]>
Commit | Line | Data |
---|---|---|
53e1b683 | 1 | /* SPDX-License-Identifier: LGPL-2.1+ */ |
7f110ff9 LP |
2 | /*** |
3 | This file is part of systemd. | |
4 | ||
036ae95a | 5 | Copyright 2008-2011 Kay Sievers |
7f110ff9 | 6 | Copyright 2012 Lennart Poettering |
7f110ff9 LP |
7 | ***/ |
8 | ||
036ae95a | 9 | /* Parts of this file are based on the GLIB utf8 validation functions. The |
7f110ff9 LP |
10 | * original license text follows. */ |
11 | ||
12 | /* gutf8.c - Operations on UTF-8 strings. | |
13 | * | |
14 | * Copyright (C) 1999 Tom Tromey | |
15 | * Copyright (C) 2000 Red Hat, Inc. | |
16 | * | |
17 | * This library is free software; you can redistribute it and/or | |
23757887 | 18 | * modify it under the terms of the GNU Library General Public |
7f110ff9 LP |
19 | * License as published by the Free Software Foundation; either |
20 | * version 2 of the License, or (at your option) any later version. | |
21 | * | |
22 | * This library is distributed in the hope that it will be useful, | |
23 | * but WITHOUT ANY WARRANTY; without even the implied warranty of | |
23757887 SK |
24 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
25 | * Library General Public License for more details. | |
7f110ff9 | 26 | * |
23757887 SK |
27 | * You should have received a copy of the GNU Library General Public |
28 | * License along with this library; if not, write to the Free Software | |
29 | * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA | |
7f110ff9 LP |
30 | */ |
31 | ||
32 | #include <errno.h> | |
7f110ff9 | 33 | #include <stdbool.h> |
cf0fbc49 TA |
34 | #include <stdlib.h> |
35 | #include <string.h> | |
7f110ff9 | 36 | |
b5efdb8a | 37 | #include "alloc-util.h" |
e4e73a63 | 38 | #include "hexdecoct.h" |
11c3a366 | 39 | #include "macro.h" |
7f110ff9 LP |
40 | #include "utf8.h" |
41 | ||
c932fb71 | 42 | bool unichar_is_valid(char32_t ch) { |
7f110ff9 LP |
43 | |
44 | if (ch >= 0x110000) /* End of unicode space */ | |
45 | return false; | |
46 | if ((ch & 0xFFFFF800) == 0xD800) /* Reserved area for UTF-16 */ | |
47 | return false; | |
48 | if ((ch >= 0xFDD0) && (ch <= 0xFDEF)) /* Reserved */ | |
49 | return false; | |
50 | if ((ch & 0xFFFE) == 0xFFFE) /* BOM (Byte Order Mark) */ | |
51 | return false; | |
52 | ||
53 | return true; | |
54 | } | |
55 | ||
c932fb71 | 56 | static bool unichar_is_control(char32_t ch) { |
ba961854 ZJS |
57 | |
58 | /* | |
59 | 0 to ' '-1 is the C0 range. | |
60 | DEL=0x7F, and DEL+1 to 0x9F is C1 range. | |
61 | '\t' is in C0 range, but more or less harmless and commonly used. | |
62 | */ | |
63 | ||
4c701096 | 64 | return (ch < ' ' && !IN_SET(ch, '\t', '\n')) || |
ba961854 ZJS |
65 | (0x7F <= ch && ch <= 0x9F); |
66 | } | |
67 | ||
7991ac34 DR |
68 | /* count of characters used to encode one unicode char */ |
69 | static int utf8_encoded_expected_len(const char *str) { | |
7e8185ef | 70 | unsigned char c; |
ba961854 | 71 | |
7e8185ef LP |
72 | assert(str); |
73 | ||
74 | c = (unsigned char) str[0]; | |
7991ac34 DR |
75 | if (c < 0x80) |
76 | return 1; | |
77 | if ((c & 0xe0) == 0xc0) | |
78 | return 2; | |
79 | if ((c & 0xf0) == 0xe0) | |
80 | return 3; | |
81 | if ((c & 0xf8) == 0xf0) | |
82 | return 4; | |
83 | if ((c & 0xfc) == 0xf8) | |
84 | return 5; | |
85 | if ((c & 0xfe) == 0xfc) | |
86 | return 6; | |
