0f3a5fefac58bd92cc2fbc8e02bc31af402f1f24
[libav.git] / libavformat / mp3.c
1 /*
2 * MP3 muxer and demuxer
3 * Copyright (c) 2003 Fabrice Bellard.
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 <strings.h>
23 #include "libavutil/avstring.h"
24 #include "libavcodec/mpegaudio.h"
25 #include "libavcodec/mpegaudiodecheader.h"
26 #include "avformat.h"
27
28 #define ID3v2_HEADER_SIZE 10
29 #define ID3v1_TAG_SIZE 128
30
31 #define ID3v1_GENRE_MAX 125
32
33 static const char * const id3v1_genre_str[ID3v1_GENRE_MAX + 1] = {
34 [0] = "Blues",
35 [1] = "Classic Rock",
36 [2] = "Country",
37 [3] = "Dance",
38 [4] = "Disco",
39 [5] = "Funk",
40 [6] = "Grunge",
41 [7] = "Hip-Hop",
42 [8] = "Jazz",
43 [9] = "Metal",
44 [10] = "New Age",
45 [11] = "Oldies",
46 [12] = "Other",
47 [13] = "Pop",
48 [14] = "R&B",
49 [15] = "Rap",
50 [16] = "Reggae",
51 [17] = "Rock",
52 [18] = "Techno",
53 [19] = "Industrial",
54 [20] = "Alternative",
55 [21] = "Ska",
56 [22] = "Death Metal",
57 [23] = "Pranks",
58 [24] = "Soundtrack",
59 [25] = "Euro-Techno",
60 [26] = "Ambient",
61 [27] = "Trip-Hop",
62 [28] = "Vocal",
63 [29] = "Jazz+Funk",
64 [30] = "Fusion",
65 [31] = "Trance",
66 [32] = "Classical",
67 [33] = "Instrumental",
68 [34] = "Acid",
69 [35] = "House",
70 [36] = "Game",
71 [37] = "Sound Clip",
72 [38] = "Gospel",
73 [39] = "Noise",
74 [40] = "AlternRock",
75 [41] = "Bass",
76 [42] = "Soul",
77 [43] = "Punk",
78 [44] = "Space",
79 [45] = "Meditative",
80 [46] = "Instrumental Pop",
81 [47] = "Instrumental Rock",
82 [48] = "Ethnic",
83 [49] = "Gothic",
84 [50] = "Darkwave",
85 [51] = "Techno-Industrial",
86 [52] = "Electronic",
87 [53] = "Pop-Folk",
88 [54] = "Eurodance",
89 [55] = "Dream",
90 [56] = "Southern Rock",
91 [57] = "Comedy",
92 [58] = "Cult",
93 [59] = "Gangsta",
94 [60] = "Top 40",
95 [61] = "Christian Rap",
96 [62] = "Pop/Funk",
97 [63] = "Jungle",
98 [64] = "Native American",
99 [65] = "Cabaret",
100 [66] = "New Wave",
101 [67] = "Psychadelic",
102 [68] = "Rave",
103 [69] = "Showtunes",
104 [70] = "Trailer",
105 [71] = "Lo-Fi",
106 [72] = "Tribal",
107 [73] = "Acid Punk",
108 [74] = "Acid Jazz",
109 [75] = "Polka",
110 [76] = "Retro",
111 [77] = "Musical",
112 [78] = "Rock & Roll",
113 [79] = "Hard Rock",
114 [80] = "Folk",
115 [81] = "Folk-Rock",
116 [82] = "National Folk",
117 [83] = "Swing",
118 [84] = "Fast Fusion",
119 [85] = "Bebob",
120 [86] = "Latin",
121 [87] = "Revival",
122 [88] = "Celtic",
123 [89] = "Bluegrass",
124 [90] = "Avantgarde",
125 [91] = "Gothic Rock",
126 [92] = "Progressive Rock",
127 [93] = "Psychedelic Rock",
128 [94] = "Symphonic Rock",
129 [95] = "Slow Rock",
130 [96] = "Big Band",
131 [97] = "Chorus",
132 [98] = "Easy Listening",
133 [99] = "Acoustic",
134 [100] = "Humour",
135 [101] = "Speech",
136 [102] = "Chanson",
137 [103] = "Opera",
138 [104] = "Chamber Music",
