Commit | Line | Data |
---|---|---|
2ba45a60 DM |
1 | /* |
2 | * TED Talks captions format decoder | |
3 | * Copyright (c) 2012 Nicolas George | |
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/bprint.h" | |
23 | #include "libavutil/log.h" | |
24 | #include "libavutil/opt.h" | |
25 | #include "avformat.h" | |
26 | #include "internal.h" | |
27 | #include "subtitles.h" | |
28 | ||
29 | typedef struct { | |
30 | AVClass *class; | |
31 | int64_t start_time; | |
32 | FFDemuxSubtitlesQueue subs; | |
33 | } TEDCaptionsDemuxer; | |
34 | ||
35 | static const AVOption tedcaptions_options[] = { | |
36 | { "start_time", "set the start time (offset) of the subtitles, in ms", | |
37 | offsetof(TEDCaptionsDemuxer, start_time), FF_OPT_TYPE_INT64, | |
38 | { .i64 = 15000 }, INT64_MIN, INT64_MAX, | |
39 | AV_OPT_FLAG_SUBTITLE_PARAM | AV_OPT_FLAG_DECODING_PARAM }, | |
40 | { NULL }, | |
41 | }; | |
42 | ||
43 | static const AVClass tedcaptions_demuxer_class = { | |
44 | .class_name = "tedcaptions_demuxer", | |
45 | .item_name = av_default_item_name, | |
46 | .option = tedcaptions_options, | |
47 | .version = LIBAVUTIL_VERSION_INT, | |
48 | }; | |
49 | ||
50 | #define BETWEEN(a, amin, amax) ((unsigned)((a) - (amin)) <= (amax) - (amin)) | |
51 | ||
52 | #define HEX_DIGIT_TEST(c) (BETWEEN(c, '0', '9') || BETWEEN((c) | 32, 'a', 'z')) | |
53 | #define HEX_DIGIT_VAL(c) ((c) <= '9' ? (c) - '0' : ((c) | 32) - 'a' + 10) | |
54 | #define ERR_CODE(c) (c < 0 ? c : AVERROR_INVALIDDATA) | |
55 | ||
56 | static void av_bprint_utf8(AVBPrint *bp, unsigned c) | |
57 | { | |
58 | int bytes, i; | |
59 | ||
60 | if (c <= 0x7F) { | |
61 | av_bprint_chars(bp, c, 1); | |
62 | return; | |
63 | } | |
64 | bytes = (av_log2(c) - 2) / 5; | |
65 | av_bprint_chars(bp, (c >> (bytes * 6)) | ((0xFF80 >> bytes) & 0xFF), 1); | |
66 | for (i = bytes - 1; i >= 0; i--) | |
67 | av_bprint_chars(bp, ((c >> (i * 6)) & 0x3F) | 0x80, 1); | |
68 | } | |
69 | ||
70 | static void next_byte(AVIOContext *pb, int *cur_byte) | |
71 | { | |
72 | uint8_t b; | |
73 | int ret = avio_read(pb, &b, 1); | |
74 | *cur_byte = ret > 0 ? b : ret == 0 ? AVERROR_EOF : ret; | |
75 | } | |
76 | ||
77 | static void skip_spaces(AVIOContext *pb, int *cur_byte) | |
78 | { | |
79 | while (*cur_byte == ' ' || *cur_byte == '\t' || | |
80 | *cur_byte == '\n' || *cur_byte == '\r') | |
81 | next_byte(pb, cur_byte); | |
82 | } | |
83 | ||
84 | static int expect_byte(AVIOContext *pb, int *cur_byte, uint8_t c) | |
85 | { | |
86 | skip_spaces(pb, cur_byte); | |
87 | if (*cur_byte != c) | |
88 | return ERR_CODE(*cur_byte); | |
89 | next_byte(pb, cur_byte); | |
90 | return 0; | |
91 | } | |
92 | ||
93 | static int parse_string(AVIOContext *pb, int *cur_byte, AVBPrint *bp, int full) | |
94 | { | |
95 | int ret; | |
96 | ||
97 | av_bprint_init(bp, 0, full ? -1 : 1); | |
98 | ret = expect_byte(pb, cur_byte, '"'); | |
99 | if (ret < 0) | |
