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, size_t 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  return FFDIFFSIGN(s1->pos, s2->pos);
151  return FFDIFFSIGN(s1->pts , s2->pts);
152 }
153 
154 static int cmp_pkt_sub_pos_ts(const void *a, const void *b)
155 {
156  const AVPacket *s1 = a;
157  const AVPacket *s2 = b;
158  if (s1->pos == s2->pos) {
159  if (s1->pts == s2->pts)
160  return 0;
161  return s1->pts > s2->pts ? 1 : -1;
162  }
163  return s1->pos > s2->pos ? 1 : -1;
164 }
165 
166 static void drop_dups(void *log_ctx, FFDemuxSubtitlesQueue *q)
167 {
168  int i, drop = 0;
169 
170  for (i = 1; i < q->nb_subs; i++) {
171  const int last_id = i - 1 - drop;
172  const AVPacket *last = &q->subs[last_id];
173 
174  if (q->subs[i].pts == last->pts &&
175  q->subs[i].duration == last->duration &&
176  q->subs[i].stream_index == last->stream_index &&
177  !strcmp(q->subs[i].data, last->data)) {
178 
179  av_packet_unref(&q->subs[i]);
180  drop++;
181  } else if (drop) {
182  q->subs[last_id + 1] = q->subs[i];
183  memset(&q->subs[i], 0, sizeof(q->subs[i])); // for safety
184  }
185  }
186 
187  if (drop) {
188  q->nb_subs -= drop;
189  av_log(log_ctx, AV_LOG_WARNING, "Dropping %d duplicated subtitle events\n", drop);
190  }
191 }
192 
194 {
195  int i;
196 
197  qsort(q->subs, q->nb_subs, sizeof(*q->subs),
200  for (i = 0; i < q->nb_subs; i++)
201  if (q->subs[i].duration < 0 && i < q->nb_subs - 1)
202  q->subs[i].duration = q->subs[i + 1].pts - q->subs[i].pts;
203 
204  if (!q->keep_duplicates)
205  drop_dups(log_ctx, q);
206 }
207 
209 {
210  AVPacket *sub = q->subs + q->current_sub_idx;
211 
212  if (q->current_sub_idx == q->nb_subs)
213  return AVERROR_EOF;
214  if (av_copy_packet(pkt, sub) < 0) {
215  return AVERROR(ENOMEM);
216  }
217 
218  pkt->dts = pkt->pts;
219  q->current_sub_idx++;
220  return 0;
221 }
222 
223 static int search_sub_ts(const FFDemuxSubtitlesQueue *q, int64_t ts)
224 {
225  int s1 = 0, s2 = q->nb_subs - 1;
226 
227  if (s2 < s1)
228  return AVERROR(ERANGE);
229 
230  for (;;) {
231  int mid;
232 
233  if (s1 == s2)
234  return s1;
235  if (s1 == s2 - 1)
236  return q->subs[s1].pts <= q->subs[s2].pts ? s1 : s2;
237  mid = (s1 + s2) / 2;
238  if (q->subs[mid].pts <= ts)
239  s1 = mid;
240  else
241  s2 = mid;
242  }
243 }
244 
246  int64_t min_ts, int64_t ts, int64_t max_ts, int flags)
247 {
248  if (flags & AVSEEK_FLAG_BYTE) {
249  return AVERROR(ENOSYS);
250  } else if (flags & AVSEEK_FLAG_FRAME) {
251  if (ts < 0 || ts >= q->nb_subs)
252  return AVERROR(ERANGE);
253  q->current_sub_idx = ts;
254  } else {
255  int i, idx = search_sub_ts(q, ts);
256  int64_t ts_selected;
257 
258  if (idx < 0)
259  return idx;
260  for (i = idx; i < q->nb_subs && q->subs[i].pts < min_ts; i++)
261  if (stream_index == -1 || q->subs[i].stream_index == stream_index)
262  idx = i;
263  for (i = idx; i > 0 && q->subs[i].pts > max_ts; i--)
264  if (stream_index == -1 || q->subs[i].stream_index == stream_index)
265  idx = i;
266 
267  ts_selected = q->subs[idx].pts;
268  if (ts_selected < min_ts || ts_selected > max_ts)
269  return AVERROR(ERANGE);
270 
271  /* look back in the latest subtitles for overlapping subtitles */
272  for (i = idx - 1; i >= 0; i--) {
273  int64_t pts = q->subs[i].pts;
274  if (q->subs[i].duration <= 0 ||
275  (stream_index != -1 && q->subs[i].stream_index != stream_index))
276  continue;
277  if (pts >= min_ts && pts > ts_selected - q->subs[i].duration)
278  idx = i;
279  else
280  break;
281  }
282 
283  /* If the queue is used to store multiple subtitles streams (like with
284  * VobSub) and the stream index is not specified, we need to make sure
285  * to focus on the smallest file position offset for a same timestamp;
286  * queue is ordered by pts and then filepos, so we can take the first
287  * entry for a given timestamp. */
288  if (stream_index == -1)
289  while (idx > 0 && q->subs[idx - 1].pts == q->subs[idx].pts)
290  idx--;
291 
292  q->current_sub_idx = idx;
293  }
294  return 0;
295 }
296 
298 {
299  int i;
300 
301  for (i = 0; i < q->nb_subs; i++)
302  av_packet_unref(&q->subs[i]);
303  av_freep(&q->subs);
304  q->nb_subs = q->allocated_size = q->current_sub_idx = 0;
305 }
306 
308 {
309  int i = 0;
310  char end_chr;
311 
312  if (!*c) // cached char?
