summaryrefslogtreecommitdiff
path: root/libavformat/mp3dec.c
diff options
context:
space:
mode:
Diffstat (limited to 'libavformat/mp3dec.c')
-rw-r--r--libavformat/mp3dec.c218
1 files changed, 148 insertions, 70 deletions
diff --git a/libavformat/mp3dec.c b/libavformat/mp3dec.c
index 14d8254f4c..e04206c991 100644
--- a/libavformat/mp3dec.c
+++ b/libavformat/mp3dec.c
@@ -2,23 +2,24 @@
* MP3 demuxer
* Copyright (c) 2003 Fabrice Bellard
*
- * This file is part of Libav.
+ * This file is part of FFmpeg.
*
- * Libav is free software; you can redistribute it and/or
+ * FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
- * Libav is distributed in the hope that it will be useful,
+ * FFmpeg is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
- * License along with Libav; if not, write to the Free Software
+ * License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
+#include "libavutil/opt.h"
#include "libavutil/avstring.h"
#include "libavutil/intreadwrite.h"
#include "libavutil/crc.h"
@@ -39,10 +40,15 @@
#define XING_TOC_COUNT 100
-typedef struct MP3DecContext {
+typedef struct {
+ AVClass *class;
+ int64_t filesize;
int xing_toc;
+ int start_pad;
+ int end_pad;
+ int usetoc;
unsigned frames; /* Total number of frames in file */
- unsigned size; /* Total number of bytes in the stream */
+ unsigned header_filesize; /* Total number of bytes in the stream */
int is_cbr;
} MP3DecContext;
@@ -53,7 +59,7 @@ static int mp3_read_probe(AVProbeData *p)
int max_frames, first_frames = 0;
int fsize, frames, sample_rate;
uint32_t header;
- uint8_t *buf, *buf0, *buf2, *end;
+ const uint8_t *buf, *buf0, *buf2, *end;
AVCodecContext avctx;
buf0 = p->buf;
@@ -66,6 +72,8 @@ static int mp3_read_probe(AVProbeData *p)
for(; buf < end; buf= buf2+1) {
buf2 = buf;
+ if(ff_mpa_check_header(AV_RB32(buf2)))
+ continue;
for(frames = 0; buf2 < end; frames++) {
header = AV_RB32(buf2);
@@ -80,31 +88,13 @@ static int mp3_read_probe(AVProbeData *p)
}
// keep this in sync with ac3 probe, both need to avoid
// issues with MPEG-files!
- if (first_frames >= 4) return AVPROBE_SCORE_EXTENSION + 1;
-
- if (max_frames) {
- int pes = 0, i;
- unsigned int code = -1;
-
-#define VIDEO_ID 0x000001e0
-#define AUDIO_ID 0x000001c0
- /* do a search for mpegps headers to be able to properly bias
- * towards mpegps if we detect this stream as both. */
- for (i = 0; i<p->buf_size; i++) {
- code = (code << 8) + p->buf[i];
- if ((code & 0xffffff00) == 0x100) {
- if ((code & 0x1f0) == VIDEO_ID) pes++;
- else if((code & 0x1e0) == AUDIO_ID) pes++;
- }
- }
-
- if (pes)
- max_frames = (max_frames + pes - 1) / pes;
- }
- if (max_frames > 500) return AVPROBE_SCORE_EXTENSION;
- else if (max_frames >= 4) return AVPROBE_SCORE_EXTENSION / 2;
- else if (max_frames >= 1) return 1;
- else return 0;
+ if (first_frames>=4) return AVPROBE_SCORE_EXTENSION + 1;