100 | goto fail; | |
101 | while (*cur_byte > 0 && *cur_byte != '"') { | |
102 | if (*cur_byte == '\\') { | |
103 | next_byte(pb, cur_byte); | |
104 | if (*cur_byte < 0) { | |
105 | ret = AVERROR_INVALIDDATA; | |
106 | goto fail; | |
107 | } | |
108 | if ((*cur_byte | 32) == 'u') { | |
109 | unsigned chr = 0, i; | |
110 | for (i = 0; i < 4; i++) { | |
111 | next_byte(pb, cur_byte); | |
112 | if (!HEX_DIGIT_TEST(*cur_byte)) { | |
113 | ret = ERR_CODE(*cur_byte); | |
114 | goto fail; | |
115 | } | |
116 | chr = chr * 16 + HEX_DIGIT_VAL(*cur_byte); | |
117 | } | |
118 | av_bprint_utf8(bp, chr); | |
119 | } else { | |
120 | av_bprint_chars(bp, *cur_byte, 1); | |
121 | } | |
122 | } else { | |
123 | av_bprint_chars(bp, *cur_byte, 1); | |
124 | } | |
125 | next_byte(pb, cur_byte); | |
126 | } | |
127 | ret = expect_byte(pb, cur_byte, '"'); | |
128 | if (ret < 0) | |
129 | goto fail; | |
130 | if (full && !av_bprint_is_complete(bp)) { | |
131 | ret = AVERROR(ENOMEM); | |
132 | goto fail; | |
133 | } | |
134 | return 0; | |
135 | ||
136 | fail: | |
137 | av_bprint_finalize(bp, NULL); | |
138 | return ret; | |
139 | } | |
140 | ||
141 | static int parse_label(AVIOContext *pb, int *cur_byte, AVBPrint *bp) | |
142 | { | |
143 | int ret; | |
144 | ||
145 | ret = parse_string(pb, cur_byte, bp, 0); | |
146 | if (ret < 0) | |
147 | return ret; | |
148 | ret = expect_byte(pb, cur_byte, ':'); | |
149 | if (ret < 0) | |
150 | return ret; | |
151 | return 0; | |
152 | } | |
153 | ||
154 | static int parse_boolean(AVIOContext *pb, int *cur_byte, int *result) | |
155 | { | |
156 | static const char * const text[] = { "false", "true" }; | |
157 | const char *p; | |
158 | int i; | |
159 | ||
160 | skip_spaces(pb, cur_byte); | |
161 | for (i = 0; i < 2; i++) { | |
162 | p = text[i]; | |
163 | if (*cur_byte != *p) | |
164 | continue; | |
165 | for (; *p; p++, next_byte(pb, cur_byte)) | |
166 | if (*cur_byte != *p) | |
167 | return AVERROR_INVALIDDATA; | |
168 | if (BETWEEN(*cur_byte | 32, 'a', 'z')) | |
169 | return AVERROR_INVALIDDATA; | |
170 | *result = i; | |
171 | return 0; | |
172 | } | |
173 | return AVERROR_INVALIDDATA; | |
174 | } | |
175 | ||
176 | static int parse_int(AVIOContext *pb, int *cur_byte, int64_t *result) | |
177 | { | |
178 | int64_t val = 0; | |
179 | ||
180 | skip_spaces(pb, cur_byte); | |
181 | if ((unsigned)*cur_byte - '0' > 9) | |
182 | return AVERROR_INVALIDDATA; | |
183 | while (BETWEEN(*cur_byte, '0', '9')) { | |
184 | val = val * 10 + (*cur_byte - '0'); | |
185 | next_byte(pb, cur_byte); | |
186 | } | |
187 | *result = val; | |
188 | return 0; | |
189 | } | |
190 | ||
191 | static int parse_file(AVIOContext *pb, FFDemuxSubtitlesQueue *subs) | |
192 | { | |
193 | int ret, cur_byte, start_of_par; | |
194 | AVBPrint label, content; | |
195 | int64_t pos, start, duration; | |
196 | AVPacket *pkt; | |
197 | ||
198 | next_byte(pb, &cur_byte); | |
