1 | /* Conversion module for UTF-32. |
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 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 ONE_DIRECTION 0 |
41 | #define FROM_DIRECTION (dir == from_utf32) |
42 | #define PREPARE_LOOP \ |
43 | enum direction dir = ((struct utf32_data *) step->__data)->dir; \ |
44 | enum variant var = ((struct utf32_data *) step->__data)->var; \ |
45 | int swap; \ |
46 | if (FROM_DIRECTION && var == UTF_32) \ |
47 | { \ |
48 | if (__glibc_unlikely (data->__invocation_counter == 0)) \ |
49 | { \ |
50 | /* We have to find out which byte order the file is encoded in. */ \ |
51 | if (inptr + 4 > inend) \ |
52 | return (inptr == inend \ |
53 | ? __GCONV_EMPTY_INPUT : __GCONV_INCOMPLETE_INPUT); \ |
54 | \ |
55 | if (get32u (inptr) == BOM) \ |
56 | /* Simply ignore the BOM character. */ \ |
57 | *inptrp = inptr += 4; \ |
58 | else if (get32u (inptr) == BOM_OE) \ |
59 | { \ |
60 | data->__flags |= __GCONV_SWAP; \ |
61 | *inptrp = inptr += 4; \ |
62 | } \ |
63 | } \ |
64 | } \ |
65 | else if (!FROM_DIRECTION && var == UTF_32 && !data->__internal_use \ |
66 | && data->__invocation_counter == 0) \ |
67 | { \ |
68 | /* Emit the Byte Order Mark. */ \ |
69 | if (__glibc_unlikely (outbuf + 4 > outend)) \ |
70 | return __GCONV_FULL_OUTPUT; \ |
71 | \ |
72 | put32u (outbuf, BOM); \ |
73 | outbuf += 4; \ |
74 | } \ |
75 | else if (__builtin_expect (data->__invocation_counter == 0, 0) \ |
76 | && ((var == UTF_32LE && BYTE_ORDER == BIG_ENDIAN) \ |
77 | || (var == UTF_32BE && BYTE_ORDER == LITTLE_ENDIAN))) \ |
78 | data->__flags |= __GCONV_SWAP; \ |
79 | swap = data->__flags & __GCONV_SWAP; |
80 | #define , var, swap |
81 | |
82 | |
83 | /* Direction of the transformation. */ |
84 | enum direction |
85 | { |
86 | illegal_dir, |
87 | to_utf32, |
88 | from_utf32 |
89 | }; |
90 | |
91 | enum variant |
92 | { |
93 | illegal_var, |
94 | UTF_32, |
95 | UTF_32LE, |
96 | UTF_32BE |
97 | }; |
98 | |
99 | struct utf32_data |
100 | { |
101 | enum direction dir; |
102 | enum variant var; |
103 | }; |
104 | |
105 | |
106 | extern int gconv_init (struct __gconv_step *step); |
107 | int |
108 | gconv_init (struct __gconv_step *step) |
109 | { |
110 | /* Determine which direction. */ |
111 | struct utf32_data *new_data; |
112 | enum direction dir = illegal_dir; |
113 | enum variant var = illegal_var; |
114 | int result; |
115 | |
116 | if (__strcasecmp (step->__from_name, "UTF-32//" ) == 0) |
117 | { |
118 | dir = from_utf32; |
119 | var = UTF_32; |
120 | } |
121 | else if (__strcasecmp (step->__to_name, "UTF-32//" ) == 0) |
122 | { |
123 | dir = to_utf32; |
124 | var = UTF_32; |
125 | } |
126 | else if (__strcasecmp (step->__from_name, "UTF-32BE//" ) == 0) |
127 | { |
128 | dir = from_utf32; |
129 | var = UTF_32BE; |
130 | } |
131 | else if (__strcasecmp (step->__to_name, "UTF-32BE//" ) == 0) |
132 | { |
133 | dir = to_utf32; |
134 | var = UTF_32BE; |
135 | } |
136 | else if (__strcasecmp (step->__from_name, "UTF-32LE//" ) == 0) |
137 | { |
138 | dir = from_utf32; |
139 | var = UTF_32LE; |
140 | } |
141 | else if (__strcasecmp (step->__to_name, "UTF-32LE//" ) == 0) |
142 | { |
143 | dir = to_utf32; |
144 | var = UTF_32LE; |
145 | } |
146 | |