+ else if(max_frames>200)return AVPROBE_SCORE_EXTENSION;
+ else if(max_frames>=4) return AVPROBE_SCORE_EXTENSION / 2;
+ else if(ff_id3v2_match(buf0, ID3v2_DEFAULT_MAGIC) && 2*ff_id3v2_tag_len(buf0) >= p->buf_size)
+ return p->buf_size < PROBE_BUF_MAX ? AVPROBE_SCORE_EXTENSION / 4 : AVPROBE_SCORE_EXTENSION - 2;
+ else if(max_frames>=1) return 1;
+ else return 0;
//mpegps_mp3_unrecognized_format.mpg has max_frames=3
}
@@ -112,22 +102,24 @@ static void read_xing_toc(AVFormatContext *s, int64_t filesize, int64_t duration
{
int i;
MP3DecContext *mp3 = s->priv_data;
+ int fill_index = mp3->usetoc && duration > 0;
if (!filesize &&
!(filesize = avio_size(s->pb))) {
av_log(s, AV_LOG_WARNING, "Cannot determine file size, skipping TOC table.\n");
- return;
+ fill_index = 0;
}
for (i = 0; i < XING_TOC_COUNT; i++) {
uint8_t b = avio_r8(s->pb);
-
- av_add_index_entry(s->streams[0],
+ if (fill_index)
+ av_add_index_entry(s->streams[0],
av_rescale(b, filesize, 256),
av_rescale(i, duration, XING_TOC_COUNT),
0, 0, AVINDEX_KEYFRAME);
}
- mp3->xing_toc = 1;
+ if (fill_index)
+ mp3->xing_toc = 1;
}
static void mp3_parse_info_tag(AVFormatContext *s, AVStream *st,
@@ -145,7 +137,7 @@ static void mp3_parse_info_tag(AVFormatContext *s, AVStream *st,
int32_t r_gain = INT32_MIN, a_gain = INT32_MIN;
MP3DecContext *mp3 = s->priv_data;
- const int64_t xing_offtbl[2][2] = {{32, 17}, {17,9}};
+ static const int64_t xing_offtbl[2][2] = {{32, 17}, {17,9}};
/* Check for Xing / Info tag */
avio_skip(s->pb, xing_offtbl[c->lsf == 1][c->nb_channels == 1]);
@@ -158,14 +150,14 @@ static void mp3_parse_info_tag(AVFormatContext *s, AVStream *st,
if (v & XING_FLAG_FRAMES)
mp3->frames = avio_rb32(s->pb);
if (v & XING_FLAG_SIZE)
- mp3->size = avio_rb32(s->pb);
- if (v & XING_FLAG_TOC && mp3->frames)
- read_xing_toc(s, mp3->size, av_rescale_q(mp3->frames,
+ mp3->header_filesize = avio_rb32(s->pb);
+ if (v & XING_FLAG_TOC)
+ read_xing_toc(s, mp3->header_filesize, av_rescale_q(mp3->frames,
(AVRational){spf, c->sample_rate},
st->time_base));
-
/* VBR quality */
- avio_rb32(s->pb);
+ if(v & 8)
+ avio_skip(s->pb, 4);
/* Encoder short version string */
memset(version, 0, sizeof(version));
@@ -208,7 +200,19 @@ static void mp3_parse_info_tag(AVFormatContext *s, AVStream *st,
avio_r8(s->pb);
/* Encoder delays */
- avio_rb24(s->pb);
+ v= avio_rb24(s->pb);
+ if(AV_RB32(version) == MKBETAG('L', 'A', 'M', 'E')
+ || AV_RB32(version) == MKBETAG('L', 'a', 'v', 'f')) {
+
+ mp3->start_pad = v>>12;
+ mp3-> end_pad = v&4095;
+ st->skip_samples = mp3->start_pad + 528 + 1;
+ if (!st->start_time)
+ st->start_time = av_rescale_q(st->skip_samples,
+ (AVRational){1, c->sample_rate},
+ st->time_base);
+ av_log(s, AV_LOG_DEBUG, "pad %d %d\n", mp3->start_pad, mp3-> end_pad);
+ }
/* Misc */
avio_r8(s->pb);
@@ -248,7 +252,7 @@ static void mp3_parse_vbri_tag(AVFormatContext *s, AVStream *st, int64_t base)
if (avio_rb16(s->pb) == 1) {
