Proper handling of trailing whitespaces
[ffmpeg-lucabe.git] / libavformat / mp3.c
blob459bf23ea270902d75b0be8ad2f0982a5b7c1164
1 /*
2 * MP3 muxer and demuxer
3 * Copyright (c) 2003 Fabrice Bellard.
5 * This file is part of FFmpeg.
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.
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.
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
22 #include "libavutil/avstring.h"
23 #include "libavcodec/mpegaudio.h"
24 #include "libavcodec/mpegaudiodecheader.h"
25 #include "avformat.h"
27 #define ID3v2_HEADER_SIZE 10
28 #define ID3v1_TAG_SIZE 128
30 #define ID3v1_GENRE_MAX 125
32 static const char *id3v1_genre_str[ID3v1_GENRE_MAX + 1] = {
33 [0] = "Blues",
34 [1] = "Classic Rock",
35 [2] = "Country",
36 [3] = "Dance",
37 [4] = "Disco",
38 [5] = "Funk",
39 [6] = "Grunge",
40 [7] = "Hip-Hop",
41 [8] = "Jazz",
42 [9] = "Metal",
43 [10] = "New Age",
44 [11] = "Oldies",
45 [12] = "Other",
46 [13] = "Pop",
47 [14] = "R&B",
48 [15] = "Rap",
49 [16] = "Reggae",
50 [17] = "Rock",
51 [18] = "Techno",
52 [19] = "Industrial",
53 [20] = "Alternative",
54 [21] = "Ska",
55 [22] = "Death Metal",
56 [23] = "Pranks",
57 [24] = "Soundtrack",
58 [25] = "Euro-Techno",
59 [26] = "Ambient",
60 [27] = "Trip-Hop",
61 [28] = "Vocal",
62 [29] = "Jazz+Funk",
63 [30] = "Fusion",
64 [31] = "Trance",
65 [32] = "Classical",
66 [33] = "Instrumental",
67 [34] = "Acid",
68 [35] = "House",
69 [36] = "Game",
70 [37] = "Sound Clip",
71 [38] = "Gospel",
72 [39] = "Noise",
73 [40] = "AlternRock",
74 [41] = "Bass",
75 [42] = "Soul",
76 [43] = "Punk",
77 [44] = "Space",
78 [45] = "Meditative",
79 [46] = "Instrumental Pop",
80 [47] = "Instrumental Rock",
81 [48] = "Ethnic",
82 [49] = "Gothic",
83 [50] = "Darkwave",
84 [51] = "Techno-Industrial",
85 [52] = "Electronic",
86 [53] = "Pop-Folk",
87 [54] = "Eurodance",
88 [55] = "Dream",
89 [56] = "Southern Rock",
90 [57] = "Comedy",
91 [58] = "Cult",
92 [59] = "Gangsta",
93 [60] = "Top 40",
94 [61] = "Christian Rap",
95 [62] = "Pop/Funk",
96 [63] = "Jungle",
97 [64] = "Native American",
98 [65] = "Cabaret",
99 [66] = "New Wave",
100 [67] = "Psychadelic",
101 [68] = "Rave",
102 [69] = "Showtunes",
103 [70] = "Trailer",
104 [71] = "Lo-Fi",
105 [72] = "Tribal",
106 [73] = "Acid Punk",
107 [74] = "Acid Jazz",
108 [75] = "Polka",
109 [76] = "Retro",