139 [105] = "Sonata",
140 [106] = "Symphony",
141 [107] = "Booty Bass",
142 [108] = "Primus",
143 [109] = "Porn Groove",
144 [110] = "Satire",
145 [111] = "Slow Jam",
146 [112] = "Club",
147 [113] = "Tango",
148 [114] = "Samba",
149 [115] = "Folklore",
150 [116] = "Ballad",
151 [117] = "Power Ballad",
152 [118] = "Rhythmic Soul",
153 [119] = "Freestyle",
154 [120] = "Duet",
155 [121] = "Punk Rock",
156 [122] = "Drum Solo",
157 [123] = "A capella",
158 [124] = "Euro-House",
159 [125] = "Dance Hall",
160 };
161
162 /* buf must be ID3v2_HEADER_SIZE byte long */
163 static int id3v2_match(const uint8_t *buf)
164 {
165 return buf[0] == 'I' &&
166 buf[1] == 'D' &&
167 buf[2] == '3' &&
168 buf[3] != 0xff &&
169 buf[4] != 0xff &&
170 (buf[6] & 0x80) == 0 &&
171 (buf[7] & 0x80) == 0 &&
172 (buf[8] & 0x80) == 0 &&
173 (buf[9] & 0x80) == 0;
174 }
175
176 static unsigned int id3v2_get_size(ByteIOContext *s, int len)
177 {
178 int v=0;
179 while(len--)
180 v= (v<<7) + (get_byte(s)&0x7F);
181 return v;
182 }
183
184 static void id3v2_read_ttag(AVFormatContext *s, int taglen, char *dst, int dstlen)
185 {
186 char *q;
187 int len;
188
189 if(dstlen > 0)
190 dst[0]= 0;
191 if(taglen < 1)
192 return;
193
194 taglen--; /* account for encoding type byte */
195 dstlen--; /* Leave space for zero terminator */
196
197 switch(get_byte(s->pb)) { /* encoding type */
198
199 case 0: /* ISO-8859-1 (0 - 255 maps directly into unicode) */
200 q = dst;
201 while(taglen--) {
202 uint8_t tmp;
203 PUT_UTF8(get_byte(s->pb), tmp, if (q - dst < dstlen - 1) *q++ = tmp;)
204 }
205 *q = '\0';
206 break;
207
208 case 3: /* UTF-8 */
209 len = FFMIN(taglen, dstlen-1);
210 get_buffer(s->pb, dst, len);
211 dst[len] = 0;
212 break;
213 }
214 }
215
216 /**
217 * ID3v2 parser
218 *
219 * Handles ID3v2.2, 2.3 and 2.4.
220 *
221 */
222
223 static void id3v2_parse(AVFormatContext *s, int len, uint8_t version, uint8_t flags)
224 {
225 int isv34, tlen;
226 uint32_t tag;
227 int64_t next;
228 char tmp[16];
229 int taghdrlen;
230 const char *reason;
231
232 switch(version) {
233 case 2:
234 if(flags & 0x40) {
235 reason = "compression";
236 goto error;
237 }
238 isv34 = 0;
239 taghdrlen = 6;
240 break;
241
242 case 3:
243 case 4:
244 isv34 = 1;
245 taghdrlen = 10;
246 break;
247
248 default:
249 reason = "version";
250 goto error;
251 }
252
253 if(flags & 0x80) {
254 reason = "unsynchronization";
255 goto error;
256 }
257
258 if(isv34 && flags & 0x40) /* Extended header present, just skip over it */
259 url_fskip(s->pb, id3v2_get_size(s->pb, 4));
260
261 while(len >= taghdrlen) {
262 if(isv34) {
263 tag = get_be32(s->pb);
264 tlen = id3v2_get_size(s->pb, 4);
265 get_be16(s->pb); /* flags */
266 } else {
267 tag = get_be24(s->pb);
268 tlen = id3v2_get_size(s->pb, 3);
269 }
270 len -= taghdrlen + tlen;
271
272 if(len < 0)
273 break;