/* skip delay and quality */
avio_skip(s->pb, 4);
- mp3->size = avio_rb32(s->pb);
+ mp3->header_filesize = avio_rb32(s->pb);
mp3->frames = avio_rb32(s->pb);
}
}
@@ -278,12 +282,12 @@ static int mp3_parse_vbr_tags(AVFormatContext *s, AVStream *st, int64_t base)
spf = c.lsf ? 576 : 1152; /* Samples per frame, layer 3 */
mp3->frames = 0;
- mp3->size = 0;
+ mp3->header_filesize = 0;
mp3_parse_info_tag(s, st, &c, spf);
mp3_parse_vbri_tag(s, st, base);
- if (!mp3->frames && !mp3->size)
+ if (!mp3->frames && !mp3->header_filesize)
return -1;
/* Skip the vbr tag frame */
@@ -292,14 +296,15 @@ static int mp3_parse_vbr_tags(AVFormatContext *s, AVStream *st, int64_t base)
if (mp3->frames)
st->duration = av_rescale_q(mp3->frames, (AVRational){spf, c.sample_rate},
st->time_base);
- if (mp3->size && mp3->frames && !mp3->is_cbr)
- st->codec->bit_rate = av_rescale(mp3->size, 8 * c.sample_rate, mp3->frames * (int64_t)spf);
+ if (mp3->header_filesize && mp3->frames && !mp3->is_cbr)
+ st->codec->bit_rate = av_rescale(mp3->header_filesize, 8 * c.sample_rate, mp3->frames * (int64_t)spf);
return 0;
}
static int mp3_read_header(AVFormatContext *s)
{
+ MP3DecContext *mp3 = s->priv_data;
AVStream *st;
int64_t off;
int ret;
@@ -310,17 +315,21 @@ static int mp3_read_header(AVFormatContext *s)
st->codec->codec_type = AVMEDIA_TYPE_AUDIO;
st->codec->codec_id = AV_CODEC_ID_MP3;
- st->need_parsing = AVSTREAM_PARSE_FULL;
+ st->need_parsing = AVSTREAM_PARSE_FULL_RAW;
st->start_time = 0;
// lcm of all mp3 sample rates
avpriv_set_pts_info(st, 64, 1, 14112000);
+ s->pb->maxsize = -1;
off = avio_tell(s->pb);
if (!av_dict_get(s->metadata, "", NULL, AV_DICT_IGNORE_SUFFIX))
ff_id3v1_read(s);
+ if(s->pb->seekable)
+ mp3->filesize = avio_size(s->pb);
+
if (mp3_parse_vbr_tags(s, st, off) < 0)
avio_seek(s->pb, off, SEEK_SET);
@@ -336,15 +345,26 @@ static int mp3_read_header(AVFormatContext *s)
static int mp3_read_packet(AVFormatContext *s, AVPacket *pkt)
{
- int ret;
-
- ret = av_get_packet(s->pb, pkt, MP3_PACKET_SIZE);
- if (ret < 0)
- return ret;
+ MP3DecContext *mp3 = s->priv_data;
+ int ret, size;
+ int64_t pos;
+
+ size= MP3_PACKET_SIZE;
+ pos = avio_tell(s->pb);
+ if(mp3->filesize > ID3v1_TAG_SIZE && pos < mp3->filesize)
+ size= FFMIN(size, mp3->filesize - pos);
+
+ ret= av_get_packet(s->pb, pkt, size);
+ if (ret <= 0) {
+ if(ret<0)
+ return ret;
+ return AVERROR_EOF;
+ }
+ pkt->flags &= ~AV_PKT_FLAG_CORRUPT;
pkt->stream_index = 0;
- if (ret > ID3v1_TAG_SIZE &&
+ if (ret >= ID3v1_TAG_SIZE &&
memcmp(&pkt->data[ret - ID3v1_TAG_SIZE], "TAG", 3) == 0)
ret -= ID3v1_TAG_SIZE;
@@ -354,38 +374,95 @@ static int mp3_read_packet(AVFormatContext *s, AVPacket *pkt)
return ret;
}
+static int check(AVFormatContext *s, int64_t pos)
+{
+ int64_t ret = avio_seek(s->pb, pos, SEEK_SET);
+ unsigned header;
+ MPADecodeHeader sd;
+ if (ret < 0)
+ return ret;
