FFmpeg
 All Data Structures Namespaces Files Functions Variables Typedefs Enumerations Enumerator Macros Groups Pages
subtitles.c
Go to the documentation of this file.
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/avassert.h"
25 #include "libavutil/avstring.h"
26 
28 {
29  int i;
30  r->pb = pb;
31  r->buf_pos = r->buf_len = 0;
32  r->type = FF_UTF_8;
33  for (i = 0; i < 2; i++)
34  r->buf[r->buf_len++] = avio_r8(r->pb);
35  if (strncmp("\xFF\xFE", r->buf, 2) == 0) {
36  r->type = FF_UTF16LE;
37  r->buf_pos += 2;
38  } else if (strncmp("\xFE\xFF", r->buf, 2) == 0) {
39  r->type = FF_UTF16BE;
40  r->buf_pos += 2;
41  } else {
42  r->buf[r->buf_len++] = avio_r8(r->pb);
43  if (strncmp("\xEF\xBB\xBF", r->buf, 3) == 0) {
44  // UTF8
45  r->buf_pos += 3;
46  }
47  }
48  if (s && (r->type == FF_UTF16LE || r->type == FF_UTF16BE))
50  "UTF16 is automatically converted to UTF8, do not specify a character encoding\n");
51 }
52 
53 void ff_text_init_buf(FFTextReader *r, void *buf, size_t size)
54 {
55  memset(&r->buf_pb, 0, sizeof(r->buf_pb));
56  ffio_init_context(&r->buf_pb, buf, size, 0, NULL, NULL, NULL, NULL);
57  ff_text_init_avio(NULL, r, &r->buf_pb);
58 }
59 
61 {
62  return avio_tell(r->pb) - r->buf_len + r->buf_pos;
63 }
64 
66 {
67  uint32_t val;
68  uint8_t tmp;
69  if (r->buf_pos < r->buf_len)
70  return r->buf[r->buf_pos++];
71  if (r->type == FF_UTF16LE) {
72  GET_UTF16(val, avio_rl16(r->pb), return 0;)
73  } else if (r->type == FF_UTF16BE) {
74  GET_UTF16(val, avio_rb16(r->pb), return 0;)
75  } else {
76  return avio_r8(r->pb);
77  }
78  if (!val)
79  return 0;
80  r->buf_pos = 0;
81  r->buf_len = 0;
82  PUT_UTF8(val, tmp, r->buf[r->buf_len++] = tmp;)
83  return r->buf[r->buf_pos++]; // buf_len is at least 1
84 }
85 
86 void ff_text_read(FFTextReader *r, char *buf, size_t size)
87 {
88  for ( ; size > 0; size--)
89  *buf++ = ff_text_r8(r);
90 }
91 
93 {
94  return r->buf_pos >= r->buf_len && avio_feof(r->pb);
95 }
96 
98 {
99  int c;
100  if (r->buf_pos < r->buf_len)
101  return r->buf[r->buf_pos];
102  c = ff_text_r8(r);
103  if (!avio_feof(r->pb)) {
104  r->buf_pos = 0;
105  r->buf_len = 1;
106  r->buf[0] = c;
107  }
108  return c;
109 }
110 
112  const uint8_t *event, int len, int merge)
113 {
114  AVPacket *subs, *sub;
115 
116  if (merge && q->nb_subs > 0) {
117  /* merge with previous event */
118 
119  int old_len;
120  sub = &q->subs[q->nb_subs - 1];
121  old_len = sub->size;
122  if (av_grow_packet(sub, len) < 0)
123  return NULL;
124  memcpy(sub->data + old_len, event, len);
125  } else {
126  /* new event */
127 
128  if (q->nb_subs >= INT_MAX/sizeof(*q->subs) - 1)
129  return NULL;
130  subs = av_fast_realloc(q->subs, &q->allocated_size,
131  (q->nb_subs + 1) * sizeof(*q->subs));
132  if (!subs)
133  return NULL;
134  q->subs = subs;
135  sub = &subs[q->nb_subs++];
136  if (av_new_packet(sub, len) < 0)
137  return NULL;
138  sub->flags |= AV_PKT_FLAG_KEY;
139  sub->pts = sub->dts = 0;
140  memcpy(sub->data, event, len);
141  }
142  return sub;
143 }
144 
145 static int cmp_pkt_sub_ts_pos(const void *a, const void *b)
146 {
147  const AVPacket *s1 = a;
148  const AVPacket *s2 = b;
149  if (s1->pts == s2->pts) {
150  if (s1->pos == s2->pos)
151  return 0;
