• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 /*
2  * Copyright (c) 2012-2013 Clément Bœsch <u pkh me>
3  *
4  * This file is part of FFmpeg.
5  *
6  * FFmpeg is free software; you can redistribute it and/or
7  * modify it under the terms of the GNU Lesser General Public
8  * License as published by the Free Software Foundation; either
9  * version 2.1 of the License, or (at your option) any later version.
10  *
11  * FFmpeg is distributed in the hope that it will be useful,
12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
14  * Lesser General Public License for more details.
15  *
16  * You should have received a copy of the GNU Lesser General Public
17  * License along with FFmpeg; if not, write to the Free Software
18  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19  */
20 
21 #include "avformat.h"
22 #include "subtitles.h"
23 #include "avio_internal.h"
24 #include "libavutil/avstring.h"
25 
ff_text_init_avio(void * s,FFTextReader * r,AVIOContext * pb)26 void ff_text_init_avio(void *s, FFTextReader *r, AVIOContext *pb)
27 {
28     int i;
29     r->pb = pb;
30     r->buf_pos = r->buf_len = 0;
31     r->type = FF_UTF_8;
32     for (i = 0; i < 2; i++)
33         r->buf[r->buf_len++] = avio_r8(r->pb);
34     if (strncmp("\xFF\xFE", r->buf, 2) == 0) {
35         r->type = FF_UTF16LE;
36         r->buf_pos += 2;
37     } else if (strncmp("\xFE\xFF", r->buf, 2) == 0) {
38         r->type = FF_UTF16BE;
39         r->buf_pos += 2;
40     } else {
41         r->buf[r->buf_len++] = avio_r8(r->pb);
42         if (strncmp("\xEF\xBB\xBF", r->buf, 3) == 0) {
43             // UTF8
44             r->buf_pos += 3;
45         }
46     }
47     if (s && (r->type == FF_UTF16LE || r->type == FF_UTF16BE))
48         av_log(s, AV_LOG_INFO,
49                "UTF16 is automatically converted to UTF8, do not specify a character encoding\n");
50 }
51 
ff_text_init_buf(FFTextReader * r,void * buf,size_t size)52 void ff_text_init_buf(FFTextReader *r, void *buf, size_t size)
53 {
54     ffio_init_context(&r->buf_pb, buf, size, 0, NULL, NULL, NULL, NULL);
55     ff_text_init_avio(NULL, r, &r->buf_pb.pub);
56 }
57 
ff_text_pos(FFTextReader * r)58 int64_t ff_text_pos(FFTextReader *r)
59 {
60     return avio_tell(r->pb) - r->buf_len + r->buf_pos;
61 }
62 
ff_text_r8(FFTextReader * r)63 int ff_text_r8(FFTextReader *r)
64 {
65     uint32_t val;
66     uint8_t tmp;
67     if (r->buf_pos < r->buf_len)
68         return r->buf[r->buf_pos++];
69     if (r->type == FF_UTF16LE) {
70         GET_UTF16(val, avio_rl16(r->pb), return 0;)
71     } else if (r->type == FF_UTF16BE) {
72         GET_UTF16(val, avio_rb16(r->pb), return 0;)
73     } else {
74         return avio_r8(r->pb);
75     }
76     if (!val)
77         return 0;
78     r->buf_pos = 0;
79     r->buf_len = 0;
80     PUT_UTF8(val, tmp, r->buf[r->buf_len++] = tmp;)
81     return r->buf[r->buf_pos++]; // buf_len is at least 1
82 }
83 
ff_text_read(FFTextReader * r,char * buf,size_t size)84 void ff_text_read(FFTextReader *r, char *buf, size_t size)
85 {
86     for ( ; size > 0; size--)
87         *buf++ = ff_text_r8(r);
88 }
89 
ff_text_eof(FFTextReader * r)90 int ff_text_eof(FFTextReader *r)
91 {
92     return r->buf_pos >= r->buf_len && avio_feof(r->pb);
93 }
94 
ff_text_peek_r8(FFTextReader * r)95 int ff_text_peek_r8(FFTextReader *r)
96 {
97     int c;
98     if (r->buf_pos < r->buf_len)
99         return r->buf[r->buf_pos];
100     c = ff_text_r8(r);
101     if (!avio_feof(r->pb)) {
102         r->buf_pos = 0;
103         r->buf_len = 1;
104         r->buf[0] = c;
105     }
106     return c;
107 }
108 
ff_subtitles_queue_insert(FFDemuxSubtitlesQueue * q,const uint8_t * event,size_t len,int merge)109 AVPacket *ff_subtitles_queue_insert(FFDemuxSubtitlesQueue *q,
110                                     const uint8_t *event, size_t len, int merge)
111 {
112     AVPacket **subs, *sub;
113 
114     if (merge && q->nb_subs > 0) {
115         /* merge with previous event */
116 
117         int old_len;
118         sub = q->subs[q->nb_subs - 1];
119         old_len = sub->size;
120         if (av_grow_packet(sub, len) < 0)
