FFmpeg coverage


Directory: ../../../ffmpeg/
File: src/libavcodec/htmlsubtitles.c
Date: 2024-04-25 15:36:26
Exec Total Coverage
Lines: 152 165 92.1%
Functions: 6 6 100.0%
Branches: 136 166 81.9%

Line Branch Exec Source
1 /*
2 * Copyright (c) 2010 Aurelien Jacobs <aurel@gnuage.org>
3 * Copyright (c) 2017 Clément Bœsch <u@pkh.me>
4 *
5 * This file is part of FFmpeg.
6 *
7 * FFmpeg is free software; you can redistribute it and/or
8 * modify it under the terms of the GNU Lesser General Public
9 * License as published by the Free Software Foundation; either
10 * version 2.1 of the License, or (at your option) any later version.
11 *
12 * FFmpeg is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15 * Lesser General Public License for more details.
16 *
17 * You should have received a copy of the GNU Lesser General Public
18 * License along with FFmpeg; if not, write to the Free Software
19 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20 */
21
22 #include "libavutil/avassert.h"
23 #include "libavutil/avstring.h"
24 #include "libavutil/common.h"
25 #include "libavutil/parseutils.h"
26 #include "htmlsubtitles.h"
27 #include <ctype.h>
28
29 81 static int html_color_parse(void *log_ctx, const char *str)
30 {
31 uint8_t rgba[4];
32 81 int nb_sharps = 0;
33
2/2
✓ Branch 0 taken 37 times.
✓ Branch 1 taken 81 times.
118 while (str[nb_sharps] == '#')
34 37 nb_sharps++;
35
2/2
✓ Branch 0 taken 36 times.
✓ Branch 1 taken 45 times.
81 str += FFMAX(0, nb_sharps - 1);
36
1/2
✗ Branch 1 not taken.
✓ Branch 2 taken 81 times.
81 if (av_parse_color(rgba, str, strcspn(str, "\" >"), log_ctx) < 0)
37 return -1;
38 81 return rgba[0] | rgba[1] << 8 | rgba[2] << 16;
39 }
40
41 732 static void rstrip_spaces_buf(AVBPrint *buf)
42 {
43
1/2
✓ Branch 1 taken 732 times.
✗ Branch 2 not taken.
732 if (av_bprint_is_complete(buf))
44
4/4
✓ Branch 0 taken 869 times.
✓ Branch 1 taken 1 times.
✓ Branch 2 taken 138 times.
✓ Branch 3 taken 731 times.
870 while (buf->len > 0 && buf->str[buf->len - 1] == ' ')
45 138 buf->str[--buf->len] = 0;
46 732 }
47
48 /*
49 * Fast code for scanning text enclosed in braces. Functionally
50 * equivalent to this sscanf call:
51 *
52 * sscanf(in, "{\\an%*1u}%n", &len) >= 0 && len > 0
53 */
54 96 static int scanbraces(const char* in) {
55
2/2
✓ Branch 0 taken 18 times.
✓ Branch 1 taken 78 times.
96 if (strncmp(in, "{\\an", 4) != 0) {
56 18 return 0;
57 }
58
1/2
✗ Branch 0 not taken.
✓ Branch 1 taken 78 times.
78 if (!av_isdigit(in[4])) {
59 return 0;
60 }
61
1/2
✗ Branch 0 not taken.
✓ Branch 1 taken 78 times.
78 if (in[5] != '}') {
62 return 0;
63 }
64 78 return 1;
65 }
66
67 /* skip all {\xxx} substrings except for {\an%d}
68 and all microdvd like styles such as {Y:xxx} */
69 96 static void handle_open_brace(AVBPrint *dst, const char **inp, int *an, int *closing_brace_missing)
70 {
71 96 const char *in = *inp;
72
73 96 *an += scanbraces(in);
74
75
1/2
✓ Branch 0 taken 96 times.
✗ Branch 1 not taken.
96 if (!*closing_brace_missing) {
76
4/4
✓ Branch 0 taken 24 times.