7e8185ef | 87 | |
7991ac34 DR |
88 | return 0; |
89 | } | |
ba961854 | 90 | |
7991ac34 | 91 | /* decode one unicode char */ |
c932fb71 SL |
92 | int utf8_encoded_to_unichar(const char *str, char32_t *ret_unichar) { |
93 | char32_t unichar; | |
94 | int len, i; | |
7e8185ef LP |
95 | |
96 | assert(str); | |
ba961854 | 97 | |
7991ac34 | 98 | len = utf8_encoded_expected_len(str); |
7e8185ef | 99 | |
7991ac34 DR |
100 | switch (len) { |
101 | case 1: | |
c932fb71 SL |
102 | *ret_unichar = (char32_t)str[0]; |
103 | return 0; | |
7991ac34 DR |
104 | case 2: |
105 | unichar = str[0] & 0x1f; | |
106 | break; | |
107 | case 3: | |
c932fb71 | 108 | unichar = (char32_t)str[0] & 0x0f; |
7991ac34 DR |
109 | break; |
110 | case 4: | |
c932fb71 | 111 | unichar = (char32_t)str[0] & 0x07; |
7991ac34 DR |
112 | break; |
113 | case 5: | |
c932fb71 | 114 | unichar = (char32_t)str[0] & 0x03; |
7991ac34 DR |
115 | break; |
116 | case 6: | |
c932fb71 | 117 | unichar = (char32_t)str[0] & 0x01; |
7991ac34 DR |
118 | break; |
119 | default: | |
7e8185ef | 120 | return -EINVAL; |
ba961854 ZJS |
121 | } |
122 | ||
7991ac34 | 123 | for (i = 1; i < len; i++) { |
c932fb71 | 124 | if (((char32_t)str[i] & 0xc0) != 0x80) |
7e8185ef | 125 | return -EINVAL; |
7991ac34 | 126 | unichar <<= 6; |
c932fb71 | 127 | unichar |= (char32_t)str[i] & 0x3f; |
7991ac34 DR |
128 | } |
129 | ||
c932fb71 SL |
130 | *ret_unichar = unichar; |
131 | ||
132 | return 0; | |
ba961854 ZJS |
133 | } |
134 | ||
0ade5ffe | 135 | bool utf8_is_printable_newline(const char* str, size_t length, bool newline) { |
6ed62be0 | 136 | const char *p; |
7f110ff9 LP |
137 | |
138 | assert(str); | |
139 | ||
6ed62be0 | 140 | for (p = str; length;) { |
c932fb71 SL |
141 | int encoded_len, r; |
142 | char32_t val; | |
7e8185ef | 143 | |
6ed62be0 | 144 | encoded_len = utf8_encoded_valid_unichar(p); |
7e8185ef | 145 | if (encoded_len < 0 || |
144b3d9e LP |
146 | (size_t) encoded_len > length) |
147 | return false; | |
148 | ||
c932fb71 SL |
149 | r = utf8_encoded_to_unichar(p, &val); |
150 | if (r < 0 || | |
f3ee6297 | 151 | unichar_is_control(val) || |
0ade5ffe | 152 | (!newline && val == '\n')) |
7991ac34 | 153 | return false; |
7f110ff9 | 154 | |
7991ac34 | 155 | length -= encoded_len; |
a7176505 | 156 | p += encoded_len; |
7f110ff9 LP |
157 | } |
158 | ||
7991ac34 | 159 | return true; |
7f110ff9 LP |
160 | } |
161 | ||
7991ac34 DR |
162 | const char *utf8_is_valid(const char *str) { |
163 | const uint8_t *p; | |
7f110ff9 LP |
164 | |
165 | assert(str); | |
166 | ||
7991ac34 | 167 | for (p = (const uint8_t*) str; *p; ) { |
faaa5728 LP |
168 | int len; |
169 | ||
170 | len = utf8_encoded_valid_unichar((const char *)p); | |
7991ac34 DR |
171 | if (len < 0) |
172 | return NULL; | |
173 | ||
174 | p += len; | |
175 | } | |
7f110ff9 | 176 | |
7991ac34 | 177 | return str; |
7f110ff9 LP |
178 | } |
179 | ||
550a40ec ZJS |
180 | char *utf8_escape_invalid(const char *str) { |
181 | char *p, *s; | |
182 | ||
183 | assert(str); | |
184 | ||
185 | p = s = malloc(strlen(str) * 4 + 1); | |
186 | if (!p) | |
187 | return NULL; | |
188 | ||
189 | while (*str) { | |
190 | int len; | |
191 | ||
192 | len = utf8_encoded_valid_unichar(str); | |
193 | if (len > 0) { | |
194 | s = mempcpy(s, str, len); | |
195 | str += len; | |