121             return NULL;
122         memcpy(sub->data + old_len, event, len);
123     } else {
124         /* new event */
125 
126         if (q->nb_subs >= INT_MAX/sizeof(*q->subs) - 1)
127             return NULL;
128         subs = av_fast_realloc(q->subs, &q->allocated_size,
129                                (q->nb_subs + 1) * sizeof(*q->subs));
130         if (!subs)
131             return NULL;
132         q->subs = subs;
133         sub = av_packet_alloc();
134         if (!sub)
135             return NULL;
136         if (av_new_packet(sub, len) < 0) {
137             av_packet_free(&sub);
138             return NULL;
139         }
140         subs[q->nb_subs++] = sub;
141         sub->flags |= AV_PKT_FLAG_KEY;
142         sub->pts = sub->dts = 0;
143         memcpy(sub->data, event, len);
144     }
145     return sub;
146 }
147 
cmp_pkt_sub_ts_pos(const void * a,const void * b)148 static int cmp_pkt_sub_ts_pos(const void *a, const void *b)
149 {
150     const AVPacket *s1 = *(const AVPacket **)a;
151     const AVPacket *s2 = *(const AVPacket **)b;
152     if (s1->pts == s2->pts)
153         return FFDIFFSIGN(s1->pos, s2->pos);
154     return FFDIFFSIGN(s1->pts , s2->pts);
155 }
156 
cmp_pkt_sub_pos_ts(const void * a,const void * b)157 static int cmp_pkt_sub_pos_ts(const void *a, const void *b)
158 {
159     const AVPacket *s1 = *(const AVPacket **)a;
160     const AVPacket *s2 = *(const AVPacket **)b;
161     if (s1->pos == s2->pos) {
162         if (s1->pts == s2->pts)
163             return 0;
164         return s1->pts > s2->pts ? 1 : -1;
165     }
166     return s1->pos > s2->pos ? 1 : -1;
167 }
168 
drop_dups(void * log_ctx,FFDemuxSubtitlesQueue * q)169 static void drop_dups(void *log_ctx, FFDemuxSubtitlesQueue *q)
170 {
171     int i, drop = 0;
172 
173     for (i = 1; i < q->nb_subs; i++) {
174         const int last_id = i - 1 - drop;
175         const AVPacket *last = q->subs[last_id];
176 
177         if (q->subs[i]->pts        == last->pts &&
178             q->subs[i]->duration   == last->duration &&
179             q->subs[i]->stream_index == last->stream_index &&
180             !strcmp(q->subs[i]->data, last->data)) {
181 
182             av_packet_free(&q->subs[i]);
183             drop++;
184         } else if (drop) {
185             q->subs[last_id + 1] = q->subs[i];
186             q->subs[i] = NULL;
187         }
188     }
189 
190     if (drop) {
191         q->nb_subs -= drop;
192         av_log(log_ctx, AV_LOG_WARNING, "Dropping %d duplicated subtitle events\n", drop);
193     }
194 }
195 
ff_subtitles_queue_finalize(void * log_ctx,FFDemuxSubtitlesQueue * q)196 void ff_subtitles_queue_finalize(void *log_ctx, FFDemuxSubtitlesQueue *q)
197 {
198     int i;
199 
200     if (!q->nb_subs)
201         return;
202 
203     qsort(q->subs, q->nb_subs, sizeof(*q->subs),
204           q->sort == SUB_SORT_TS_POS ? cmp_pkt_sub_ts_pos
205                                      : cmp_pkt_sub_pos_ts);
206     for (i = 0; i < q->nb_subs; i++)
207         if (q->subs[i]->duration < 0 && i < q->nb_subs - 1 && q->subs[i + 1]->pts - (uint64_t)q->subs[i]->pts <= INT64_MAX)
208             q->subs[i]->duration = q->subs[i + 1]->pts - q->subs[i]->pts;
209 
210     if (!q->keep_duplicates)
211         drop_dups(log_ctx, q);
212 }
213 
ff_subtitles_queue_read_packet(FFDemuxSubtitlesQueue * q,AVPacket * pkt)214 int ff_subtitles_queue_read_packet(FFDemuxSubtitlesQueue *q, AVPacket *pkt)
215 {
216     AVPacket *sub;
217     int ret;
218 
219     if (q->current_sub_idx == q->nb_subs)
220         return AVERROR_EOF;
221     sub = q->subs[q->current_sub_idx];
222     if ((ret = av_packet_ref(pkt, sub)) < 0) {
223         return ret;
224     }
225 
226     pkt->dts = pkt->pts;
227     q->current_sub_idx++;
228     return 0;
229 }
230 
search_sub_ts(const FFDemuxSubtitlesQueue * q,int64_t ts)231 static int search_sub_ts(const FFDemuxSubtitlesQueue *q, int64_t ts)
232 {
233     int s1 = 0, s2 = q->nb_subs - 1;
234 
235     if (s2 < s1)
236         return AVERROR(ERANGE);
237 
238     for (;;) {
239         int mid;
240 
241         if (s1 == s2)