199 | ret = expect_byte(pb, &cur_byte, '{'); | |
200 | if (ret < 0) | |
201 | return AVERROR_INVALIDDATA; | |
202 | ret = parse_label(pb, &cur_byte, &label); | |
203 | if (ret < 0 || strcmp(label.str, "captions")) | |
204 | return AVERROR_INVALIDDATA; | |
205 | ret = expect_byte(pb, &cur_byte, '['); | |
206 | if (ret < 0) | |
207 | return AVERROR_INVALIDDATA; | |
208 | while (1) { | |
209 | content.size = 0; | |
210 | start = duration = AV_NOPTS_VALUE; | |
211 | ret = expect_byte(pb, &cur_byte, '{'); | |
212 | if (ret < 0) | |
213 | return ret; | |
214 | pos = avio_tell(pb) - 1; | |
215 | while (1) { | |
216 | ret = parse_label(pb, &cur_byte, &label); | |
217 | if (ret < 0) | |
218 | return ret; | |
219 | if (!strcmp(label.str, "startOfParagraph")) { | |
220 | ret = parse_boolean(pb, &cur_byte, &start_of_par); | |
221 | if (ret < 0) | |
222 | return ret; | |
223 | } else if (!strcmp(label.str, "content")) { | |
224 | ret = parse_string(pb, &cur_byte, &content, 1); | |
225 | if (ret < 0) | |
226 | return ret; | |
227 | } else if (!strcmp(label.str, "startTime")) { | |
228 | ret = parse_int(pb, &cur_byte, &start); | |
229 | if (ret < 0) | |
230 | return ret; | |
231 | } else if (!strcmp(label.str, "duration")) { | |
232 | ret = parse_int(pb, &cur_byte, &duration); | |
233 | if (ret < 0) | |
234 | return ret; | |
235 | } else { | |
236 | return AVERROR_INVALIDDATA; | |
237 | } | |
238 | skip_spaces(pb, &cur_byte); | |
239 | if (cur_byte != ',') | |
240 | break; | |
241 | next_byte(pb, &cur_byte); | |
242 | } | |
243 | ret = expect_byte(pb, &cur_byte, '}'); | |
244 | if (ret < 0) | |
245 | return ret; | |
246 | ||
247 | if (!content.size || start == AV_NOPTS_VALUE || | |
248 | duration == AV_NOPTS_VALUE) | |
249 | return AVERROR_INVALIDDATA; | |
250 | pkt = ff_subtitles_queue_insert(subs, content.str, content.len, 0); | |
251 | if (!pkt) | |
252 | return AVERROR(ENOMEM); | |
253 | pkt->pos = pos; | |
254 | pkt->pts = start; | |
255 | pkt->duration = duration; | |
256 | av_bprint_finalize(&content, NULL); | |
257 | ||
258 | skip_spaces(pb, &cur_byte); | |
259 | if (cur_byte != ',') | |
260 | break; | |
261 | next_byte(pb, &cur_byte); | |
262 | } | |
263 | ret = expect_byte(pb, &cur_byte, ']'); | |
264 | if (ret < 0) | |
265 | return ret; | |
266 | ret = expect_byte(pb, &cur_byte, '}'); | |
267 | if (ret < 0) | |
268 | return ret; | |
269 | skip_spaces(pb, &cur_byte); | |
270 | if (cur_byte != AVERROR_EOF) | |
271 | return ERR_CODE(cur_byte); | |
272 | return 0; | |
273 | } | |
274 | ||
275 | static av_cold int tedcaptions_read_header(AVFormatContext *avf) | |
276 | { | |
277 | TEDCaptionsDemuxer *tc = avf->priv_data; | |
278 | AVStream *st; | |
279 | int ret, i; | |
280 | AVPacket *last; | |
281 | ||
282 | ret = parse_file(avf->pb, &tc->subs); | |
283 | if (ret < 0) { | |
284 | if (ret == AVERROR_INVALIDDATA) | |
285 | av_log(avf, AV_LOG_ERROR, "Syntax error near offset %"PRId64".\n", | |