274
275 next = url_ftell(s->pb) + tlen;
276
277 switch(tag) {
278 case MKBETAG('T', 'I', 'T', '2'):
279 case MKBETAG(0, 'T', 'T', '2'):
280 id3v2_read_ttag(s, tlen, s->title, sizeof(s->title));
281 break;
282 case MKBETAG('T', 'P', 'E', '1'):
283 case MKBETAG(0, 'T', 'P', '1'):
284 id3v2_read_ttag(s, tlen, s->author, sizeof(s->author));
285 break;
286 case MKBETAG('T', 'A', 'L', 'B'):
287 case MKBETAG(0, 'T', 'A', 'L'):
288 id3v2_read_ttag(s, tlen, s->album, sizeof(s->album));
289 break;
290 case MKBETAG('T', 'C', 'O', 'N'):
291 case MKBETAG(0, 'T', 'C', 'O'):
292 id3v2_read_ttag(s, tlen, s->genre, sizeof(s->genre));
293 break;
294 case MKBETAG('T', 'C', 'O', 'P'):
295 case MKBETAG(0, 'T', 'C', 'R'):
296 id3v2_read_ttag(s, tlen, s->copyright, sizeof(s->copyright));
297 break;
298 case MKBETAG('T', 'R', 'C', 'K'):
299 case MKBETAG(0, 'T', 'R', 'K'):
300 id3v2_read_ttag(s, tlen, tmp, sizeof(tmp));
301 s->track = atoi(tmp);
302 break;
303 case 0:
304 /* padding, skip to end */
305 url_fskip(s->pb, len);
306 len = 0;
307 continue;
308 }
309 /* Skip to end of tag */
310 url_fseek(s->pb, next, SEEK_SET);
311 }
312
313 if(version == 4 && flags & 0x10) /* Footer preset, always 10 bytes, skip over it */
314 url_fskip(s->pb, 10);
315 return;
316
317 error:
318 av_log(s, AV_LOG_INFO, "ID3v2.%d tag skipped, cannot handle %s\n", version, reason);
319 url_fskip(s->pb, len);
320 }
321
322 static void id3v1_get_string(char *str, int str_size,
323 const uint8_t *buf, int buf_size)
324 {
325 int i, c;
326 char *q;
327
328 q = str;
329 for(i = 0; i < buf_size; i++) {
330 c = buf[i];
331 if (c == '\0')
332 break;
333 if ((q - str) >= str_size - 1)
334 break;
335 *q++ = c;
336 }
337 *q = '\0';
338 }
339
340 /* 'buf' must be ID3v1_TAG_SIZE byte long */
341 static int id3v1_parse_tag(AVFormatContext *s, const uint8_t *buf)
342 {
343 char str[5];
344 int genre;
345
346 if (!(buf[0] == 'T' &&
347 buf[1] == 'A' &&
348 buf[2] == 'G'))
349 return -1;
350 id3v1_get_string(s->title, sizeof(s->title), buf + 3, 30);
351 id3v1_get_string(s->author, sizeof(s->author), buf + 33, 30);
352 id3v1_get_string(s->album, sizeof(s->album), buf + 63, 30);
353 id3v1_get_string(str, sizeof(str), buf + 93, 4);
354 s->year = atoi(str);
355 id3v1_get_string(s->comment, sizeof(s->comment), buf + 97, 30);
356 if (buf[125] == 0 && buf[126] != 0)
357 s->track = buf[126];
358 genre = buf[127];
359 if (genre <= ID3v1_GENRE_MAX)
360 av_strlcpy(s->genre, id3v1_genre_str[genre], sizeof(s->genre));
361 return 0;
362 }
363
364 /* mp3 read */
365
366 static int mp3_read_probe(AVProbeData *p)
367 {
368 int max_frames, first_frames = 0;
369 int fsize, frames, sample_rate;
370 uint32_t header;
371 uint8_t *buf, *buf2, *end;
372 AVCodecContext avctx;
373
374 if(id3v2_match(p->buf))
375 return AVPROBE_SCORE_MAX/2+1; // this must be less than mpeg-ps because some retards put id3v2 tags before mpeg-ps files