242             return s1;
243         if (s1 == s2 - 1)
244             return q->subs[s1]->pts <= q->subs[s2]->pts ? s1 : s2;
245         mid = (s1 + s2) / 2;
246         if (q->subs[mid]->pts <= ts)
247             s1 = mid;
248         else
249             s2 = mid;
250     }
251 }
252 
ff_subtitles_queue_seek(FFDemuxSubtitlesQueue * q,AVFormatContext * s,int stream_index,int64_t min_ts,int64_t ts,int64_t max_ts,int flags)253 int ff_subtitles_queue_seek(FFDemuxSubtitlesQueue *q, AVFormatContext *s, int stream_index,
254                             int64_t min_ts, int64_t ts, int64_t max_ts, int flags)
255 {
256     if (flags & AVSEEK_FLAG_BYTE) {
257         return AVERROR(ENOSYS);
258     } else if (flags & AVSEEK_FLAG_FRAME) {
259         if (ts < 0 || ts >= q->nb_subs)
260             return AVERROR(ERANGE);
261         q->current_sub_idx = ts;
262     } else {
263         int i, idx = search_sub_ts(q, ts);
264         int64_t ts_selected;
265 
266         if (idx < 0)
267             return idx;
268         for (i = idx; i < q->nb_subs && q->subs[i]->pts < min_ts; i++)
269             if (stream_index == -1 || q->subs[i]->stream_index == stream_index)
270                 idx = i;
271         for (i = idx; i > 0 && q->subs[i]->pts > max_ts; i--)
272             if (stream_index == -1 || q->subs[i]->stream_index == stream_index)
273                 idx = i;
274 
275         ts_selected = q->subs[idx]->pts;
276         if (ts_selected < min_ts || ts_selected > max_ts)
277             return AVERROR(ERANGE);
278 
279         /* look back in the latest subtitles for overlapping subtitles */
280         for (i = idx - 1; i >= 0; i--) {
281             int64_t pts = q->subs[i]->pts;
282             if (q->subs[i]->duration <= 0 ||
283                 (stream_index != -1 && q->subs[i]->stream_index != stream_index))
284                 continue;
285             if (pts >= min_ts && pts > ts_selected - q->subs[i]->duration)
286                 idx = i;
287             else
288                 break;
289         }
290 
291         /* If the queue is used to store multiple subtitles streams (like with
292          * VobSub) and the stream index is not specified, we need to make sure
293          * to focus on the smallest file position offset for a same timestamp;
294          * queue is ordered by pts and then filepos, so we can take the first
295          * entry for a given timestamp. */
296         if (stream_index == -1)
297             while (idx > 0 && q->subs[idx - 1]->pts == q->subs[idx]->pts)
298                 idx--;
299 
300         q->current_sub_idx = idx;
301     }
302     return 0;
303 }
304 
ff_subtitles_queue_clean(FFDemuxSubtitlesQueue * q)305 void ff_subtitles_queue_clean(FFDemuxSubtitlesQueue *q)
306 {
307     int i;
308 
309     for (i = 0; i < q->nb_subs; i++)
310         av_packet_free(&q->subs[i]);
311     av_freep(&q->subs);
312     q->nb_subs = q->allocated_size = q->current_sub_idx = 0;
313 }
314 
ff_subtitles_read_packet(AVFormatContext * s,AVPacket * pkt)315 int ff_subtitles_read_packet(AVFormatContext *s, AVPacket *pkt)
316 {
317     FFDemuxSubtitlesQueue *q = s->priv_data;
318     return ff_subtitles_queue_read_packet(q, pkt);
319 }
320 
ff_subtitles_read_seek(AVFormatContext * s,int stream_index,int64_t min_ts,int64_t ts,int64_t max_ts,int flags)321 int ff_subtitles_read_seek(AVFormatContext *s, int stream_index,
322                            int64_t min_ts, int64_t ts, int64_t max_ts, int flags)
323 {
324     FFDemuxSubtitlesQueue *q = s->priv_data;
325     return ff_subtitles_queue_seek(q, s, stream_index,
326                                    min_ts, ts, max_ts, flags);
327 }
328 
ff_subtitles_read_close(AVFormatContext * s)329 int ff_subtitles_read_close(AVFormatContext *s)
330 {
331     FFDemuxSubtitlesQueue *q = s->priv_data;
332     ff_subtitles_queue_clean(q);
333     return 0;
334 }
335 
ff_smil_extract_next_text_chunk(FFTextReader * tr,AVBPrint * buf,char * c)336 int ff_smil_extract_next_text_chunk(FFTextReader *tr, AVBPrint *buf, char *c)
337 {
338     int i = 0;
339     char end_chr;
340 
341     if (!*c) // cached char?
