]>
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" |
3f536d5b | 38 | #include "gunicode.h" |
e4e73a63 | 39 | #include "hexdecoct.h" |
11c3a366 | 40 | #include "macro.h" |
7f110ff9 LP |
41 | #include "utf8.h" |
42 | ||
c932fb71 | 43 | bool unichar_is_valid(char32_t ch) { |
7f110ff9 LP |
44 | |
45 | if (ch >= 0x110000) /* End of unicode space */ | |
46 | return false; | |
47 | if ((ch & 0xFFFFF800) == 0xD800) /* Reserved area for UTF-16 */ | |
48 | return false; | |
49 | if ((ch >= 0xFDD0) && (ch <= 0xFDEF)) /* Reserved */ | |
50 | return false; | |
51 | if ((ch & 0xFFFE) == 0xFFFE) /* BOM (Byte Order Mark) */ | |
52 | return false; | |
53 | ||
54 | return true; | |
55 | } | |
56 | ||
c932fb71 | 57 | static bool unichar_is_control(char32_t ch) { |
ba961854 ZJS |
58 | |
59 | /* | |
60 | 0 to ' '-1 is the C0 range. | |
61 | DEL=0x7F, and DEL+1 to 0x9F is C1 range. | |
62 | '\t' is in C0 range, but more or less harmless and commonly used. | |
63 | */ | |
64 | ||
4c701096 | 65 | return (ch < ' ' && !IN_SET(ch, '\t', '\n')) || |
ba961854 ZJS |
66 | (0x7F <= ch && ch <= 0x9F); |
67 | } | |
68 | ||
7991ac34 DR |
69 | /* count of characters used to encode one unicode char */ |
70 | static int utf8_encoded_expected_len(const char *str) { | |
7e8185ef | 71 | unsigned char c; |
ba961854 | 72 | |
7e8185ef LP |
73 | assert(str); |
74 | ||
75 | c = (unsigned char) str[0]; | |
7991ac34 DR |
76 | if (c < 0x80) |
77 | return 1; | |
78 | if ((c & 0xe0) == 0xc0) | |
79 | return 2; | |
80 | if ((c & 0xf0) == 0xe0) | |
81 | return 3; | |
82 | if ((c & 0xf8) == 0xf0) | |
83 | return 4; | |
84 | if ((c & 0xfc) == 0xf8) | |
85 | return 5; | |
86 | if ((c & 0xfe) == 0xfc) | |
87 | return 6; | |
7e8185ef | 88 | |
7991ac34 DR |
89 | return 0; |
90 | } | |
ba961854 | 91 | |
7991ac34 | 92 | /* decode one unicode char */ |
c932fb71 SL |
93 | int utf8_encoded_to_unichar(const char *str, char32_t *ret_unichar) { |
94 | char32_t unichar; | |
95 | int len, i; | |
7e8185ef LP |
96 | |
97 | assert(str); | |
ba961854 | 98 | |
7991ac34 | 99 | len = utf8_encoded_expected_len(str); |
7e8185ef | 100 | |
7991ac34 DR |
101 | switch (len) { |
102 | case 1: | |
c932fb71 SL |
103 | *ret_unichar = (char32_t)str[0]; |
104 | return 0; | |
7991ac34 DR |
105 | case 2: |
106 | unichar = str[0] & 0x1f; | |
107 | break; | |
108 | case 3: | |
c932fb71 | 109 | unichar = (char32_t)str[0] & 0x0f; |
7991ac34 DR |
110 | break; |
111 | case 4: | |
c932fb71 | 112 | unichar = (char32_t)str[0] & 0x07; |
7991ac34 DR |
113 | break; |
114 | case 5: | |
c932fb71 | 115 | unichar = (char32_t)str[0] & 0x03; |
7991ac34 DR |
116 | break; |
117 | case 6: | |
c932fb71 | 118 | unichar = (char32_t)str[0] & 0x01; |
7991ac34 DR |
119 | break; |
120 | default: | |
7e8185ef | 121 | return -EINVAL; |
ba961854 ZJS |
122 | } |
123 | ||
7991ac34 | 124 | for (i = 1; i < len; i++) { |
c932fb71 | 125 | if (((char32_t)str[i] & 0xc0) != 0x80) |
7e8185ef | 126 | return -EINVAL; |
7991ac34 | 127 | unichar <<= 6; |
c932fb71 | 128 | unichar |= (char32_t)str[i] & 0x3f; |
