY_INPUT : __GCONV_INCOMPLETE_INPUT); \
55 if (get16u (inptr) == BOM) \
56 /* Simply ignore the BOM character. */ \
57 *inptrp = inptr += 2; \
58 else if (get16u (inptr) == BOM_OE) \
59 { \
60 ((struct unicode_data *) step->__data)->swap = 1; \
61 *inptrp = inptr += 2; \
62 } \
63 } \
64 } \
65 else if (!data->__internal_use && data->__invocation_counter == 0) \
66 { \
67 /* Emit the Byte Order Mark. */ \
68 if (__builtin_expect (outbuf + 2 > outend, 0)) \
69 return __GCONV_FULL_OUTPUT; \
71 put16u (outbuf, BOM); \
72 outbuf += 2; \
73 } \
74 swap = ((struct unicode_data *) step->__data)->swap;
75 #define EXTRA_LOOP_ARGS , swap
78 /* Direction of the transformation. */
79 enum direction
81 illegal_dir,
82 to_unicode,
83 from_unicode
86 struct unicode_data
88 enum direction dir;
89 int swap;
93 extern int gconv_init (struct __gconv_step *step);
94 int
95 gconv_init (struct __gconv_step *step)
97 /* Determine which direction. */
98 struct unicode_data *new_data;
99 enum direction dir = illegal_dir;
100 int result;
102 if (strcmp (step->__from_name, "UNICODE//") == 0)
103 dir = from_unicode;
104 else
105 dir = to_unicode;
107 new_data = (struct unicode_data *) malloc (sizeof (struct unicode_data));
109 result = __GCONV_NOMEM;
110 if (new_data != NULL)
112 new_data->dir = dir;
113 new_data->swap = 0;
114 step->__data = new_data;
116 if (dir == from_unicode)
118 step->__min_needed_from = MIN_NEEDED_FROM;
119 step->__max_needed_from = MIN_NEEDED_FROM;
120 step->__min_needed_to = MIN_NEEDED_TO;
121 step->__max_needed_to = MIN_NEEDED_TO;
123 else
125 step->__min_needed_from = MIN_NEEDED_TO;
126 step->__max_needed_from = MIN_NEEDED_TO;
127 step->__min_needed_to = MIN_NEEDED_FROM;
128 step->__max_needed_to = MIN_NEEDED_FROM;
131 step->__stateful = 0;
133 result = __GCONV_OK;
136 return result;
140 extern void gconv_end (struct __gconv_step *data);
141 void
142 gconv_end (struct __gconv_step *data)
144 free (data->__data);
148 /* Convert from the internal (UCS4-like) format to UCS2. */
149 #define MIN_NEEDED_INPUT MIN_NEEDED_TO
150 #define MIN_NEEDED_OUTPUT MIN_NEEDED_FROM
151 #define LOOPFCT TO_LOOP
152 #define BODY \
154 uint32_t c = get32 (inptr); \
156 if (__builtin_expect (c >= 0x10000, 0)) \
158 UNICODE_TAG_HANDLER (c, 4); \
159 STANDARD_TO_LOOP_ERR_HANDLER (4); \
161 else if (__builtin_expect (c >= 0xd800 && c < 0xe000, 0)) \
163 /* Surrogate characters in UCS-4 input are not valid. \
164 We must catch this, because the UCS-2 output might be \
165 interpreted as UTF-16 by other programs. If we let \
166 surrogates pass through, attackers could make a security \
167 hole exploit by synthesizing any desired plane 1-16 \
168 character. */ \
169 result = __GCONV_ILLEGAL_INPUT; \
170 if (! ignore_errors_p ()) \
171 break; \
172 inptr += 4; \
173 ++*irreversible; \
174 continue; \
176 else \
178 put16 (outptr, c); \
179 outptr += 2; \
182 inptr += 4; \
184 #define LOOP_NEED_FLAGS
185 #define EXTRA_LOOP_DECLS \
186 , int swap
187 #include <iconv/loop.c>
190 /* Convert from UCS2 to the internal (UCS4-like) format. */
191 #define MIN_NEEDED_INPUT MIN_NEEDED_FROM
192 #define MIN_NEEDED_OUTPUT MIN_NEEDED_TO
193 #define LOOPFCT FROM_LOOP
194 #define BODY \
196 uint16_t u1 = get16 (inptr); \
198 if (swap) \
199 u1 = bswap_16 (u1); \
201 if (__builtin_expect (u1 >= 0xd800 && u1 < 0xe000, 0)) \
203 /* Surrogate characters in UCS-2 input are not valid. Reject \
204 them. (Catching this here is not security relevant.) */ \
205 STANDARD_FROM_LOOP_ERR_HANDLER (2); \
208 put32 (outptr, u1); \
210 inptr += 2; \
211 outptr += 4; \
213 #define LOOP_NEED_FLAGS
214 #define EXTRA_LOOP_DECLS \
215 , int swap
216 #include <iconv/loop.c>
219 /* Now define the toplevel functions. */
220 #include <iconv/skeleton.c>