376
377 max_frames = 0;
378 buf = p->buf;
379 end = buf + p->buf_size - sizeof(uint32_t);
380
381 for(; buf < end; buf= buf2+1) {
382 buf2 = buf;
383
384 for(frames = 0; buf2 < end; frames++) {
385 header = AV_RB32(buf2);
386 fsize = ff_mpa_decode_header(&avctx, header, &sample_rate, &sample_rate, &sample_rate, &sample_rate);
387 if(fsize < 0)
388 break;
389 buf2 += fsize;
390 }
391 max_frames = FFMAX(max_frames, frames);
392 if(buf == p->buf)
393 first_frames= frames;
394 }
395 if (first_frames>=3) return AVPROBE_SCORE_MAX/2+1;
396 else if(max_frames>500)return AVPROBE_SCORE_MAX/2;
397 else if(max_frames>=3) return AVPROBE_SCORE_MAX/4;
398 else if(max_frames>=1) return 1;
399 else return 0;
400 }
401
402 /**
403 * Try to find Xing/Info/VBRI tags and compute duration from info therein
404 */
405 static void mp3_parse_vbr_tags(AVFormatContext *s, AVStream *st, int64_t base)
406 {
407 uint32_t v, spf;
408 int frames = -1; /* Total number of frames in file */
409 const int64_t xing_offtbl[2][2] = {{32, 17}, {17,9}};
410 MPADecodeContext c;
411
412 v = get_be32(s->pb);
413 if(ff_mpa_check_header(v) < 0)
414 return;
415
416 ff_mpegaudio_decode_header(&c, v);
417 if(c.layer != 3)
418 return;
419
420 /* Check for Xing / Info tag */
421 url_fseek(s->pb, xing_offtbl[c.lsf == 1][c.nb_channels == 1], SEEK_CUR);
422 v = get_be32(s->pb);
423 if(v == MKBETAG('X', 'i', 'n', 'g') || v == MKBETAG('I', 'n', 'f', 'o')) {
424 v = get_be32(s->pb);
425 if(v & 0x1)
426 frames = get_be32(s->pb);
427 }
428
429 /* Check for VBRI tag (always 32 bytes after end of mpegaudio header) */
430 url_fseek(s->pb, base + 4 + 32, SEEK_SET);
431 v = get_be32(s->pb);
432 if(v == MKBETAG('V', 'B', 'R', 'I')) {
433 /* Check tag version */
434 if(get_be16(s->pb) == 1) {
435 /* skip delay, quality and total bytes */
436 url_fseek(s->pb, 8, SEEK_CUR);
437 frames = get_be32(s->pb);
438 }
439 }
440
441 if(frames < 0)
442 return;
443
444 spf = c.lsf ? 576 : 1152; /* Samples per frame, layer 3 */
445 st->duration = av_rescale_q(frames, (AVRational){spf, c.sample_rate},
446 st->time_base);
447 }
448
449 static int mp3_read_header(AVFormatContext *s,
450 AVFormatParameters *ap)
451 {
452 AVStream *st;
453 uint8_t buf[ID3v1_TAG_SIZE];
454 int len, ret, filesize;
455 int64_t off;
456
457 st = av_new_stream(s, 0);
458 if (!st)
459 return AVERROR(ENOMEM);
460
461 st->codec->codec_type = CODEC_TYPE_AUDIO;
462 st->codec->codec_id = CODEC_ID_MP3;
463 st->need_parsing = AVSTREAM_PARSE_FULL;
464 st->start_time = 0;
465
466 /* try to get the TAG */
467 if (!url_is_streamed(s->pb)) {
468 /* XXX: change that */
469 filesize = url_fsize(s->pb);
470 if (filesize > 128) {
471 url_fseek(s->pb, filesize - 128, SEEK_SET);
472 ret = get_buffer(s->pb, buf, ID3v1_TAG_SIZE);
473 if (ret == ID3v1_TAG_SIZE) {