152  return s1->pos > s2->pos ? 1 : -1;
153  }
154  return s1->pts > s2->pts ? 1 : -1;
155 }
156 
157 static int cmp_pkt_sub_pos_ts(const void *a, const void *b)
158 {
159  const AVPacket *s1 = a;
160  const AVPacket *s2 = 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 
170 {
171  int i;
172 
173  qsort(q->subs, q->nb_subs, sizeof(*q->subs),
176  for (i = 0; i < q->nb_subs; i++)
177  if (q->subs[i].duration == -1 && i < q->nb_subs - 1)
178  q->subs[i].duration = q->subs[i + 1].pts - q->subs[i].pts;
179 }
180 
182 {
183  AVPacket *sub = q->subs + q->current_sub_idx;
184 
185  if (q->current_sub_idx == q->nb_subs)
186  return AVERROR_EOF;
187  if (av_copy_packet(pkt, sub) < 0) {
188  return AVERROR(ENOMEM);
189  }
190 
191  pkt->dts = pkt->pts;
192  q->current_sub_idx++;
193  return 0;
194 }
195 
196 static int search_sub_ts(const FFDemuxSubtitlesQueue *q, int64_t ts)
197 {
198  int s1 = 0, s2 = q->nb_subs - 1;
199 
200  if (s2 < s1)
201  return AVERROR(ERANGE);
202 
203  for (;;) {
204  int mid;
205 
206  if (s1 == s2)
207  return s1;
208  if (s1 == s2 - 1)
209  return q->subs[s1].pts <= q->subs[s2].pts ? s1 : s2;
210  mid = (s1 + s2) / 2;
211  if (q->subs[mid].pts <= ts)
212  s1 = mid;
213  else
214  s2 = mid;
215  }
216 }
217 
219  int64_t min_ts, int64_t ts, int64_t max_ts, int flags)
220 {
221  if (flags & AVSEEK_FLAG_BYTE) {
222  return AVERROR(ENOSYS);
223  } else if (flags & AVSEEK_FLAG_FRAME) {
224  if (ts < 0 || ts >= q->nb_subs)
225  return AVERROR(ERANGE);
226  q->current_sub_idx = ts;
227  } else {
228  int i, idx = search_sub_ts(q, ts);
229  int64_t ts_selected;
230 
231  if (idx < 0)
232  return idx;
233  for (i = idx; i < q->nb_subs && q->subs[i].pts < min_ts; i++)
234  if (stream_index == -1 || q->subs[i].stream_index == stream_index)
235  idx = i;
236  for (i = idx; i > 0 && q->subs[i].pts > max_ts; i--)
237  if (stream_index == -1 || q->subs[i].stream_index == stream_index)
238  idx = i;
239 
240  ts_selected = q->subs[idx].pts;
241  if (ts_selected < min_ts || ts_selected > max_ts)
242  return AVERROR(ERANGE);
243 
244  /* look back in the latest subtitles for overlapping subtitles */
245  for (i = idx - 1; i >= 0; i--) {
246  int64_t pts = q->subs[i].pts;
247  if (q->subs[i].duration <= 0 ||
248  (stream_index != -1 && q->subs[i].stream_index != stream_index))
249  continue;
250  if (pts >= min_ts && pts > ts_selected - q->subs[i].duration)
251  idx = i;
252  else
253  break;
254  }
255 
256  /* If the queue is used to store multiple subtitles streams (like with
257  * VobSub) and the stream index is not specified, we need to make sure
258  * to focus on the smallest file position offset for a same timestamp;
259  * queue is ordered by pts and then filepos, so we can take the first
260  * entry for a given timestamp. */
261  if (stream_index == -1)
262  while (idx > 0 && q->subs[idx - 1].pts == q->subs[idx].pts)
263  idx--;
264 
265  q->current_sub_idx = idx;
266  }
267  return 0;
268 }
269 
271 {
272  int i;
273 
274  for (i = 0; i < q->nb_subs; i++)
275  av_free_packet(&q->subs[i]);
276  av_freep(&q->subs);
277  q->nb_subs = q->allocated_size = q->current_sub_idx = 0;
278 }
279 
281 {
282  int i = 0;
283  char end_chr;
284 
285  if (!*c) // cached char?