110 [77] = "Musical",
111 [78] = "Rock & Roll",
112 [79] = "Hard Rock",
113 [80] = "Folk",
114 [81] = "Folk-Rock",
115 [82] = "National Folk",
116 [83] = "Swing",
117 [84] = "Fast Fusion",
118 [85] = "Bebob",
119 [86] = "Latin",
120 [87] = "Revival",
121 [88] = "Celtic",
122 [89] = "Bluegrass",
123 [90] = "Avantgarde",
124 [91] = "Gothic Rock",
125 [92] = "Progressive Rock",
126 [93] = "Psychedelic Rock",
127 [94] = "Symphonic Rock",
128 [95] = "Slow Rock",
129 [96] = "Big Band",
130 [97] = "Chorus",
131 [98] = "Easy Listening",
132 [99] = "Acoustic",
133 [100] = "Humour",
134 [101] = "Speech",
135 [102] = "Chanson",
136 [103] = "Opera",
137 [104] = "Chamber Music",
138 [105] = "Sonata",
139 [106] = "Symphony",
140 [107] = "Booty Bass",
141 [108] = "Primus",
142 [109] = "Porn Groove",
143 [110] = "Satire",
144 [111] = "Slow Jam",
145 [112] = "Club",
146 [113] = "Tango",
147 [114] = "Samba",
148 [115] = "Folklore",
149 [116] = "Ballad",
150 [117] = "Power Ballad",
151 [118] = "Rhythmic Soul",
152 [119] = "Freestyle",
153 [120] = "Duet",
154 [121] = "Punk Rock",
155 [122] = "Drum Solo",
156 [123] = "A capella",
157 [124] = "Euro-House",
158 [125] = "Dance Hall",
161 /* buf must be ID3v2_HEADER_SIZE byte long */
162 static int id3v2_match(const uint8_t *buf)
164 return buf[0] == 'I' &&
165 buf[1] == 'D' &&
166 buf[2] == '3' &&
167 buf[3] != 0xff &&
168 buf[4] != 0xff &&
169 (buf[6] & 0x80) == 0 &&
170 (buf[7] & 0x80) == 0 &&
171 (buf[8] & 0x80) == 0 &&
172 (buf[9] & 0x80) == 0;
175 static unsigned int id3v2_get_size(ByteIOContext *s, int len)
177 int v=0;
178 while(len--)
179 v= (v<<7) + (get_byte(s)&0x7F);
180 return v;
183 static void id3v2_read_ttag(AVFormatContext *s, int taglen, char *dst, int dstlen)
185 char *q;
186 int len;
188 if(taglen < 1)
189 return;
191 taglen--; /* account for encoding type byte */
192 dstlen--; /* Leave space for zero terminator */
194 switch(get_byte(s->pb)) { /* encoding type */
196 case 0: /* ISO-8859-1 (0 - 255 maps directly into unicode) */
197 q = dst;
198 while(taglen--) {
199 uint8_t tmp;
200 PUT_UTF8(get_byte(s->pb), tmp, if (q - dst < dstlen - 1) *q++ = tmp;)
202 *q = '\0';
203 break;
205 case 3: /* UTF-8 */
206 len = FFMIN(taglen, dstlen);
207 get_buffer(s->pb, dst, len);
208 dst[len] = 0;
209 break;
214 * ID3v2 parser
216 * Handles ID3v2.2, 2.3 and 2.4.