+ header = avio_rb32(s->pb);
+ if (ff_mpa_check_header(header) < 0)
+ return -1;
+ if (avpriv_mpegaudio_decode_header(&sd, header) == 1)
+ return -1;
+ return sd.frame_size;
+}
+
static int mp3_seek(AVFormatContext *s, int stream_index, int64_t timestamp,
int flags)
{
MP3DecContext *mp3 = s->priv_data;
- AVIndexEntry *ie;
+ AVIndexEntry *ie, ie1;
AVStream *st = s->streams[0];
int64_t ret = av_index_search_timestamp(st, timestamp, flags);
- uint32_t header = 0;
+ int i, j;
+ int dir = (flags&AVSEEK_FLAG_BACKWARD) ? -1 : 1;
+
+ if (mp3->is_cbr && st->duration > 0 && mp3->header_filesize > s->data_offset) {
+ int64_t filesize = avio_size(s->pb);
+ int64_t duration;
+ if (filesize <= s->data_offset)
+ filesize = mp3->header_filesize;
+ filesize -= s->data_offset;
+ duration = av_rescale(st->duration, filesize, mp3->header_filesize - s->data_offset);
+ ie = &ie1;
+ timestamp = av_clip64(timestamp, 0, duration);
+ ie->timestamp = timestamp;
+ ie->pos = av_rescale(timestamp, filesize, duration) + s->data_offset;
+ } else if (mp3->xing_toc) {
+ if (ret < 0)
+ return ret;
+
+ ie = &st->index_entries[ret];
+ } else {
+ st->skip_samples = timestamp <= 0 ? mp3->start_pad + 528 + 1 : 0;
- if (!mp3->xing_toc)
- return AVERROR(ENOSYS);
-
- if (ret < 0)
- return ret;
+ return -1;
+ }
- ie = &st->index_entries[ret];
+ if (dir < 0)
+ avio_seek(s->pb, FFMAX(ie->pos - 4096, 0), SEEK_SET);
ret = avio_seek(s->pb, ie->pos, SEEK_SET);
if (ret < 0)
return ret;
- while (!s->pb->eof_reached) {
- header = (header << 8) + avio_r8(s->pb);
- if (ff_mpa_check_header(header) >= 0) {
- ff_update_cur_dts(s, st, ie->timestamp);
- ret = avio_seek(s->pb, -4, SEEK_CUR);
- return (ret >= 0) ? 0 : ret;
+#define MIN_VALID 3
+ for(i=0; i<4096; i++) {
+ int64_t pos = ie->pos + i*dir;
+ for(j=0; j<MIN_VALID; j++) {
+ ret = check(s, pos);
+ if(ret < 0)
+ break;
+ pos += ret;
}
+ if(j==MIN_VALID)
+ break;
}
+ if(j!=MIN_VALID)
+ i=0;
- return AVERROR_EOF;
+ ret = avio_seek(s->pb, ie->pos + i*dir, SEEK_SET);
+ if (ret < 0)
+ return ret;
+ ff_update_cur_dts(s, st, ie->timestamp);
+ st->skip_samples = ie->timestamp <= 0 ? mp3->start_pad + 528 + 1 : 0;
+ return 0;
}
+static const AVOption options[] = {
+ { "usetoc", "use table of contents", offsetof(MP3DecContext, usetoc), AV_OPT_TYPE_INT, {.i64 = -1}, -1, 1, AV_OPT_FLAG_DECODING_PARAM},
+ { NULL },
+};
+
+static const AVClass demuxer_class = {
+ .class_name = "mp3",
+ .item_name = av_default_item_name,
+ .option = options,
+ .version = LIBAVUTIL_VERSION_INT,
+ .category = AV_CLASS_CATEGORY_DEMUXER,
+};
+
AVInputFormat ff_mp3_demuxer = {
.name = "mp3",
.long_name = NULL_IF_CONFIG_SMALL("MP2/3 (MPEG audio layer 2/3)"),
@@ -396,4 +473,5 @@ AVInputFormat ff_mp3_demuxer = {
.priv_data_size = sizeof(MP3DecContext),
.flags = AVFMT_GENERIC_INDEX,
.extensions = "mp2,mp3,m2a,mpa", /* XXX: use probe */
+ .priv_class = &demuxer_class,
};