196 | } else { | |
3c6d3052 | 197 | s = stpcpy(s, UTF8_REPLACEMENT_CHARACTER); |
550a40ec ZJS |
198 | str += 1; |
199 | } | |
200 | } | |
7e8185ef | 201 | |
550a40ec ZJS |
202 | *s = '\0'; |
203 | ||
204 | return p; | |
205 | } | |
206 | ||
fec84576 WC |
207 | char *utf8_escape_non_printable(const char *str) { |
208 | char *p, *s; | |
209 | ||
210 | assert(str); | |
211 | ||
212 | p = s = malloc(strlen(str) * 4 + 1); | |
213 | if (!p) | |
214 | return NULL; | |
215 | ||
216 | while (*str) { | |
217 | int len; | |
218 | ||
219 | len = utf8_encoded_valid_unichar(str); | |
220 | if (len > 0) { | |
221 | if (utf8_is_printable(str, len)) { | |
222 | s = mempcpy(s, str, len); | |
223 | str += len; | |
224 | } else { | |
3c6d3052 | 225 | while (len > 0) { |
fec84576 WC |
226 | *(s++) = '\\'; |
227 | *(s++) = 'x'; | |
228 | *(s++) = hexchar((int) *str >> 4); | |
229 | *(s++) = hexchar((int) *str); | |
fec84576 | 230 | |
3c6d3052 | 231 | str += 1; |
313cefa1 | 232 | len--; |
3c6d3052 | 233 | } |
fec84576 WC |
234 | } |
235 | } else { | |
3c6d3052 | 236 | s = stpcpy(s, UTF8_REPLACEMENT_CHARACTER); |
fec84576 WC |
237 | str += 1; |
238 | } | |
239 | } | |
240 | ||
241 | *s = '\0'; | |
242 | ||
243 | return p; | |
244 | } | |
245 | ||
7f110ff9 LP |
246 | char *ascii_is_valid(const char *str) { |
247 | const char *p; | |
248 | ||
249 | assert(str); | |
250 | ||
251 | for (p = str; *p; p++) | |
252 | if ((unsigned char) *p >= 128) | |
253 | return NULL; | |
254 | ||
255 | return (char*) str; | |
256 | } | |
257 | ||
2bb4c7e3 TG |
258 | /** |
259 | * utf8_encode_unichar() - Encode single UCS-4 character as UTF-8 | |
260 | * @out_utf8: output buffer of at least 4 bytes or NULL | |
261 | * @g: UCS-4 character to encode | |
262 | * | |
263 | * This encodes a single UCS-4 character as UTF-8 and writes it into @out_utf8. | |
264 | * The length of the character is returned. It is not zero-terminated! If the | |
265 | * output buffer is NULL, only the length is returned. | |
266 | * | |
267 | * Returns: The length in bytes that the UTF-8 representation does or would | |
268 | * occupy. | |
269 | */ | |
c932fb71 | 270 | size_t utf8_encode_unichar(char *out_utf8, char32_t g) { |
f3ee6297 | 271 | |
2bb4c7e3 TG |
272 | if (g < (1 << 7)) { |
273 | if (out_utf8) | |
274 | out_utf8[0] = g & 0x7f; | |
e7eebcfc | 275 | return 1; |
2bb4c7e3 TG |
276 | } else if (g < (1 << 11)) { |
277 | if (out_utf8) { | |
278 | out_utf8[0] = 0xc0 | ((g >> 6) & 0x1f); | |
279 | out_utf8[1] = 0x80 | (g & 0x3f); | |
280 | } | |
e7eebcfc | 281 | return 2; |
2bb4c7e3 TG |
282 | } else if (g < (1 << 16)) { |
283 | if (out_utf8) { | |
284 | out_utf8[0] = 0xe0 | ((g >> 12) & 0x0f); | |
285 | out_utf8[1] = 0x80 | ((g >> 6) & 0x3f); | |
286 | out_utf8[2] = 0x80 | (g & 0x3f); | |
287 | } | |
e7eebcfc | 288 | return 3; |
2bb4c7e3 TG |
289 | } else if (g < (1 << 21)) { |
290 | if (out_utf8) { | |
291 | out_utf8[0] = 0xf0 | ((g >> 18) & 0x07); | |
292 | out_utf8[1] = 0x80 | ((g >> 12) & 0x3f); | |
293 | out_utf8[2] = 0x80 | ((g >> 6) & 0x3f); | |
294 | out_utf8[3] = 0x80 | (g & 0x3f); | |
295 | } | |
296 | return 4; | |
e7eebcfc | 297 | } |
f3ee6297 LP |
298 | |
299 | return 0; | |
e7eebcfc LP |
300 | } |
301 | ||
2e3d0692 | 302 | char *utf16_to_utf8(const void *s, size_t length) { |
2e3d0692 | 303 | const uint8_t *f; |