✓ Branch 1 taken 72 times.
✓ Branch 2 taken 12 times.
✓ Branch 3 taken 12 times.
96 if ( (*an != 1 && in[1] == '\\')
77
4/6
✓ Branch 0 taken 84 times.
✗ Branch 1 not taken.
✓ Branch 2 taken 6 times.
✓ Branch 3 taken 78 times.
✓ Branch 4 taken 6 times.
✗ Branch 5 not taken.
84 || (in[1] && strchr("CcFfoPSsYy", in[1]) && in[2] == ':')) {
78 18 char *bracep = strchr(in+2, '}');
79
1/2
✓ Branch 0 taken 18 times.
✗ Branch 1 not taken.
18 if (bracep) {
80 18 *inp = bracep;
81 18 return;
82 } else
83 *closing_brace_missing = 1;
84 }
85 }
86
87 78 av_bprint_chars(dst, *in, 1);
88 }
89
90 struct font_tag {
91 char face[128];
92 int size;
93 uint32_t color;
94 };
95
96 /*
97 * Fast code for scanning the rest of a tag. Functionally equivalent to
98 * this sscanf call:
99 *
100 * sscanf(in, "%127[^<>]>%n", buffer, lenp) == 2
101 */
102 626 static int scantag(const char* in, char* buffer, int* lenp) {
103 int len;
104
105
1/2
✓ Branch 0 taken 5011 times.
✗ Branch 1 not taken.
5011 for (len = 0; len < 128; len++) {
106 5011 const char c = *in++;
107
4/4
✓ Branch 0 taken 9 times.
✓ Branch 1 taken 11 times.
✓ Branch 2 taken 606 times.
✓ Branch 3 taken 4385 times.
5011 switch (c) {
108 9 case '\0':
109 9 return 0;
110 11 case '<':
111 11 return 0;
112 606 case '>':
113 606 buffer[len] = '\0';
114 606 *lenp = len+1;
115 606 return 1;
116 4385 default:
117 4385 break;
118 }
119 4385 buffer[len] = c;
120 }
121 return 0;
122 }
123
124 /*
125 * The general politic of the convert is to mask unsupported tags or formatting
126 * errors (but still alert the user/subtitles writer with an error/warning)
127 * without dropping any actual text content for the final user.
128 */
129 337 int ff_htmlmarkup_to_ass(void *log_ctx, AVBPrint *dst, const char *in)
130 {
131 char *param, buffer[128];
132 337 int len, tag_close, sptr = 0, line_start = 1, an = 0, end = 0;
133 337 int closing_brace_missing = 0;
134 int i, likely_a_tag;
135
136 /*
137 * state stack is only present for fonts since they are the only tags where
138 * the state is not binary. Here is a typical use case:
139 *
140 * <font color="red" size=10>
141 * red 10
142 * <font size=50> RED AND BIG </font>
143 * red 10 again
144 * </font>
145 *
146 * On the other hand, using the state system for all the tags should be
147 * avoided because it breaks wrongly nested tags such as:
148 *
149 * <b> foo <i> bar </b> bla </i>
150 *
151 * We don't want to break here; instead, we will treat all these tags as
152 * binary state markers. Basically, "<b>" will activate bold, and "</b>"
153 * will deactivate it, whatever the current state.
154 *
155 * This will also prevents cases where we have a random closing tag
156 * remaining after the opening one was dropped. Yes, this happens and we
157 * still don't want to print a "</b>" at the end of the dialog event.
158 */
159 struct font_tag stack[16];
160
161 337 memset(&stack[0], 0, sizeof(stack[0]));
162
163
3/4
✓ Branch 0 taken 27050 times.
✗ Branch 1 not taken.
✓ Branch 2 taken 26713 times.
✓ Branch 3 taken 337 times.
27050 for (; !end && *in; in++) {
164
5/6
✗ Branch 0 not taken.
✓ Branch 1 taken 395 times.