220 static void id3v2_parse(AVFormatContext *s, int len, uint8_t version, uint8_t flags)
222 int isv34, tlen;
223 uint32_t tag;
224 offset_t next;
225 char tmp[16];
226 int taghdrlen;
227 const char *reason;
229 switch(version) {
230 case 2:
231 if(flags & 0x40) {
232 reason = "compression";
233 goto error;
235 isv34 = 0;
236 taghdrlen = 6;
237 break;
239 case 3:
240 case 4:
241 isv34 = 1;
242 taghdrlen = 10;
243 break;
245 default:
246 reason = "version";
247 goto error;
250 if(flags & 0x80) {
251 reason = "unsynchronization";
252 goto error;
255 if(isv34 && flags & 0x40) /* Extended header present, just skip over it */
256 url_fskip(s->pb, id3v2_get_size(s->pb, 4));
258 while(len >= taghdrlen) {
259 if(isv34) {
260 tag = get_be32(s->pb);
261 tlen = id3v2_get_size(s->pb, 4);
262 get_be16(s->pb); /* flags */
263 } else {
264 tag = get_be24(s->pb);
265 tlen = id3v2_get_size(s->pb, 3);
267 len -= taghdrlen + tlen;
269 if(len < 0)
270 break;
272 next = url_ftell(s->pb) + tlen;
274 switch(tag) {
275 case MKBETAG('T', 'I', 'T', '2'):
276 case MKBETAG(0, 'T', 'T', '2'):
277 id3v2_read_ttag(s, tlen, s->title, sizeof(s->title));
278 break;
279 case MKBETAG('T', 'P', 'E', '1'):
280 case MKBETAG(0, 'T', 'P', '1'):
281 id3v2_read_ttag(s, tlen, s->author, sizeof(s->author));
282 break;
283 case MKBETAG('T', 'A', 'L', 'B'):
284 case MKBETAG(0, 'T', 'A', 'L'):
285 id3v2_read_ttag(s, tlen, s->album, sizeof(s->album));
286 break;
287 case MKBETAG('T', 'C', 'O', 'N'):
288 case MKBETAG(0, 'T', 'C', 'O'):
289 id3v2_read_ttag(s, tlen, s->genre, sizeof(s->genre));
290 break;
291 case MKBETAG('T', 'C', 'O', 'P'):
292 case MKBETAG(0, 'T', 'C', 'R'):
293 id3v2_read_ttag(s, tlen, s->copyright, sizeof(s->copyright));
294 break;
295 case MKBETAG('T', 'R', 'C', 'K'):
296 case MKBETAG(0, 'T', 'R', 'K'):
297 id3v2_read_ttag(s, tlen, tmp, sizeof(tmp));
298 s->track = atoi(tmp);
299 break;
300 case 0:
301 /* padding, skip to end */
302 url_fskip(s->pb, len);
303 len = 0;
304 continue;
306 /* Skip to end of tag */
307 url_fseek(s->pb, next, SEEK_SET);
310 if(version == 4 && flags & 0x10) /* Footer preset, always 10 bytes, skip over it */
311 url_fskip(s->pb, 10);
312 return;
314 error:
315 av_log(s, AV_LOG_INFO, "ID3v2.%d tag skipped, cannot handle %s\n", version, reason);
316 url_fskip(s->pb, len);
319 static void id3v1_get_string(char *str, int str_size,
320 const uint8_t *buf, int buf_size)
322 int i, c;
323 char *q;
325 q = str;
326 for(i = 0; i < buf_size; i++) {
327 c = buf[i];
328 if (c == '\0')
329 break;
330 if ((q - str) >= str_size - 1)
331 break;
332 *q++ = c;
334 *q = '\0';
337 /* 'buf' must be ID3v1_TAG_SIZE byte long */
338 static int id3v1_parse_tag(AVFormatContext *s, const uint8_t *buf)
340 char str[5];
341 int genre;
343 if (!(buf[0] == 'T' &&
344 buf[1] == 'A' &&
345 buf[2] == 'G'))
346 return -1;
347 id3v1_get_string(s->title, sizeof(s->title), buf + 3, 30);