474 id3v1_parse_tag(s, buf);
475 }
476 url_fseek(s->pb, 0, SEEK_SET);
477 }
478 }
479
480 /* if ID3v2 header found, skip it */
481 ret = get_buffer(s->pb, buf, ID3v2_HEADER_SIZE);
482 if (ret != ID3v2_HEADER_SIZE)
483 return -1;
484 if (id3v2_match(buf)) {
485 /* parse ID3v2 header */
486 len = ((buf[6] & 0x7f) << 21) |
487 ((buf[7] & 0x7f) << 14) |
488 ((buf[8] & 0x7f) << 7) |
489 (buf[9] & 0x7f);
490 id3v2_parse(s, len, buf[3], buf[5]);
491 } else {
492 url_fseek(s->pb, 0, SEEK_SET);
493 }
494
495 off = url_ftell(s->pb);
496 mp3_parse_vbr_tags(s, st, off);
497 url_fseek(s->pb, off, SEEK_SET);
498
499 /* the parameters will be extracted from the compressed bitstream */
500 return 0;
501 }
502
503 #define MP3_PACKET_SIZE 1024
504
505 static int mp3_read_packet(AVFormatContext *s, AVPacket *pkt)
506 {
507 int ret, size;
508 // AVStream *st = s->streams[0];
509
510 size= MP3_PACKET_SIZE;
511
512 ret= av_get_packet(s->pb, pkt, size);
513
514 pkt->stream_index = 0;
515 if (ret <= 0) {
516 return AVERROR(EIO);
517 }
518 /* note: we need to modify the packet size here to handle the last
519 packet */
520 pkt->size = ret;
521 return ret;
522 }
523
524 #if defined(CONFIG_MP2_MUXER) || defined(CONFIG_MP3_MUXER)
525 static void id3v1_create_tag(AVFormatContext *s, uint8_t *buf)
526 {
527 int v, i;
528
529 memset(buf, 0, ID3v1_TAG_SIZE); /* fail safe */
530 buf[0] = 'T';
531 buf[1] = 'A';
532 buf[2] = 'G';
533 strncpy(buf + 3, s->title, 30);
534 strncpy(buf + 33, s->author, 30);
535 strncpy(buf + 63, s->album, 30);
536 v = s->year;
537 if (v > 0) {
538 for(i = 0;i < 4; i++) {
539 buf[96 - i] = '0' + (v % 10);
540 v = v / 10;
541 }
542 }
543 strncpy(buf + 97, s->comment, 30);
544 if (s->track != 0) {
545 buf[125] = 0;
546 buf[126] = s->track;
547 }
548 for(i = 0; i <= ID3v1_GENRE_MAX; i++) {
549 if (!strcasecmp(s->genre, id3v1_genre_str[i])) {
550 buf[127] = i;
551 break;
552 }
553 }
554 }
555
556 /* simple formats */
557
558 static void id3v2_put_size(AVFormatContext *s, int size)
559 {
560 put_byte(s->pb, size >> 21 & 0x7f);
561 put_byte(s->pb, size >> 14 & 0x7f);
562 put_byte(s->pb, size >> 7 & 0x7f);
563 put_byte(s->pb, size & 0x7f);
564 }
565
566 static void id3v2_put_ttag(AVFormatContext *s, const char *string, uint32_t tag)
567 {
568 int len = strlen(string);
569 put_be32(s->pb, tag);
570 id3v2_put_size(s, len + 1);
571 put_be16(s->pb, 0);
572 put_byte(s->pb, 3); /* UTF-8 */
573 put_buffer(s->pb, string, len);
574 }
575
576
577 /**
578 * Write an ID3v2.4 header at beginning of stream
579 */
580
581 static int mp3_write_header(struct AVFormatContext *s)
582 {
583 int totlen = 0;
584 char tracktxt[10];
585 char yeartxt[10];
586
587 if(s->track)
588 snprintf(tracktxt, sizeof(tracktxt), "%d", s->track);
589 if(s->year)
590 snprintf( yeartxt, sizeof(yeartxt) , "%d", s->year );
591