7991ac34 DR |
129 | } |
130 | ||
c932fb71 SL |
131 | *ret_unichar = unichar; |
132 | ||
133 | return 0; | |
ba961854 ZJS |
134 | } |
135 | ||
0ade5ffe | 136 | bool utf8_is_printable_newline(const char* str, size_t length, bool newline) { |
6ed62be0 | 137 | const char *p; |
7f110ff9 LP |
138 | |
139 | assert(str); | |
140 | ||
6ed62be0 | 141 | for (p = str; length;) { |
c932fb71 SL |
142 | int encoded_len, r; |
143 | char32_t val; | |
7e8185ef | 144 | |
6ed62be0 | 145 | encoded_len = utf8_encoded_valid_unichar(p); |
7e8185ef | 146 | if (encoded_len < 0 || |
144b3d9e LP |
147 | (size_t) encoded_len > length) |
148 | return false; | |
149 | ||
c932fb71 SL |
150 | r = utf8_encoded_to_unichar(p, &val); |
151 | if (r < 0 || | |
f3ee6297 | 152 | unichar_is_control(val) || |
0ade5ffe | 153 | (!newline && val == '\n')) |
7991ac34 | 154 | return false; |
7f110ff9 | 155 | |
7991ac34 | 156 | length -= encoded_len; |
a7176505 | 157 | p += encoded_len; |
7f110ff9 LP |
158 | } |
159 | ||
7991ac34 | 160 | return true; |
7f110ff9 LP |
161 | } |
162 | ||
7991ac34 DR |
163 | const char *utf8_is_valid(const char *str) { |
164 | const uint8_t *p; | |
7f110ff9 LP |
165 | |
166 | assert(str); | |
167 | ||
7991ac34 | 168 | for (p = (const uint8_t*) str; *p; ) { |
faaa5728 LP |
169 | int len; |
170 | ||
171 | len = utf8_encoded_valid_unichar((const char *)p); | |
7991ac34 DR |
172 | if (len < 0) |
173 | return NULL; | |
174 | ||
175 | p += len; | |
176 | } | |
7f110ff9 | 177 | |
7991ac34 | 178 | return str; |
7f110ff9 LP |
179 | } |
180 | ||
550a40ec ZJS |
181 | char *utf8_escape_invalid(const char *str) { |
182 | char *p, *s; | |
183 | ||
184 | assert(str); | |
185 | ||
186 | p = s = malloc(strlen(str) * 4 + 1); | |
187 | if (!p) | |
188 | return NULL; | |
189 | ||
190 | while (*str) { | |
191 | int len; | |
192 | ||
193 | len = utf8_encoded_valid_unichar(str); | |
194 | if (len > 0) { | |
195 | s = mempcpy(s, str, len); | |
196 | str += len; | |
197 | } else { | |
3c6d3052 | 198 | s = stpcpy(s, UTF8_REPLACEMENT_CHARACTER); |
550a40ec ZJS |
199 | str += 1; |
200 | } | |
201 | } | |
7e8185ef | 202 | |
550a40ec ZJS |
203 | *s = '\0'; |
204 | ||
205 | return p; | |
206 | } | |
207 | ||
fec84576 WC |
208 | char *utf8_escape_non_printable(const char *str) { |
209 | char *p, *s; | |
210 | ||
211 | assert(str); | |
212 | ||
213 | p = s = malloc(strlen(str) * 4 + 1); | |
214 | if (!p) | |
215 | return NULL; | |
216 | ||
217 | while (*str) { | |
218 | int len; | |
219 | ||
220 | len = utf8_encoded_valid_unichar(str); | |
221 | if (len > 0) { | |
222 | if (utf8_is_printable(str, len)) { | |
223 | s = mempcpy(s, str, len); | |
224 | str += len; | |
225 | } else { | |
3c6d3052 | 226 | while (len > 0) { |
fec84576 WC |
227 | *(s++) = '\\'; |
228 | *(s++) = 'x'; | |
229 | *(s++) = hexchar((int) *str >> 4); | |
230 | *(s++) = hexchar((int) *str); | |
fec84576 | 231 | |
3c6d3052 | 232 | str += 1; |
313cefa1 | 233 | len--; |
3c6d3052 | 234 | } |
fec84576 WC |
235 | } |
236 | } else { | |
3c6d3052 | 237 | s = stpcpy(s, UTF8_REPLACEMENT_CHARACTER); |