348 id3v1_get_string(s->author, sizeof(s->author), buf + 33, 30);
349 id3v1_get_string(s->album, sizeof(s->album), buf + 63, 30);
350 id3v1_get_string(str, sizeof(str), buf + 93, 4);
351 s->year = atoi(str);
352 id3v1_get_string(s->comment, sizeof(s->comment), buf + 97, 30);
353 if (buf[125] == 0 && buf[126] != 0)
354 s->track = buf[126];
355 genre = buf[127];
356 if (genre <= ID3v1_GENRE_MAX)
357 av_strlcpy(s->genre, id3v1_genre_str[genre], sizeof(s->genre));
358 return 0;
361 static void id3v1_create_tag(AVFormatContext *s, uint8_t *buf)
363 int v, i;
365 memset(buf, 0, ID3v1_TAG_SIZE); /* fail safe */
366 buf[0] = 'T';
367 buf[1] = 'A';
368 buf[2] = 'G';
369 strncpy(buf + 3, s->title, 30);
370 strncpy(buf + 33, s->author, 30);
371 strncpy(buf + 63, s->album, 30);
372 v = s->year;
373 if (v > 0) {
374 for(i = 0;i < 4; i++) {
375 buf[96 - i] = '0' + (v % 10);
376 v = v / 10;
379 strncpy(buf + 97, s->comment, 30);
380 if (s->track != 0) {
381 buf[125] = 0;
382 buf[126] = s->track;
384 for(i = 0; i <= ID3v1_GENRE_MAX; i++) {
385 if (!strcasecmp(s->genre, id3v1_genre_str[i])) {
386 buf[127] = i;
387 break;
392 /* mp3 read */
394 static int mp3_read_probe(AVProbeData *p)
396 int max_frames, first_frames = 0;
397 int fsize, frames, sample_rate;
398 uint32_t header;
399 uint8_t *buf, *buf2, *end;
400 AVCodecContext avctx;
402 if(id3v2_match(p->buf))
403 return AVPROBE_SCORE_MAX/2+1; // this must be less than mpeg-ps because some retards put id3v2 tags before mpeg-ps files
405 max_frames = 0;
406 buf = p->buf;
407 end = buf + p->buf_size - sizeof(uint32_t);
409 for(; buf < end; buf= buf2+1) {
410 buf2 = buf;
412 for(frames = 0; buf2 < end; frames++) {
413 header = AV_RB32(buf2);
414 fsize = ff_mpa_decode_header(&avctx, header, &sample_rate);
415 if(fsize < 0)
416 break;
417 buf2 += fsize;
419 max_frames = FFMAX(max_frames, frames);
420 if(buf == p->buf)
421 first_frames= frames;
423 if (first_frames>=3) return AVPROBE_SCORE_MAX/2+1;
424 else if(max_frames>500)return AVPROBE_SCORE_MAX/2;
425 else if(max_frames>=3) return AVPROBE_SCORE_MAX/4;
426 else if(max_frames>=1) return 1;
427 else return 0;
431 * Try to find Xing/Info/VBRI tags and compute duration from info therein
433 static void mp3_parse_vbr_tags(AVFormatContext *s, AVStream *st, offset_t base)
435 uint32_t v, spf;
436 int frames = -1; /* Total number of frames in file */
437 const offset_t xing_offtbl[2][2] = {{32, 17}, {17,9}};
438 MPADecodeContext c;
440 v = get_be32(s->pb);
441 if(ff_mpa_check_header(v) < 0)
442 return;
444 ff_mpegaudio_decode_header(&c, v);
445 if(c.layer != 3)
446 return;
448 /* Check for Xing / Info tag */
449 url_fseek(s->pb, xing_offtbl[c.lsf == 1][c.nb_channels == 1], SEEK_CUR);
450 v = get_be32(s->pb);
451 if(v == MKBETAG('X', 'i', 'n', 'g') || v == MKBETAG('I', 'n', 'f', 'o')) {
452 v = get_be32(s->pb);
453 if(v & 0x1)
454 frames = get_be32(s->pb);
457 /* Check for VBRI tag (always 32 bytes after end of mpegaudio header) */
