1 /* Conversion module for Unicode
2 Copyright (C) 1999-2023 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 <https://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 0xfeff
29 /* And in the other endian format. */
30 #define BOM_OE 0xfffe
31
32
33 /* Definitions used in the body of the `gconv' function. */
34 #define FROM_LOOP from_unicode_loop
35 #define TO_LOOP to_unicode_loop
36 #define DEFINE_INIT 0
37 #define DEFINE_FINI 0
38 #define MIN_NEEDED_FROM 2
39 #define MIN_NEEDED_TO 4
40 #define ONE_DIRECTION 0
41 #define FROM_DIRECTION (dir == from_unicode)
42 #define PREPARE_LOOP \
43 enum direction dir = ((struct unicode_data *) step->__data)->dir; \
44 int swap; \
45 if (FROM_DIRECTION) \
46 { \
47 if (data->__invocation_counter == 0) \
48 { \
49 /* We have to find out which byte order the file is encoded in. */ \
50 if (inptr + 2 > inend) \
51 return (inptr == inend \
52 ? __GCONV_EMPTY_INPUT : __GCONV_INCOMPLETE_INPUT); \
53 \
54 if (get16 (inptr) == BOM) \
55 /* Simply ignore the BOM character. */ \
56 *inptrp = inptr += 2; \
57 else if (get16 (inptr) == BOM_OE) \
58 { \
59 data->__flags |= __GCONV_SWAP; \
60 *inptrp = inptr += 2; \
61 } \
62 } \
63 } \
64 else if (!data->__internal_use && data->__invocation_counter == 0) \
65 { \
66 /* Emit the Byte Order Mark. */ \
67 if (__glibc_unlikely (outbuf + 2 > outend)) \
68 return __GCONV_FULL_OUTPUT; \
69 \
70 put16 (outbuf, BOM); \
71 outbuf += 2; \
72 } \
73 swap = data->__flags & __GCONV_SWAP;
74 #define EXTRA_LOOP_ARGS , swap
75
76
77 /* Direction of the transformation. */
78 enum direction
79 {
80 illegal_dir,
81 to_unicode,
82 from_unicode
83 };
84
85 struct unicode_data
86 {
87 enum direction dir;
88 };
89
90
91 extern int gconv_init (struct __gconv_step *step);
92 int
93 gconv_init (struct __gconv_step *step)
94 {
95 /* Determine which direction. */
96 struct unicode_data *new_data;
97 enum direction dir = illegal_dir;
98 int result;
99
100 if (strcmp (step->__from_name, "UNICODE//") == 0)
101 dir = from_unicode;
102 else
103 dir = to_unicode;
104
105 new_data = (struct unicode_data *) malloc (sizeof (struct unicode_data));
106
107 result = __GCONV_NOMEM;
108 if (new_data != NULL)
109 {
110 new_data->dir = dir;
111 step->__data = new_data;
112
113 if (dir == from_unicode)
114 {
115 step->__min_needed_from = MIN_NEEDED_FROM;
116 step->__max_needed_from = MIN_NEEDED_FROM;
117 step->__min_needed_to = MIN_NEEDED_TO;
118 step->__max_needed_to = MIN_NEEDED_TO;
119 }
120 else
121 {
122 step->__min_needed_from = MIN_NEEDED_TO;
123 step->__max_needed_from = MIN_NEEDED_TO;
124 step->__min_needed_to = MIN_NEEDED_FROM;
125 step->__max_needed_to = MIN_NEEDED_FROM;
126 }
127
128 step->__stateful = 0;
129
130 result = __GCONV_OK;
131 }
132
133 return result;
134 }
135
136
137 extern void gconv_end (struct __gconv_step *data);
138 void
139 gconv_end (struct __gconv_step *data)
140 {
141 free (data->__data);
142 }
143
144
145 /* Convert from the internal (UCS4-like) format to UCS2. */
146 #define MIN_NEEDED_INPUT MIN_NEEDED_TO
147 #define MIN_NEEDED_OUTPUT MIN_NEEDED_FROM
148 #define LOOPFCT TO_LOOP
149 #define BODY \
150 { \
151 uint32_t c = get32 (inptr); \
152 \
153 if (__glibc_unlikely (c >= 0x10000)) \
154 { \
155 UNICODE_TAG_HANDLER (c, 4); \
156 STANDARD_TO_LOOP_ERR_HANDLER (4); \
157 } \
158 else if (__glibc_unlikely (c >= 0xd800 && c < 0xe000)) \
159 { \
160 /* Surrogate characters in UCS-4 input are not valid. \
161 We must catch this, because the UCS-2 output might be \
162 interpreted as UTF-16 by other programs. If we let \
163 surrogates pass through, attackers could make a security \
164 hole exploit by synthesizing any desired plane 1-16 \
165 character. */ \
166 result = __GCONV_ILLEGAL_INPUT; \
167 if (! ignore_errors_p ()) \
168 break; \
169 inptr += 4; \
170 ++*irreversible; \
171 continue; \
172 } \
173 else \
174 { \
175 put16 (outptr, c); \
176 outptr += 2; \
177 } \
178 \
179 inptr += 4; \
180 }
181 #define LOOP_NEED_FLAGS
182 #define EXTRA_LOOP_DECLS \
183 , int swap
184 #include <iconv/loop.c>
185
186
187 /* Convert from UCS2 to the internal (UCS4-like) format. */
188 #define MIN_NEEDED_INPUT MIN_NEEDED_FROM
189 #define MIN_NEEDED_OUTPUT MIN_NEEDED_TO
190 #define LOOPFCT FROM_LOOP
191 #define BODY \
192 { \
193 uint16_t u1 = get16 (inptr); \
194 \
195 if (swap) \
196 u1 = bswap_16 (u1); \
197 \
198 if (__glibc_unlikely (u1 >= 0xd800 && u1 < 0xe000)) \
199 { \
200 /* Surrogate characters in UCS-2 input are not valid. Reject \
201 them. (Catching this here is not security relevant.) */ \
202 STANDARD_FROM_LOOP_ERR_HANDLER (2); \
203 } \
204 \
205 put32 (outptr, u1); \
206 \
207 inptr += 2; \
208 outptr += 4; \
209 }
210 #define LOOP_NEED_FLAGS
211 #define EXTRA_LOOP_DECLS \
212 , int swap
213 #include <iconv/loop.c>
214
215
216 /* Now define the toplevel functions. */
217 #include <iconv/skeleton.c>