✓ Branch 2 taken 3767 times.
✓ Branch 3 taken 96 times.
✓ Branch 4 taken 626 times.
✓ Branch 5 taken 21829 times.
26713 switch (*in) {
165 case '\r':
166 break;
167 395 case '\n':
168
1/2
✗ Branch 0 not taken.
✓ Branch 1 taken 395 times.
395 if (line_start) {
169 end = 1;
170 break;
171 }
172 395 rstrip_spaces_buf(dst);
173 395 av_bprintf(dst, "\\N");
174 395 line_start = 1;
175 395 break;
176 3767 case ' ':
177
2/2
✓ Branch 0 taken 3707 times.
✓ Branch 1 taken 60 times.
3767 if (!line_start)
178 3707 av_bprint_chars(dst, *in, 1);
179 3767 break;
180 96 case '{':
181 96 handle_open_brace(dst, &in, &an, &closing_brace_missing);
182 96 break;
183 626 case '<':
184 /*
185 * "<<" are likely latin guillemets in ASCII or some kind of random
186 * style effect; see sub/badsyntax.srt in the FATE samples
187 * directory for real test cases.
188 */
189
190 626 likely_a_tag = 1;
191
2/2
✓ Branch 0 taken 27 times.
✓ Branch 1 taken 626 times.
653 for (i = 0; in[1] == '<'; i++) {
192 27 av_bprint_chars(dst, '<', 1);
193 27 likely_a_tag = 0;
194 27 in++;
195 }
196
197 626 tag_close = in[1] == '/';
198
2/2
✓ Branch 0 taken 288 times.
✓ Branch 1 taken 338 times.
626 if (tag_close)
199 288 likely_a_tag = 1;
200
201
1/2
✗ Branch 0 not taken.
✓ Branch 1 taken 626 times.
626 av_assert0(in[0] == '<');
202
203 626 len = 0;
204
205
3/4
✓ Branch 1 taken 606 times.
✓ Branch 2 taken 20 times.
✓ Branch 3 taken 606 times.
✗ Branch 4 not taken.
1232 if (scantag(in+tag_close+1, buffer, &len) && len > 0) {
206 606 const int skip = len + tag_close;
207 606 const char *tagname = buffer;
208
2/2
✓ Branch 0 taken 5 times.
✓ Branch 1 taken 606 times.
611 while (*tagname == ' ') {
209 5 likely_a_tag = 0;
210 5 tagname++;
211 }
212
2/2
✓ Branch 0 taken 148 times.
✓ Branch 1 taken 458 times.
606 if ((param = strchr(tagname, ' ')))
213 148 *param++ = 0;
214
215 /* Check if this is likely a tag */
216 #define LIKELY_A_TAG_CHAR(x) (((x) >= '0' && (x) <= '9') || \
217 ((x) >= 'a' && (x) <= 'z') || \
218 ((x) >= 'A' && (x) <= 'Z') || \
219 (x) == '_' || (x) == '/')
220
2/2
✓ Branch 0 taken 2228 times.
✓ Branch 1 taken 597 times.
2825 for (i = 0; tagname[i]; i++) {
221
15/16
✓ Branch 0 taken 2218 times.
✓ Branch 1 taken 10 times.
✓ Branch 2 taken 2193 times.
✓ Branch 3 taken 25 times.
✓ Branch 4 taken 2115 times.
✓ Branch 5 taken 88 times.
✗ Branch 6 not taken.
✓ Branch 7 taken 2115 times.
✓ Branch 8 taken 77 times.
✓ Branch 9 taken 11 times.
✓ Branch 10 taken 66 times.
✓ Branch 11 taken 11 times.
✓ Branch 12 taken 11 times.
✓ Branch 13 taken 66 times.
✓ Branch 14 taken 9 times.
✓ Branch 15 taken 2 times.
2228 if (!LIKELY_A_TAG_CHAR(tagname[i])) {
222 9 likely_a_tag = 0;
223 9 break;
224 }
225 }
226
227
2/2
✓ Branch 1 taken 258 times.