458 url_fseek(s->pb, base + 4 + 32, SEEK_SET);
459 v = get_be32(s->pb);
460 if(v == MKBETAG('V', 'B', 'R', 'I')) {
461 /* Check tag version */
462 if(get_be16(s->pb) == 1) {
463 /* skip delay, quality and total bytes */
464 url_fseek(s->pb, 8, SEEK_CUR);
465 frames = get_be32(s->pb);
469 if(frames < 0)
470 return;
472 spf = c.lsf ? 576 : 1152; /* Samples per frame, layer 3 */
473 st->duration = av_rescale_q(frames, (AVRational){spf, c.sample_rate},
474 st->time_base);
477 static int mp3_read_header(AVFormatContext *s,
478 AVFormatParameters *ap)
480 AVStream *st;
481 uint8_t buf[ID3v1_TAG_SIZE];
482 int len, ret, filesize;
483 offset_t off;
485 st = av_new_stream(s, 0);
486 if (!st)
487 return AVERROR(ENOMEM);
489 st->codec->codec_type = CODEC_TYPE_AUDIO;
490 st->codec->codec_id = CODEC_ID_MP3;
491 st->need_parsing = AVSTREAM_PARSE_FULL;
492 st->start_time = 0;
494 /* try to get the TAG */
495 if (!url_is_streamed(s->pb)) {
496 /* XXX: change that */
497 filesize = url_fsize(s->pb);
498 if (filesize > 128) {
499 url_fseek(s->pb, filesize - 128, SEEK_SET);
500 ret = get_buffer(s->pb, buf, ID3v1_TAG_SIZE);
501 if (ret == ID3v1_TAG_SIZE) {
502 id3v1_parse_tag(s, buf);
504 url_fseek(s->pb, 0, SEEK_SET);
508 /* if ID3v2 header found, skip it */
509 ret = get_buffer(s->pb, buf, ID3v2_HEADER_SIZE);
510 if (ret != ID3v2_HEADER_SIZE)
511 return -1;
512 if (id3v2_match(buf)) {
513 /* parse ID3v2 header */
514 len = ((buf[6] & 0x7f) << 21) |
515 ((buf[7] & 0x7f) << 14) |
516 ((buf[8] & 0x7f) << 7) |
517 (buf[9] & 0x7f);
518 id3v2_parse(s, len, buf[3], buf[5]);
519 } else {
520 url_fseek(s->pb, 0, SEEK_SET);
523 off = url_ftell(s->pb);
524 mp3_parse_vbr_tags(s, st, off);
525 url_fseek(s->pb, off, SEEK_SET);
527 /* the parameters will be extracted from the compressed bitstream */
528 return 0;
531 #define MP3_PACKET_SIZE 1024
533 static int mp3_read_packet(AVFormatContext *s, AVPacket *pkt)
535 int ret, size;
536 // AVStream *st = s->streams[0];
538 size= MP3_PACKET_SIZE;
540 ret= av_get_packet(s->pb, pkt, size);
542 pkt->stream_index = 0;
543 if (ret <= 0) {
544 return AVERROR(EIO);
546 /* note: we need to modify the packet size here to handle the last
547 packet */
548 pkt->size = ret;
549 return ret;
552 static int mp3_read_close(AVFormatContext *s)
554 return 0;
557 #ifdef CONFIG_MUXERS
558 /* simple formats */
560 static void id3v2_put_size(AVFormatContext *s, int size)
562 put_byte(s->pb, size >> 21 & 0x7f);
563 put_byte(s->pb, size >> 14 & 0x7f);
564 put_byte(s->pb, size >> 7 & 0x7f);
565 put_byte(s->pb, size & 0x7f);
568 static void id3v2_put_ttag(AVFormatContext *s, const char *string, uint32_t tag)
570 int len = strlen(string);
571 put_be32(s->pb, tag);
572 id3v2_put_size(s, len + 1);
573 put_be16(s->pb, 0);
574 put_byte(s->pb, 3); /* UTF-8 */
575 put_buffer(s->pb, string, len);
580 * Write an ID3v2.4 header at beginning of stream