286 | avio_tell(avf->pb)); | |
287 | ff_subtitles_queue_clean(&tc->subs); | |
288 | return ret; | |
289 | } | |
290 | ff_subtitles_queue_finalize(&tc->subs); | |
291 | for (i = 0; i < tc->subs.nb_subs; i++) | |
292 | tc->subs.subs[i].pts += tc->start_time; | |
293 | ||
294 | last = &tc->subs.subs[tc->subs.nb_subs - 1]; | |
295 | st = avformat_new_stream(avf, NULL); | |
296 | if (!st) | |
297 | return AVERROR(ENOMEM); | |
298 | st->codec->codec_type = AVMEDIA_TYPE_SUBTITLE; | |
299 | st->codec->codec_id = AV_CODEC_ID_TEXT; | |
300 | avpriv_set_pts_info(st, 64, 1, 1000); | |
301 | st->probe_packets = 0; | |
302 | st->start_time = 0; | |
303 | st->duration = last->pts + last->duration; | |
304 | st->cur_dts = 0; | |
305 | ||
306 | return 0; | |
307 | } | |
308 | ||
309 | static int tedcaptions_read_packet(AVFormatContext *avf, AVPacket *packet) | |
310 | { | |
311 | TEDCaptionsDemuxer *tc = avf->priv_data; | |
312 | ||
313 | return ff_subtitles_queue_read_packet(&tc->subs, packet); | |
314 | } | |
315 | ||
316 | static int tedcaptions_read_close(AVFormatContext *avf) | |
317 | { | |
318 | TEDCaptionsDemuxer *tc = avf->priv_data; | |
319 | ||
320 | ff_subtitles_queue_clean(&tc->subs); | |
321 | return 0; | |
322 | } | |
323 | ||
324 | static av_cold int tedcaptions_read_probe(AVProbeData *p) | |
325 | { | |
326 | static const char *const tags[] = { | |
327 | "\"captions\"", "\"duration\"", "\"content\"", | |
328 | "\"startOfParagraph\"", "\"startTime\"", | |
329 | }; | |
330 | unsigned i, count = 0; | |
331 | const char *t; | |
332 | ||
333 | if (p->buf[strspn(p->buf, " \t\r\n")] != '{') | |
334 | return 0; | |
335 | for (i = 0; i < FF_ARRAY_ELEMS(tags); i++) { | |
336 | if (!(t = strstr(p->buf, tags[i]))) | |
337 | continue; | |
338 | t += strlen(tags[i]); | |
339 | t += strspn(t, " \t\r\n"); | |
340 | if (*t == ':') | |
341 | count++; | |
342 | } | |
343 | return count == FF_ARRAY_ELEMS(tags) ? AVPROBE_SCORE_MAX : | |
344 | count ? AVPROBE_SCORE_EXTENSION : 0; | |
345 | } | |
346 | ||
347 | static int tedcaptions_read_seek(AVFormatContext *avf, int stream_index, | |
348 | int64_t min_ts, int64_t ts, int64_t max_ts, | |
349 | int flags) | |
350 | { | |
351 | TEDCaptionsDemuxer *tc = avf->priv_data; | |
352 | return ff_subtitles_queue_seek(&tc->subs, avf, stream_index, | |
353 | min_ts, ts, max_ts, flags); | |
354 | } | |
355 | ||
356 | AVInputFormat ff_tedcaptions_demuxer = { | |
357 | .name = "tedcaptions", | |
358 | .long_name = NULL_IF_CONFIG_SMALL("TED Talks captions"), | |
359 | .priv_data_size = sizeof(TEDCaptionsDemuxer), | |
360 | .priv_class = &tedcaptions_demuxer_class, | |
361 | .read_header = tedcaptions_read_header, | |
362 | .read_packet = tedcaptions_read_packet, | |
363 | .read_close = tedcaptions_read_close, | |
364 | .read_probe = tedcaptions_read_probe, | |
365 | .read_seek2 = tedcaptions_read_seek, | |
366 | }; |