1/* Conversion module for Unicode
2 Copyright (C) 1999-2021 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 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#include <byteswap.h>
21#include <dlfcn.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 endian format. */
31#define BOM_OE 0xfffe
32
33
34/* Definitions used in the body of the `gconv' function. */
35#define FROM_LOOP from_unicode_loop
36#define TO_LOOP to_unicode_loop
37#define DEFINE_INIT 0
38#define DEFINE_FINI 0
39#define MIN_NEEDED_FROM 2
40#define MIN_NEEDED_TO 4
41#define ONE_DIRECTION 0
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); \
54 \
55 if (get16u (inptr) == BOM) \
56 /* Simply ignore the BOM character. */ \
57 *inptrp = inptr += 2; \
58 else if (get16u (inptr) == BOM_OE) \
59 { \
60 data->__flags |= __GCONV_SWAP; \
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 (__glibc_unlikely (outbuf + 2 > outend)) \
69 return __GCONV_FULL_OUTPUT; \
70 \
71 put16u (outbuf, BOM); \
72 outbuf += 2; \
73 } \
74 swap = data->__flags & __GCONV_SWAP;
75#define EXTRA_LOOP_ARGS , swap
76
77
78/* Direction of the transformation. */
79enum direction
80{
81 illegal_dir,
82 to_unicode,
83 from_unicode
84};
85
86struct unicode_data
87{
88 enum direction dir;
89};
90
91
92extern int gconv_init (struct __gconv_step *step);
93int
94gconv_init (struct __gconv_step *step)
95{
96 /* Determine which direction. */
97 struct unicode_data *new_data;
98 enum direction dir = illegal_dir;
99 int result;
100
101 if (strcmp (step->__from_name, "UNICODE//") == 0)
102 dir = from_unicode;
103 else
104 dir = to_unicode;
105
106 new_data = (struct unicode_data *) malloc (sizeof (struct unicode_data));
107
108 result = __GCONV_NOMEM;
109 if (new_data != NULL)
110 {
111 new_data->dir = dir;
112 step->__data = new_data;
113
114 if (dir == from_unicode)
115 {
116 step->__min_needed_from = MIN_NEEDED_FROM;
117 step->__max_needed_from = MIN_NEEDED_FROM;
118 step->__min_needed_to = MIN_NEEDED_TO;
119 step->__max_needed_to = MIN_NEEDED_TO;
120 }
121 else
122 {
123 step->__min_needed_from = MIN_NEEDED_TO;
124 step->__max_needed_from = MIN_NEEDED_TO;
125 step->__min_needed_to = MIN_NEEDED_FROM;
126 step->__max_needed_to = MIN_NEEDED_FROM;
127 }
128
129 step->__stateful = 0;
130
131 result = __GCONV_OK;
132 }
133
134 return result;
135}
136
137
138extern void gconv_end (struct __gconv_step *data);
139void
140gconv_end (struct __gconv_step *data)
141{
142 free (data->__data);
143}
144
145
146/* Convert from the internal (UCS4-like) format to UCS2. */
147#define MIN_NEEDED_INPUT MIN_NEEDED_TO
148#define MIN_NEEDED_OUTPUT MIN_NEEDED_FROM
149#define LOOPFCT TO_LOOP
150#define BODY \
151 { \
152 uint32_t c = get32 (inptr); \
153 \
154 if (__glibc_unlikely (c >= 0x10000)) \
155 { \
156 UNICODE_TAG_HANDLER (c, 4); \
157 STANDARD_TO_LOOP_ERR_HANDLER (4); \
158 } \
159 else if (__glibc_unlikely (c >= 0xd800 && c < 0xe000)) \
160 { \
161 /* Surrogate characters in UCS-4 input are not valid. \
162 We must catch this, because the UCS-2 output might be \
163 interpreted as UTF-16 by other programs. If we let \
164 surrogates pass through, attackers could make a security \
165 hole exploit by synthesizing any desired plane 1-16 \
166 character. */ \
167 result = __GCONV_ILLEGAL_INPUT; \
168 if (! ignore_errors_p ()) \
169 break; \
170 inptr += 4; \
171 ++*irreversible; \
172 continue; \
173 } \
174 else \
175 { \
176 put16 (outptr, c); \
177 outptr += 2; \
178 } \
179 \
180 inptr += 4; \
181 }
182#define LOOP_NEED_FLAGS
183#define EXTRA_LOOP_DECLS \
184 , int swap
185#include <iconv/loop.c>
186
187
188/* Convert from UCS2 to the internal (UCS4-like) format. */
189#define MIN_NEEDED_INPUT MIN_NEEDED_FROM
190#define MIN_NEEDED_OUTPUT MIN_NEEDED_TO
191#define LOOPFCT FROM_LOOP
192#define BODY \
193 { \
194 uint16_t u1 = get16 (inptr); \
195 \
196 if (swap) \
197 u1 = bswap_16 (u1); \
198 \
199 if (__glibc_unlikely (u1 >= 0xd800 && u1 < 0xe000)) \
200 { \
201 /* Surrogate characters in UCS-2 input are not valid. Reject \
202 them. (Catching this here is not security relevant.) */ \
203 STANDARD_FROM_LOOP_ERR_HANDLER (2); \
204 } \
205 \
206 put32 (outptr, u1); \
207 \
208 inptr += 2; \
209 outptr += 4; \
210 }
211#define LOOP_NEED_FLAGS
212#define EXTRA_LOOP_DECLS \
213 , int swap
214#include <iconv/loop.c>
215
216
217/* Now define the toplevel functions. */
218#include <iconv/skeleton.c>
219