fec84576 WC |
238 | str += 1; |
239 | } | |
240 | } | |
241 | ||
242 | *s = '\0'; | |
243 | ||
244 | return p; | |
245 | } | |
246 | ||
7f110ff9 LP |
247 | char *ascii_is_valid(const char *str) { |
248 | const char *p; | |
249 | ||
250 | assert(str); | |
251 | ||
252 | for (p = str; *p; p++) | |
253 | if ((unsigned char) *p >= 128) | |
254 | return NULL; | |
255 | ||
256 | return (char*) str; | |
257 | } | |
258 | ||
2bb4c7e3 TG |
259 | /** |
260 | * utf8_encode_unichar() - Encode single UCS-4 character as UTF-8 | |
261 | * @out_utf8: output buffer of at least 4 bytes or NULL | |
262 | * @g: UCS-4 character to encode | |
263 | * | |
264 | * This encodes a single UCS-4 character as UTF-8 and writes it into @out_utf8. | |
265 | * The length of the character is returned. It is not zero-terminated! If the | |
266 | * output buffer is NULL, only the length is returned. | |
267 | * | |
268 | * Returns: The length in bytes that the UTF-8 representation does or would | |
269 | * occupy. | |
270 | */ | |
c932fb71 | 271 | size_t utf8_encode_unichar(char *out_utf8, char32_t g) { |
f3ee6297 | 272 | |
2bb4c7e3 TG |
273 | if (g < (1 << 7)) { |
274 | if (out_utf8) | |
275 | out_utf8[0] = g & 0x7f; | |
e7eebcfc | 276 | return 1; |
2bb4c7e3 TG |
277 | } else if (g < (1 << 11)) { |
278 | if (out_utf8) { | |
279 | out_utf8[0] = 0xc0 | ((g >> 6) & 0x1f); | |
280 | out_utf8[1] = 0x80 | (g & 0x3f); | |
281 | } | |
e7eebcfc | 282 | return 2; |
2bb4c7e3 TG |
283 | } else if (g < (1 << 16)) { |
284 | if (out_utf8) { | |
285 | out_utf8[0] = 0xe0 | ((g >> 12) & 0x0f); | |
286 | out_utf8[1] = 0x80 | ((g >> 6) & 0x3f); | |
287 | out_utf8[2] = 0x80 | (g & 0x3f); | |
288 | } | |
e7eebcfc | 289 | return 3; |
2bb4c7e3 TG |
290 | } else if (g < (1 << 21)) { |
291 | if (out_utf8) { | |
292 | out_utf8[0] = 0xf0 | ((g >> 18) & 0x07); | |
293 | out_utf8[1] = 0x80 | ((g >> 12) & 0x3f); | |
294 | out_utf8[2] = 0x80 | ((g >> 6) & 0x3f); | |
295 | out_utf8[3] = 0x80 | (g & 0x3f); | |
296 | } | |
297 | return 4; | |
e7eebcfc | 298 | } |
f3ee6297 LP |
299 | |
300 | return 0; | |
e7eebcfc LP |
301 | } |
302 | ||
2e3d0692 | 303 | char *utf16_to_utf8(const void *s, size_t length) { |
2e3d0692 | 304 | const uint8_t *f; |
e7eebcfc | 305 | char *r, *t; |
2e3d0692 | 306 | |
04166cb7 | 307 | r = new(char, (length * 4 + 1) / 2 + 1); |
2e3d0692 LP |
308 | if (!r) |
309 | return NULL; | |
310 | ||
04166cb7 TG |
311 | f = s; |
312 | t = r; | |
313 | ||
314 | while (f < (const uint8_t*) s + length) { | |
c932fb71 | 315 | char16_t w1, w2; |
04166cb7 TG |
316 | |
317 | /* see RFC 2781 section 2.2 */ | |
318 | ||
319 | w1 = f[1] << 8 | f[0]; | |
320 | f += 2; | |
321 | ||
322 | if (!utf16_is_surrogate(w1)) { | |
dcd12626 | 323 | t += utf8_encode_unichar(t, w1); |
04166cb7 TG |
324 | |
325 | continue; | |
326 | } | |
327 | ||
328 | if (utf16_is_trailing_surrogate(w1)) | |
329 | continue; | |
330 | else if (f >= (const uint8_t*) s + length) | |
331 | break; | |
332 | ||
333 | w2 = f[1] << 8 | f[0]; | |
334 | f += 2; | |