e7eebcfc | 304 | char *r, *t; |
2e3d0692 | 305 | |
04166cb7 | 306 | r = new(char, (length * 4 + 1) / 2 + 1); |
2e3d0692 LP |
307 | if (!r) |
308 | return NULL; | |
309 | ||
04166cb7 TG |
310 | f = s; |
311 | t = r; | |
312 | ||
313 | while (f < (const uint8_t*) s + length) { | |
c932fb71 | 314 | char16_t w1, w2; |
04166cb7 TG |
315 | |
316 | /* see RFC 2781 section 2.2 */ | |
317 | ||
318 | w1 = f[1] << 8 | f[0]; | |
319 | f += 2; | |
320 | ||
321 | if (!utf16_is_surrogate(w1)) { | |
dcd12626 | 322 | t += utf8_encode_unichar(t, w1); |
04166cb7 TG |
323 | |
324 | continue; | |
325 | } | |
326 | ||
327 | if (utf16_is_trailing_surrogate(w1)) | |
328 | continue; | |
329 | else if (f >= (const uint8_t*) s + length) | |
330 | break; | |
331 | ||
332 | w2 = f[1] << 8 | f[0]; | |
333 | f += 2; | |
334 | ||
335 | if (!utf16_is_trailing_surrogate(w2)) { | |
336 | f -= 2; | |
337 | continue; | |
338 | } | |
339 | ||
340 | t += utf8_encode_unichar(t, utf16_surrogate_pair_to_unichar(w1, w2)); | |
341 | } | |
2e3d0692 LP |
342 | |
343 | *t = 0; | |
7b4d7cc0 | 344 | return r; |
2e3d0692 | 345 | } |
02a36bc9 | 346 | |
02a36bc9 | 347 | /* expected size used to encode one unicode char */ |
c932fb71 | 348 | static int utf8_unichar_to_encoded_len(char32_t unichar) { |
7e8185ef | 349 | |
02a36bc9 DR |
350 | if (unichar < 0x80) |
351 | return 1; | |
352 | if (unichar < 0x800) | |
353 | return 2; | |
354 | if (unichar < 0x10000) | |
355 | return 3; | |
356 | if (unichar < 0x200000) | |
357 | return 4; | |
358 | if (unichar < 0x4000000) | |
359 | return 5; | |
7e8185ef | 360 | |
02a36bc9 DR |
361 | return 6; |
362 | } | |
363 | ||
364 | /* validate one encoded unicode char and return its length */ | |
365 | int utf8_encoded_valid_unichar(const char *str) { | |
c932fb71 SL |
366 | int len, i, r; |
367 | char32_t unichar; | |
7e8185ef LP |
368 | |
369 | assert(str); | |
02a36bc9 DR |
370 | |
371 | len = utf8_encoded_expected_len(str); | |
372 | if (len == 0) | |
7e8185ef | 373 | return -EINVAL; |
02a36bc9 DR |
374 | |
375 | /* ascii is valid */ | |
376 | if (len == 1) | |
377 | return 1; | |
378 | ||
379 | /* check if expected encoded chars are available */ | |
380 | for (i = 0; i < len; i++) | |
381 | if ((str[i] & 0x80) != 0x80) | |
7e8185ef | 382 | return -EINVAL; |
02a36bc9 | 383 | |
c932fb71 SL |
384 | r = utf8_encoded_to_unichar(str, &unichar); |
385 | if (r < 0) | |
386 | return r; | |
02a36bc9 DR |
387 | |
388 | /* check if encoded length matches encoded value */ | |
389 | if (utf8_unichar_to_encoded_len(unichar) != len) | |
7e8185ef | 390 | return -EINVAL; |
02a36bc9 DR |
391 | |
392 | /* check if value has valid range */ | |
f3ee6297 | 393 | if (!unichar_is_valid(unichar)) |
7e8185ef | 394 | return -EINVAL; |
02a36bc9 DR |
395 | |
396 | return len; | |
397 | } | |
65ee8660 LP |
398 | |
399 | size_t utf8_n_codepoints(const char *str) { | |
400 | size_t n = 0; | |
401 | ||
402 | /* Returns the number of UTF-8 codepoints in this string, or (size_t) -1 if the string is not valid UTF-8. */ | |
403 | ||
404 | while (*str != 0) { | |
405 | int k; | |
406 | ||
407 | k = utf8_encoded_valid_unichar(str); | |
408 | if (k < 0) | |
409 | return (size_t) -1; | |
410 | ||
411 | str += k; | |
412 | n++; | |
413 | } | |
414 | ||
415 | return n; | |
416 | } |