313  *c = ff_text_r8(tr);
314  if (!*c)
315  return 0;
316 
317  end_chr = *c == '<' ? '>' : '<';
318  do {
319  av_bprint_chars(buf, *c, 1);
320  *c = ff_text_r8(tr);
321  i++;
322  } while (*c != end_chr && *c);
323  if (end_chr == '>') {
324  av_bprint_chars(buf, '>', 1);
325  *c = 0;
326  }
327  return i;
328 }
329 
330 const char *ff_smil_get_attr_ptr(const char *s, const char *attr)
331 {
332  int in_quotes = 0;
333  const size_t len = strlen(attr);
334 
335  while (*s) {
336  while (*s) {
337  if (!in_quotes && av_isspace(*s))
338  break;
339  in_quotes ^= *s == '"'; // XXX: support escaping?
340  s++;
341  }
342  while (av_isspace(*s))
343  s++;
344  if (!av_strncasecmp(s, attr, len) && s[len] == '=')
345  return s + len + 1 + (s[len + 1] == '"');
346  }
347  return NULL;
348 }
349 
350 static inline int is_eol(char c)
351 {
352  return c == '\r' || c == '\n';
353 }
354 
356 {
357  char eol_buf[5], last_was_cr = 0;
358  int n = 0, i = 0, nb_eol = 0;
359 
360  av_bprint_clear(buf);
361 
362  for (;;) {
363  char c = ff_text_r8(tr);
364 
365  if (!c)
366  break;
367 
368  /* ignore all initial line breaks */
369  if (n == 0 && is_eol(c))
370  continue;
371 
372  /* line break buffering: we don't want to add the trailing \r\n */
373  if (is_eol(c)) {
374  nb_eol += c == '\n' || last_was_cr;
375  if (nb_eol == 2)
376  break;
377  eol_buf[i++] = c;
378  if (i == sizeof(eol_buf) - 1)
379  break;
380  last_was_cr = c == '\r';
381  continue;
382  }
383 
384  /* only one line break followed by data: we flush the line breaks
385  * buffer */
386  if (i) {
387  eol_buf[i] = 0;
388  av_bprintf(buf, "%s", eol_buf);
389  i = nb_eol = 0;
390  }
391 
392  av_bprint_chars(buf, c, 1);
393  n++;
394  }
395 }
396 
398 {
399  FFTextReader tr;
400  tr.buf_pos = tr.buf_len = 0;
401  tr.type = 0;
402  tr.pb = pb;
404 }
405 
406 ptrdiff_t ff_subtitles_read_line(FFTextReader *tr, char *buf, size_t size)
407 {
408  size_t cur = 0;
409  if (!size)
410  return 0;
411  while (cur + 1 < size) {
412  unsigned char c = ff_text_r8(tr);
413  if (!c)
414  return ff_text_eof(tr) ? cur : AVERROR_INVALIDDATA;
415  if (c == '\r' || c == '\n')
416  break;
417  buf[cur++] = c;
418  buf[cur] = '\0';
419  }
420  if (ff_text_peek_r8(tr) == '\r')
421  ff_text_r8(tr);
422  if (ff_text_peek_r8(tr) == '\n')
423  ff_text_r8(tr);
424  return cur;
425 }
int keep_duplicates
set to 1 to keep duplicated subtitle events
Definition: subtitles.h:108
#define NULL
Definition: coverity.c:32
const char const char void * val
Definition: avisynth_c.h:771
const char * s
Definition: avisynth_c.h:768
Bytestream IO Context.
Definition: avio.h:147
#define AVERROR_INVALIDDATA
Invalid data found when processing input.