147 | result = __GCONV_NOCONV; |
148 | if (__builtin_expect (dir, to_utf32) != illegal_dir) |
149 | { |
150 | new_data = (struct utf32_data *) malloc (sizeof (struct utf32_data)); |
151 | |
152 | result = __GCONV_NOMEM; |
153 | if (new_data != NULL) |
154 | { |
155 | new_data->dir = dir; |
156 | new_data->var = var; |
157 | step->__data = new_data; |
158 | |
159 | if (dir == from_utf32) |
160 | { |
161 | step->__min_needed_from = MIN_NEEDED_FROM; |
162 | step->__max_needed_from = MIN_NEEDED_FROM; |
163 | step->__min_needed_to = MIN_NEEDED_TO; |
164 | step->__max_needed_to = MIN_NEEDED_TO; |
165 | } |
166 | else |
167 | { |
168 | step->__min_needed_from = MIN_NEEDED_TO; |
169 | step->__max_needed_from = MIN_NEEDED_TO; |
170 | step->__min_needed_to = MIN_NEEDED_FROM; |
171 | step->__max_needed_to = MIN_NEEDED_FROM; |
172 | } |
173 | |
174 | step->__stateful = 0; |
175 | |
176 | result = __GCONV_OK; |
177 | } |
178 | } |
179 | |
180 | return result; |
181 | } |
182 | |
183 | |
184 | extern void gconv_end (struct __gconv_step *data); |
185 | void |
186 | gconv_end (struct __gconv_step *data) |
187 | { |
188 | free (data->__data); |
189 | } |
190 | |
191 | |
192 | /* Convert from the internal (UCS4-like) format to UTF-32. */ |
193 | #define MIN_NEEDED_INPUT MIN_NEEDED_TO |
194 | #define MIN_NEEDED_OUTPUT MIN_NEEDED_FROM |
195 | #define LOOPFCT TO_LOOP |
196 | #define BODY \ |
197 | { \ |
198 | uint32_t c = get32 (inptr); \ |
199 | \ |
200 | if (__glibc_unlikely (c >= 0x110000)) \ |
201 | { \ |
202 | STANDARD_TO_LOOP_ERR_HANDLER (4); \ |
203 | } \ |
204 | else if (__glibc_unlikely (c >= 0xd800 && c < 0xe000)) \ |
205 | { \ |
206 | /* Surrogate characters in UCS-4 input are not valid. \ |
207 | We must catch this. If we let surrogates pass through, \ |
208 | attackers could make a security hole exploit by \ |
209 | generating "irregular UTF-32" sequences. */ \ |
210 | result = __GCONV_ILLEGAL_INPUT; \ |
211 | if (! ignore_errors_p ()) \ |
212 | break; \ |
213 | inptr += 4; \ |
214 | ++*irreversible; \ |
215 | continue; \ |
216 | } \ |
217 | \ |
218 | if (swap) \ |
219 | c = bswap_32 (c); \ |
220 | put32 (outptr, c); \ |
221 | \ |
222 | outptr += 4; \ |
223 | inptr += 4; \ |
224 | } |
225 | #define LOOP_NEED_FLAGS |
226 | #define \ |
227 | , enum variant var, int swap |
228 | #include <iconv/loop.c> |
229 | |
230 | |
231 | /* Convert from UTF-32 to the internal (UCS4-like) format. */ |
232 | #define MIN_NEEDED_INPUT MIN_NEEDED_FROM |
233 | #define MIN_NEEDED_OUTPUT MIN_NEEDED_TO |
234 | #define LOOPFCT FROM_LOOP |
235 | #define BODY \ |
236 | { \ |
237 | uint32_t u1 = get32 (inptr); \ |
238 | \ |
239 | if (swap) \ |
240 | u1 = bswap_32 (u1); \ |
241 | \ |
242 | if (__glibc_unlikely (u1 >= 0x110000 || (u1 >= 0xd800 && u1 < 0xe000))) \ |
243 | { \ |
244 | /* This is illegal. */ \ |
245 | STANDARD_FROM_LOOP_ERR_HANDLER (4); \ |
246 | } \ |
247 | \ |
248 | put32 (outptr, u1); \ |
249 | inptr += 4; \ |
250 | outptr += 4; \ |
251 | } |
252 | #define LOOP_NEED_FLAGS |
253 | #define \ |
254 | , enum variant var, int swap |
255 | #include <iconv/loop.c> |
256 | |
257 | |
258 | /* Now define the toplevel functions. */ |
259 | #include <iconv/skeleton.c> |
260 | |