✓ Branch 2 taken 348 times.
606 if (!av_strcasecmp(tagname, "font")) {
228
3/4
✓ Branch 0 taken 129 times.
✓ Branch 1 taken 129 times.
✓ Branch 2 taken 129 times.
✗ Branch 3 not taken.
387 if (tag_close && sptr > 0) {
229 129 struct font_tag *cur_tag = &stack[sptr--];
230 129 struct font_tag *last_tag = &stack[sptr];
231
232
2/2
✓ Branch 0 taken 75 times.
✓ Branch 1 taken 54 times.
129 if (cur_tag->size) {
233
2/2
✓ Branch 0 taken 63 times.
✓ Branch 1 taken 12 times.
75 if (!last_tag->size)
234 63 av_bprintf(dst, "{\\fs}");
235
2/2
✓ Branch 0 taken 6 times.
✓ Branch 1 taken 6 times.
12 else if (last_tag->size != cur_tag->size)
236 6 av_bprintf(dst, "{\\fs%d}", last_tag->size);
237 }
238
239
2/2
✓ Branch 0 taken 103 times.
✓ Branch 1 taken 26 times.
129 if (cur_tag->color & 0xff000000) {
240
2/2
✓ Branch 0 taken 74 times.
✓ Branch 1 taken 29 times.
103 if (!(last_tag->color & 0xff000000))
241 74 av_bprintf(dst, "{\\c}");
242
2/2
✓ Branch 0 taken 7 times.
✓ Branch 1 taken 22 times.
29 else if (last_tag->color != cur_tag->color)
243 7 av_bprintf(dst, "{\\c&H%"PRIX32"&}", last_tag->color & 0xffffff);
244 }
245
246
2/2
✓ Branch 0 taken 6 times.
✓ Branch 1 taken 123 times.
129 if (cur_tag->face[0]) {
247
1/2
✓ Branch 0 taken 6 times.
✗ Branch 1 not taken.
6 if (!last_tag->face[0])
248 6 av_bprintf(dst, "{\\fn}");
249 else if (strcmp(last_tag->face, cur_tag->face))
250 av_bprintf(dst, "{\\fn%s}", last_tag->face);
251 }
252
2/4
✓ Branch 0 taken 129 times.
✗ Branch 1 not taken.
✓ Branch 2 taken 129 times.
✗ Branch 3 not taken.
129 } else if (!tag_close && sptr < FF_ARRAY_ELEMS(stack) - 1) {
253 129 struct font_tag *new_tag = &stack[sptr + 1];
254
255 129 *new_tag = stack[sptr++];
256
257
2/2
✓ Branch 0 taken 157 times.
✓ Branch 1 taken 129 times.
286 while (param) {
258
2/2
✓ Branch 1 taken 69 times.
✓ Branch 2 taken 88 times.
157 if (!av_strncasecmp(param, "size=", 5)) {
259
2/2
✓ Branch 0 taken 24 times.
✓ Branch 1 taken 45 times.
69 param += 5 + (param[5] == '"');
260
1/2
✓ Branch 0 taken 69 times.
✗ Branch 1 not taken.
69 if (sscanf(param, "%u", &new_tag->size) == 1)
261 69 av_bprintf(dst, "{\\fs%u}", new_tag->size);
262
2/2
✓ Branch 1 taken 81 times.
✓ Branch 2 taken 7 times.
88 } else if (!av_strncasecmp(param, "color=", 6)) {
263 int color;
264
2/2
✓ Branch 0 taken 54 times.
✓ Branch 1 taken 27 times.
81 param += 6 + (param[6] == '"');
265 81 color = html_color_parse(log_ctx, param);
266
1/2
✓ Branch 0 taken 81 times.
✗ Branch 1 not taken.
81 if (color >= 0) {
267 81 new_tag->color = 0xff000000 | color;
268 81 av_bprintf(dst, "{\\c&H%"PRIX32"&}", new_tag->color & 0xffffff);
269 }
270
2/2
✓ Branch 1 taken 6 times.