Definition: error.h:59
void av_bprintf(AVBPrint *buf, const char *fmt,...)
Definition: bprint.c:94
#define AV_LOG_WARNING
Something somehow does not look correct.
Definition: log.h:182
int64_t pos
byte position in stream, -1 if unknown
Definition: avcodec.h:1621
static int cmp_pkt_sub_pos_ts(const void *a, const void *b)
Definition: subtitles.c:154
static int search_sub_ts(const FFDemuxSubtitlesQueue *q, int64_t ts)
Definition: subtitles.c:223
void ff_subtitles_read_chunk(AVIOContext *pb, AVBPrint *buf)
Same as ff_subtitles_read_text_chunk(), but read from an AVIOContext.
Definition: subtitles.c:397
int size
Definition: avcodec.h:1602
const char * b
Definition: vf_curves.c:113
static av_const int av_isspace(int c)
Locale-independent conversion of ASCII isspace.
Definition: avstring.h:222
void ff_subtitles_queue_clean(FFDemuxSubtitlesQueue *q)
Remove and destroy all the subtitles packets.
Definition: subtitles.c:297
int av_strncasecmp(const char *a, const char *b, size_t n)
Locale-independent case-insensitive compare.
Definition: avstring.c:223
int ff_text_peek_r8(FFTextReader *r)
Like ff_text_r8(), but don't remove the byte from the buffer.
Definition: subtitles.c:97
#define PUT_UTF8(val, tmp, PUT_BYTE)
Convert a 32-bit Unicode character to its UTF-8 encoded form (up to 4 bytes long).
Definition: common.h:414
int allocated_size
allocated size for subs
Definition: subtitles.h:105
static AVPacket pkt
unsigned int avio_rb16(AVIOContext *s)
Definition: aviobuf.c:742
Format I/O context.
Definition: avformat.h:1338
const char * ff_smil_get_attr_ptr(const char *s, const char *attr)
SMIL helper to point on the value of an attribute in the given tag.
Definition: subtitles.c:330
uint8_t
int ff_subtitles_queue_read_packet(FFDemuxSubtitlesQueue *q, AVPacket *pkt)
Generic read_packet() callback for subtitles demuxers using this queue system.
Definition: subtitles.c:208
int64_t duration
Duration of this packet in AVStream->time_base units, 0 if unknown.
Definition: avcodec.h:1619
AVIOContext * pb
Definition: subtitles.h:42
enum sub_sort sort
sort method to use when finalizing subtitles
Definition: subtitles.h:107
uint8_t * data
Definition: avcodec.h:1601
#define AVERROR_EOF
End of file.
Definition: error.h:55
ptrdiff_t size
Definition: opengl_enc.c:101
static av_always_inline int64_t avio_tell(AVIOContext *s)
ftell() equivalent for AVIOContext.
Definition: avio.h:511
#define av_log(a,...)
#define AV_PKT_FLAG_KEY
The packet contains a keyframe.
Definition: avcodec.h:1633
#define GET_UTF16(val, GET_16BIT, ERROR)
Convert a UTF-16 character (2 or 4 bytes) to its 32-bit UCS-4 encoded form.
Definition: common.h:386
int av_new_packet(AVPacket *pkt, int size)
Allocate the payload of a packet and initialize its fields with default values.
Definition: avpacket.c:86
static void drop_dups(void *log_ctx, FFDemuxSubtitlesQueue *q)
Definition: subtitles.c:166
#define s2
Definition: regdef.h:39
static int is_eol(char c)
Definition: subtitles.c:350
int ff_text_eof(FFTextReader *r)
Return non-zero if EOF was reached.
Definition: subtitles.c:92
#define AVERROR(e)
Definition: error.h:43
ptrdiff_t ff_subtitles_read_line(FFTextReader *tr, char *buf, size_t size)
Read a line of text.
Definition: subtitles.c:406
const char * r
Definition: vf_curves.c:111
simple assert() macros that are a bit more flexible than ISO C assert().
void ff_text_init_avio(void *s, FFTextReader *r, AVIOContext *pb)
Initialize the FFTextReader from the given AVIOContext.
Definition: subtitles.c:27
int flags
A combination of AV_PKT_FLAG values.
Definition: avcodec.h:1607
unsigned char buf[8]
Definition: subtitles.h:43
int avio_r8(AVIOContext *s)
Definition: aviobuf.c:595
#define FFDIFFSIGN(x, y)
Comparator.
Definition: common.h:92
int64_t ff_text_pos(FFTextReader *r)
Return the byte position of the next byte returned by ff_text_r8().