342         *c = ff_text_r8(tr);
343     if (!*c)
344         return 0;
345 
346     end_chr = *c == '<' ? '>' : '<';
347     do {
348         av_bprint_chars(buf, *c, 1);
349         *c = ff_text_r8(tr);
350         i++;
351     } while (*c != end_chr && *c);
352     if (end_chr == '>') {
353         av_bprint_chars(buf, '>', 1);
354         *c = 0;
355     }
356     return i;
357 }
358 
ff_smil_get_attr_ptr(const char * s,const char * attr)359 const char *ff_smil_get_attr_ptr(const char *s, const char *attr)
360 {
361     int in_quotes = 0;
362     const size_t len = strlen(attr);
363 
364     while (*s) {
365         while (*s) {
366             if (!in_quotes && av_isspace(*s))
367                 break;
368             in_quotes ^= *s == '"'; // XXX: support escaping?
369             s++;
370         }
371         while (av_isspace(*s))
372             s++;
373         if (!av_strncasecmp(s, attr, len) && s[len] == '=')
374             return s + len + 1 + (s[len + 1] == '"');
375     }
376     return NULL;
377 }
378 
is_eol(char c)379 static inline int is_eol(char c)
380 {
381     return c == '\r' || c == '\n';
382 }
383 
ff_subtitles_read_text_chunk(FFTextReader * tr,AVBPrint * buf)384 void ff_subtitles_read_text_chunk(FFTextReader *tr, AVBPrint *buf)
385 {
386     char eol_buf[5], last_was_cr = 0;
387     int n = 0, i = 0, nb_eol = 0;
388 
389     av_bprint_clear(buf);
390 
391     for (;;) {
392         char c = ff_text_r8(tr);
393 
394         if (!c)
395             break;
396 
397         /* ignore all initial line breaks */
398         if (n == 0 && is_eol(c))
399             continue;
400 
401         /* line break buffering: we don't want to add the trailing \r\n */
402         if (is_eol(c)) {
403             nb_eol += c == '\n' || last_was_cr;
404             if (nb_eol == 2)
405                 break;
406             eol_buf[i++] = c;
407             if (i == sizeof(eol_buf) - 1)
408                 break;
409             last_was_cr = c == '\r';
410             continue;
411         }
412 
413         /* only one line break followed by data: we flush the line breaks
414          * buffer */
415         if (i) {
416             eol_buf[i] = 0;
417             av_bprintf(buf, "%s", eol_buf);
418             i = nb_eol = 0;
419         }
420 
421         av_bprint_chars(buf, c, 1);
422         n++;
423     }
424 }
425 
ff_subtitles_read_chunk(AVIOContext * pb,AVBPrint * buf)426 void ff_subtitles_read_chunk(AVIOContext *pb, AVBPrint *buf)
427 {
428     FFTextReader tr;
429     tr.buf_pos = tr.buf_len = 0;
430     tr.type = 0;
431     tr.pb = pb;
432     ff_subtitles_read_text_chunk(&tr, buf);
433 }
434 
ff_subtitles_read_line(FFTextReader * tr,char * buf,size_t size)435 ptrdiff_t ff_subtitles_read_line(FFTextReader *tr, char *buf, size_t size)
436 {
437     size_t cur = 0;
438     if (!size)
439         return 0;
440     buf[0] = '\0';
441     while (cur + 1 < size) {
442         unsigned char c = ff_text_r8(tr);
443         if (!c)
444             return ff_text_eof(tr) ? cur : AVERROR_INVALIDDATA;
445         if (c == '\r' || c == '\n')
446             break;
447         buf[cur++] = c;
448         buf[cur] = '\0';
449     }
450     while (ff_text_peek_r8(tr) == '\r')
451         ff_text_r8(tr);
452     if (ff_text_peek_r8(tr) == '\n')
453         ff_text_r8(tr);
454     return cur;
455 }
456