✓ Branch 2 taken 1 times.
7 } else if (!av_strncasecmp(param, "face=", 5)) {
271
1/2
✓ Branch 0 taken 6 times.
✗ Branch 1 not taken.
6 param += 5 + (param[5] == '"');
272 6 len = strcspn(param,
273
1/2
✓ Branch 0 taken 6 times.
✗ Branch 1 not taken.
6 param[-1] == '"' ? "\"" :" ");
274 6 av_strlcpy(new_tag->face, param,
275
1/2
✓ Branch 0 taken 6 times.
✗ Branch 1 not taken.
6 FFMIN(sizeof(new_tag->face), len+1));
276 6 param += len;
277 6 av_bprintf(dst, "{\\fn%s}", new_tag->face);
278 }
279
2/2
✓ Branch 0 taken 28 times.
✓ Branch 1 taken 129 times.
157 if ((param = strchr(param, ' ')))
280 28 param++;
281 }
282 }
283 258 in += skip;
284
6/6
✓ Branch 0 taken 347 times.
✓ Branch 1 taken 1 times.
✓ Branch 2 taken 202 times.
✓ Branch 3 taken 145 times.
✓ Branch 4 taken 200 times.
✓ Branch 5 taken 2 times.
348 } else if (tagname[0] && !tagname[1] && strchr("bisu", av_tolower(tagname[0]))) {
285 200 av_bprintf(dst, "{\\%c%d}", (char)av_tolower(tagname[0]), !tag_close);
286 200 in += skip;
287
2/2
✓ Branch 1 taken 6 times.
✓ Branch 2 taken 142 times.
148 } else if (!av_strncasecmp(tagname, "br", 2) &&
288
5/6
✓ Branch 0 taken 3 times.
✓ Branch 1 taken 3 times.
✓ Branch 2 taken 1 times.
✓ Branch 3 taken 2 times.
✓ Branch 4 taken 1 times.
✗ Branch 5 not taken.
6 (!tagname[2] || (tagname[2] == '/' && !tagname[3]))) {
289 4 av_bprintf(dst, "\\N");
290 4 in += skip;
291
2/2
✓ Branch 0 taken 129 times.
✓ Branch 1 taken 15 times.
144 } else if (likely_a_tag) {
292
2/2
✓ Branch 0 taken 69 times.
✓ Branch 1 taken 60 times.
129 if (!tag_close) // warn only once
293 69 av_log(log_ctx, AV_LOG_WARNING, "Unrecognized tag %s\n", tagname);
294 129 in += skip;
295 } else {
296 15 av_bprint_chars(dst, '<', 1);
297 }
298 } else {
299 20 av_bprint_chars(dst, *in, 1);
300 }
301 626 break;
302 21829 default:
303 21829 av_bprint_chars(dst, *in, 1);
304 21829 break;
305 }
306
5/6
✓ Branch 0 taken 22946 times.
✓ Branch 1 taken 3767 times.
✓ Branch 2 taken 22946 times.
✗ Branch 3 not taken.
✓ Branch 4 taken 22551 times.
✓ Branch 5 taken 395 times.
26713 if (*in != ' ' && *in != '\r' && *in != '\n')
307 22551 line_start = 0;
308 }
309
310
1/2
✗ Branch 1 not taken.
✓ Branch 2 taken 337 times.
337 if (!av_bprint_is_complete(dst))
311 return AVERROR(ENOMEM);
312
313
3/4
✓ Branch 0 taken 336 times.
✓ Branch 1 taken 1 times.
✗ Branch 2 not taken.
✓ Branch 3 taken 336 times.
337 while (dst->len >= 2 && !strncmp(&dst->str[dst->len - 2], "\\N", 2))
314 dst->len -= 2;
315 337 dst->str[dst->len] = 0;
316 337 rstrip_spaces_buf(dst);
317
318 337 return 0;
319 }
320