Definition: subtitles.c:60
AVIOContext buf_pb
Definition: subtitles.h:45
static int cmp_pkt_sub_ts_pos(const void *a, const void *b)
Definition: subtitles.c:145
void * av_fast_realloc(void *ptr, unsigned int *size, size_t min_size)
Reallocate the given buffer if it is not large enough, otherwise do nothing.
Definition: mem.c:480
int n
Definition: avisynth_c.h:684
void ff_text_read(FFTextReader *r, char *buf, size_t size)
Read the given number of bytes (in UTF-8).
Definition: subtitles.c:86
int ff_text_r8(FFTextReader *r)
Return the next byte.
Definition: subtitles.c:65
#define AV_LOG_INFO
Standard information.
Definition: log.h:187
int ff_subtitles_queue_seek(FFDemuxSubtitlesQueue *q, AVFormatContext *s, int stream_index, int64_t min_ts, int64_t ts, int64_t max_ts, int flags)
Update current_sub_idx to emulate a seek.
Definition: subtitles.c:245
int ff_smil_extract_next_text_chunk(FFTextReader *tr, AVBPrint *buf, char *c)
SMIL helper to load next chunk ("<...>" or untagged content) in buf.
Definition: subtitles.c:307
void ff_subtitles_read_text_chunk(FFTextReader *tr, AVBPrint *buf)
Read a subtitles chunk from FFTextReader.
Definition: subtitles.c:355
void av_packet_unref(AVPacket *pkt)
Wipe the packet.
Definition: avpacket.c:567
int av_copy_packet(AVPacket *dst, const AVPacket *src)
Copy packet, including contents.
Definition: avpacket.c:264
void * buf
Definition: avisynth_c.h:690
#define AVSEEK_FLAG_BYTE
seeking based on position in bytes
Definition: avformat.h:2387
#define s1
Definition: regdef.h:38
static void merge(GetBitContext *gb, uint8_t *dst, uint8_t *src, int size)
Merge two consequent lists of equal size depending on bits read.
Definition: bink.c:217
static int64_t pts
Global timestamp for the audio frames.
static int flags
Definition: cpu.c:47
void av_bprint_clear(AVBPrint *buf)
Reset the string to "" but keep internal allocated data.
Definition: bprint.c:227
unsigned int avio_rl16(AVIOContext *s)
Definition: aviobuf.c:710
Main libavformat public API header.
int ffio_init_context(AVIOContext *s, unsigned char *buffer, int buffer_size, int write_flag, void *opaque, int(*read_packet)(void *opaque, uint8_t *buf, int buf_size), int(*write_packet)(void *opaque, uint8_t *buf, int buf_size), int64_t(*seek)(void *opaque, int64_t offset, int whence))
Definition: aviobuf.c:81
static double c[64]
AVPacket * subs
array of subtitles packets
Definition: subtitles.h:103
void ff_text_init_buf(FFTextReader *r, void *buf, size_t size)
Similar to ff_text_init_avio(), but sets it up to read from a bounded buffer.
Definition: subtitles.c:53
int av_grow_packet(AVPacket *pkt, int grow_by)
Increase packet size, correctly zeroing padding.
Definition: avpacket.c:109
int current_sub_idx
current position for the read packet callback
Definition: subtitles.h:106
int len
static uint8_t tmp[8]
Definition: des.c:38
#define AVSEEK_FLAG_FRAME
seeking based on frame number
Definition: avformat.h:2389
int64_t dts
Decompression timestamp in AVStream->time_base units; the time at which the packet is decompressed...
Definition: avcodec.h:1600
#define av_freep(p)
int avio_feof(AVIOContext *s)
feof() equivalent for AVIOContext.
Definition: aviobuf.c:328
int stream_index
Definition: avcodec.h:1603
AVPacket * ff_subtitles_queue_insert(FFDemuxSubtitlesQueue *q, const uint8_t *event, size_t len, int merge)
Insert a new subtitle event.
Definition: subtitles.c:111
This structure stores compressed data.
Definition: avcodec.h:1578
int nb_subs
number of subtitles packets
Definition: subtitles.h:104
sort by timestamps, then position
Definition: subtitles.h:30
void ff_subtitles_queue_finalize(void *log_ctx, FFDemuxSubtitlesQueue *q)
Set missing durations, sort subtitles by PTS (and then byte position), and drop duplicated events...
Definition: subtitles.c:193
int64_t pts
Presentation timestamp in AVStream->time_base units; the time at which the decompressed packet will b...
Definition: avcodec.h:1594
void av_bprint_chars(AVBPrint *buf, char c, unsigned n)
Append char c n times to a print buffer.
Definition: bprint.c:140