583 static int mp3_write_header(struct AVFormatContext *s)
585 int totlen = 0;
586 char tracktxt[10];
587 char yeartxt[10];
589 if(s->track)
590 snprintf(tracktxt, sizeof(tracktxt), "%d", s->track);
591 if(s->year)
592 snprintf( yeartxt, sizeof(yeartxt) , "%d", s->year );
594 if(s->title[0]) totlen += 11 + strlen(s->title);
595 if(s->author[0]) totlen += 11 + strlen(s->author);
596 if(s->album[0]) totlen += 11 + strlen(s->album);
597 if(s->genre[0]) totlen += 11 + strlen(s->genre);
598 if(s->copyright[0]) totlen += 11 + strlen(s->copyright);
599 if(s->track) totlen += 11 + strlen(tracktxt);
600 if(s->year) totlen += 11 + strlen(yeartxt);
601 if(!(s->streams[0]->codec->flags & CODEC_FLAG_BITEXACT))
602 totlen += strlen(LIBAVFORMAT_IDENT) + 11;
604 if(totlen == 0)
605 return 0;
607 put_be32(s->pb, MKBETAG('I', 'D', '3', 0x04)); /* ID3v2.4 */
608 put_byte(s->pb, 0);
609 put_byte(s->pb, 0); /* flags */
611 id3v2_put_size(s, totlen);
613 if(s->title[0]) id3v2_put_ttag(s, s->title, MKBETAG('T', 'I', 'T', '2'));
614 if(s->author[0]) id3v2_put_ttag(s, s->author, MKBETAG('T', 'P', 'E', '1'));
615 if(s->album[0]) id3v2_put_ttag(s, s->album, MKBETAG('T', 'A', 'L', 'B'));
616 if(s->genre[0]) id3v2_put_ttag(s, s->genre, MKBETAG('T', 'C', 'O', 'N'));
617 if(s->copyright[0]) id3v2_put_ttag(s, s->copyright, MKBETAG('T', 'C', 'O', 'P'));
618 if(s->track) id3v2_put_ttag(s, tracktxt, MKBETAG('T', 'R', 'C', 'K'));
619 if(s->year) id3v2_put_ttag(s, yeartxt, MKBETAG('T', 'Y', 'E', 'R'));
620 if(!(s->streams[0]->codec->flags & CODEC_FLAG_BITEXACT))
621 id3v2_put_ttag(s, LIBAVFORMAT_IDENT, MKBETAG('T', 'E', 'N', 'C'));
622 return 0;
625 static int mp3_write_packet(struct AVFormatContext *s, AVPacket *pkt)
627 put_buffer(s->pb, pkt->data, pkt->size);
628 put_flush_packet(s->pb);
629 return 0;
632 static int mp3_write_trailer(struct AVFormatContext *s)
634 uint8_t buf[ID3v1_TAG_SIZE];
636 /* write the id3v1 tag */
637 if (s->title[0] != '\0') {
638 id3v1_create_tag(s, buf);
639 put_buffer(s->pb, buf, ID3v1_TAG_SIZE);
640 put_flush_packet(s->pb);
642 return 0;
644 #endif //CONFIG_MUXERS
646 #ifdef CONFIG_MP3_DEMUXER
647 AVInputFormat mp3_demuxer = {
648 "mp3",
649 "MPEG audio",
651 mp3_read_probe,
652 mp3_read_header,
653 mp3_read_packet,
654 mp3_read_close,
655 .flags= AVFMT_GENERIC_INDEX,
656 .extensions = "mp2,mp3,m2a", /* XXX: use probe */
658 #endif
659 #ifdef CONFIG_MP2_MUXER
660 AVOutputFormat mp2_muxer = {
661 "mp2",
662 "MPEG audio layer 2",
663 "audio/x-mpeg",
664 #ifdef CONFIG_LIBMP3LAME
665 "mp2,m2a",
666 #else
667 "mp2,mp3,m2a",
668 #endif
670 CODEC_ID_MP2,
671 CODEC_ID_NONE,
672 NULL,
673 mp3_write_packet,
674 mp3_write_trailer,
676 #endif
677 #ifdef CONFIG_MP3_MUXER
678 AVOutputFormat mp3_muxer = {
679 "mp3",
680 "MPEG audio layer 3",
681 "audio/x-mpeg",
682 "mp3",
684 CODEC_ID_MP3,
685 CODEC_ID_NONE,
686 mp3_write_header,
687 mp3_write_packet,
688 mp3_write_trailer,
690 #endif