592 if(s->title[0]) totlen += 11 + strlen(s->title);
593 if(s->author[0]) totlen += 11 + strlen(s->author);
594 if(s->album[0]) totlen += 11 + strlen(s->album);
595 if(s->genre[0]) totlen += 11 + strlen(s->genre);
596 if(s->copyright[0]) totlen += 11 + strlen(s->copyright);
597 if(s->track) totlen += 11 + strlen(tracktxt);
598 if(s->year) totlen += 11 + strlen(yeartxt);
599 if(!(s->streams[0]->codec->flags & CODEC_FLAG_BITEXACT))
600 totlen += strlen(LIBAVFORMAT_IDENT) + 11;
601
602 if(totlen == 0)
603 return 0;
604
605 put_be32(s->pb, MKBETAG('I', 'D', '3', 0x04)); /* ID3v2.4 */
606 put_byte(s->pb, 0);
607 put_byte(s->pb, 0); /* flags */
608
609 id3v2_put_size(s, totlen);
610
611 if(s->title[0]) id3v2_put_ttag(s, s->title, MKBETAG('T', 'I', 'T', '2'));
612 if(s->author[0]) id3v2_put_ttag(s, s->author, MKBETAG('T', 'P', 'E', '1'));
613 if(s->album[0]) id3v2_put_ttag(s, s->album, MKBETAG('T', 'A', 'L', 'B'));
614 if(s->genre[0]) id3v2_put_ttag(s, s->genre, MKBETAG('T', 'C', 'O', 'N'));
615 if(s->copyright[0]) id3v2_put_ttag(s, s->copyright, MKBETAG('T', 'C', 'O', 'P'));
616 if(s->track) id3v2_put_ttag(s, tracktxt, MKBETAG('T', 'R', 'C', 'K'));
617 if(s->year) id3v2_put_ttag(s, yeartxt, MKBETAG('T', 'Y', 'E', 'R'));
618 if(!(s->streams[0]->codec->flags & CODEC_FLAG_BITEXACT))
619 id3v2_put_ttag(s, LIBAVFORMAT_IDENT, MKBETAG('T', 'E', 'N', 'C'));
620 return 0;
621 }
622
623 static int mp3_write_packet(struct AVFormatContext *s, AVPacket *pkt)
624 {
625 put_buffer(s->pb, pkt->data, pkt->size);
626 put_flush_packet(s->pb);
627 return 0;
628 }
629
630 static int mp3_write_trailer(struct AVFormatContext *s)
631 {
632 uint8_t buf[ID3v1_TAG_SIZE];
633
634 /* write the id3v1 tag */
635 if (s->title[0] != '\0') {
636 id3v1_create_tag(s, buf);
637 put_buffer(s->pb, buf, ID3v1_TAG_SIZE);
638 put_flush_packet(s->pb);
639 }
640 return 0;
641 }
642 #endif /* defined(CONFIG_MP2_MUXER) || defined(CONFIG_MP3_MUXER) */
643
644 #ifdef CONFIG_MP3_DEMUXER
645 AVInputFormat mp3_demuxer = {
646 "mp3",
647 NULL_IF_CONFIG_SMALL("MPEG audio"),
648 0,
649 mp3_read_probe,
650 mp3_read_header,
651 mp3_read_packet,
652 .flags= AVFMT_GENERIC_INDEX,
653 .extensions = "mp2,mp3,m2a", /* XXX: use probe */
654 };
655 #endif
656 #ifdef CONFIG_MP2_MUXER
657 AVOutputFormat mp2_muxer = {
658 "mp2",
659 NULL_IF_CONFIG_SMALL("MPEG audio layer 2"),
660 "audio/x-mpeg",
661 #ifdef CONFIG_LIBMP3LAME
662 "mp2,m2a",
663 #else
664 "mp2,mp3,m2a",
665 #endif
666 0,
667 CODEC_ID_MP2,
668 CODEC_ID_NONE,
669 NULL,
670 mp3_write_packet,
671 mp3_write_trailer,
672 };
673 #endif
674 #ifdef CONFIG_MP3_MUXER
675 AVOutputFormat mp3_muxer = {
676 "mp3",
677 NULL_IF_CONFIG_SMALL("MPEG audio layer 3"),
678 "audio/x-mpeg",
679 "mp3",
680 0,
681 CODEC_ID_MP3,
682 CODEC_ID_NONE,
683 mp3_write_header,
684 mp3_write_packet,
685 mp3_write_trailer,
686 };
687 #endif