286  *c = ff_text_r8(tr);
287  if (!*c)
288  return 0;
289 
290  end_chr = *c == '<' ? '>' : '<';
291  do {
292  av_bprint_chars(buf, *c, 1);
293  *c = ff_text_r8(tr);
294  i++;
295  } while (*c != end_chr && *c);
296  if (end_chr == '>') {
297  av_bprint_chars(buf, '>', 1);
298  *c = 0;
299  }
300  return i;
301 }
302 
303 const char *ff_smil_get_attr_ptr(const char *s, const char *attr)
304 {
305  int in_quotes = 0;
306  const int len = strlen(attr);
307 
308  while (*s) {
309  while (*s) {
310  if (!in_quotes && av_isspace(*s))
311  break;
312  in_quotes ^= *s == '"'; // XXX: support escaping?
313  s++;
314  }
315  while (av_isspace(*s))
316  s++;
317  if (!av_strncasecmp(s, attr, len) && s[len] == '=')
318  return s + len + 1 + (s[len + 1] == '"');
319  }
320  return NULL;
321 }
322 
323 static inline int is_eol(char c)
324 {
325  return c == '\r' || c == '\n';
326 }
327 
329 {
330  char eol_buf[5], last_was_cr = 0;
331  int n = 0, i = 0, nb_eol = 0;
332 
333  av_bprint_clear(buf);
334 
335  for (;;) {
336  char c = ff_text_r8(tr);
337 
338  if (!c)
339  break;
340 
341  /* ignore all initial line breaks */
342  if (n == 0 && is_eol(c))
343  continue;
344 
345  /* line break buffering: we don't want to add the trailing \r\n */
346  if (is_eol(c)) {
347  nb_eol += c == '\n' || last_was_cr;
348  if (nb_eol == 2)
349  break;
350  eol_buf[i++] = c;
351  if (i == sizeof(eol_buf) - 1)
352  break;
353  last_was_cr = c == '\r';
354  continue;
355  }
356 
357  /* only one line break followed by data: we flush the line breaks
358  * buffer */
359  if (i) {
360  eol_buf[i] = 0;
361  av_bprintf(buf, "%s", eol_buf);
362  i = nb_eol = 0;
363  }
364 
365  av_bprint_chars(buf, c, 1);
366  n++;
367  }
368 }
369 
371 {
372  FFTextReader tr;
373  tr.buf_pos = tr.buf_len = 0;
374  tr.type = 0;
375  tr.pb = pb;
377 }
378 
379 ptrdiff_t ff_subtitles_read_line(FFTextReader *tr, char *buf, size_t size)
380 {
381  size_t cur = 0;
382  if (!size)
383  return 0;
384  while (cur + 1 < size) {
385  unsigned char c = ff_text_r8(tr);
386  if (!c)
387  return ff_text_eof(tr) ? cur : AVERROR_INVALIDDATA;
388  if (c == '\r' || c == '\n')
389  break;
390  buf[cur++] = c;
391  buf[cur] = '\0';
392  }
393  if (ff_text_peek_r8(tr) == '\r')
394  ff_text_r8(tr);
395  if (ff_text_peek_r8(tr) == '\n')
396  ff_text_r8(tr);
397  return cur;
398 }