335 | ||
336 | if (!utf16_is_trailing_surrogate(w2)) { | |
337 | f -= 2; | |
338 | continue; | |
339 | } | |
340 | ||
341 | t += utf8_encode_unichar(t, utf16_surrogate_pair_to_unichar(w1, w2)); | |
342 | } | |
2e3d0692 LP |
343 | |
344 | *t = 0; | |
7b4d7cc0 | 345 | return r; |
2e3d0692 | 346 | } |
02a36bc9 | 347 | |
02a36bc9 | 348 | /* expected size used to encode one unicode char */ |
c932fb71 | 349 | static int utf8_unichar_to_encoded_len(char32_t unichar) { |
7e8185ef | 350 | |
02a36bc9 DR |
351 | if (unichar < 0x80) |
352 | return 1; | |
353 | if (unichar < 0x800) | |
354 | return 2; | |
355 | if (unichar < 0x10000) | |
356 | return 3; | |
357 | if (unichar < 0x200000) | |
358 | return 4; | |
359 | if (unichar < 0x4000000) | |
360 | return 5; | |
7e8185ef | 361 | |
02a36bc9 DR |
362 | return 6; |
363 | } | |
364 | ||
365 | /* validate one encoded unicode char and return its length */ | |
366 | int utf8_encoded_valid_unichar(const char *str) { | |
c932fb71 SL |
367 | int len, i, r; |
368 | char32_t unichar; | |
7e8185ef LP |
369 | |
370 | assert(str); | |
02a36bc9 DR |
371 | |
372 | len = utf8_encoded_expected_len(str); | |
373 | if (len == 0) | |
7e8185ef | 374 | return -EINVAL; |
02a36bc9 DR |
375 | |
376 | /* ascii is valid */ | |
377 | if (len == 1) | |
378 | return 1; | |
379 | ||
380 | /* check if expected encoded chars are available */ | |
381 | for (i = 0; i < len; i++) | |
382 | if ((str[i] & 0x80) != 0x80) | |
7e8185ef | 383 | return -EINVAL; |
02a36bc9 | 384 | |
c932fb71 SL |
385 | r = utf8_encoded_to_unichar(str, &unichar); |
386 | if (r < 0) | |
387 | return r; | |
02a36bc9 DR |
388 | |
389 | /* check if encoded length matches encoded value */ | |
390 | if (utf8_unichar_to_encoded_len(unichar) != len) | |
7e8185ef | 391 | return -EINVAL; |
02a36bc9 DR |
392 | |
393 | /* check if value has valid range */ | |
f3ee6297 | 394 | if (!unichar_is_valid(unichar)) |
7e8185ef | 395 | return -EINVAL; |
02a36bc9 DR |
396 | |
397 | return len; | |
398 | } | |
65ee8660 LP |
399 | |
400 | size_t utf8_n_codepoints(const char *str) { | |
401 | size_t n = 0; | |
402 | ||
403 | /* Returns the number of UTF-8 codepoints in this string, or (size_t) -1 if the string is not valid UTF-8. */ | |
404 | ||
405 | while (*str != 0) { | |
406 | int k; | |
407 | ||
408 | k = utf8_encoded_valid_unichar(str); | |
409 | if (k < 0) | |
410 | return (size_t) -1; | |
411 | ||
412 | str += k; | |
413 | n++; | |
414 | } | |
415 | ||
416 | return n; | |
417 | } | |
3f536d5b LP |
418 | |
419 | size_t utf8_console_width(const char *str) { | |
420 | size_t n = 0; | |
421 | ||
422 | /* Returns the approximate width a string will take on screen when printed on a character cell | |
423 | * terminal/console. */ | |
424 | ||
425 | while (*str != 0) { | |
426 | char32_t c; | |
427 | ||
428 | if (utf8_encoded_to_unichar(str, &c) < 0) | |
429 | return (size_t) -1; | |
430 | ||
431 | str = utf8_next_char(str); | |
432 | ||
433 | n += unichar_iswide(c) ? 2 : 1; | |
434 | } | |
435 | ||
436 | return n; | |
437 | } |