mirror of
https://git.ffmpeg.org/ffmpeg.git
synced 2024-12-24 16:22:37 +00:00
976a8b2179
* qatar/master: (40 commits) H.264: template left MB handling H.264: faster fill_decode_caches H.264: faster write_back_* H.264: faster fill_filter_caches H.264: make filter_mb_fast support the case of unavailable top mb Do not include log.h in avutil.h Do not include pixfmt.h in avutil.h Do not include rational.h in avutil.h Do not include mathematics.h in avutil.h Do not include intfloat_readwrite.h in avutil.h Remove return statements following infinite loops without break RTSP: Doxygen comment cleanup doxygen: Escape '\' in Doxygen documentation. md5: cosmetics md5: use AV_WL32 to write result md5: add fate test md5: include correct headers md5: fix test program doxygen: Drop array size declarations from Doxygen parameter names. doxygen: Fix parameter names to match the function prototypes. ... Conflicts: libavcodec/x86/dsputil_mmx.c libavformat/flvenc.c libavformat/oggenc.c libavformat/wtv.c Merged-by: Michael Niedermayer <michaelni@gmx.at>
403 lines
13 KiB
C
403 lines
13 KiB
C
/*
|
|
* Core Audio Format demuxer
|
|
* Copyright (c) 2007 Justin Ruggles
|
|
* Copyright (c) 2009 Peter Ross
|
|
*
|
|
* This file is part of FFmpeg.
|
|
*
|
|
* 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.
|
|
*
|
|
* 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 FFmpeg; if not, write to the Free Software
|
|
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
|
*/
|
|
|
|
/**
|
|
* @file
|
|
* Core Audio Format demuxer
|
|
*/
|
|
|
|
#include "avformat.h"
|
|
#include "riff.h"
|
|
#include "isom.h"
|
|
#include "libavutil/intreadwrite.h"
|
|
#include "libavutil/intfloat_readwrite.h"
|
|
#include "libavutil/dict.h"
|
|
#include "caf.h"
|
|
|
|
typedef struct {
|
|
int bytes_per_packet; ///< bytes in a packet, or 0 if variable
|
|
int frames_per_packet; ///< frames in a packet, or 0 if variable
|
|
int64_t num_bytes; ///< total number of bytes in stream
|
|
|
|
int64_t packet_cnt; ///< packet counter
|
|
int64_t frame_cnt; ///< frame counter
|
|
|
|
int64_t data_start; ///< data start position, in bytes
|
|
int64_t data_size; ///< raw data size, in bytes
|
|
} CaffContext;
|
|
|
|
static int probe(AVProbeData *p)
|
|
{
|
|
if (AV_RB32(p->buf) == MKBETAG('c','a','f','f') && AV_RB16(&p->buf[4]) == 1)
|
|
return AVPROBE_SCORE_MAX;
|
|
return 0;
|
|
}
|
|
|
|
/** Read audio description chunk */
|
|
static int read_desc_chunk(AVFormatContext *s)
|
|
{
|
|
AVIOContext *pb = s->pb;
|
|
CaffContext *caf = s->priv_data;
|
|
AVStream *st;
|
|
int flags;
|
|
|
|
/* new audio stream */
|
|
st = av_new_stream(s, 0);
|
|
if (!st)
|
|
return AVERROR(ENOMEM);
|
|
|
|
/* parse format description */
|
|
st->codec->codec_type = AVMEDIA_TYPE_AUDIO;
|
|
st->codec->sample_rate = av_int2dbl(avio_rb64(pb));
|
|
st->codec->codec_tag = avio_rb32(pb);
|
|
flags = avio_rb32(pb);
|
|
caf->bytes_per_packet = avio_rb32(pb);
|
|
st->codec->block_align = caf->bytes_per_packet;
|
|
caf->frames_per_packet = avio_rb32(pb);
|
|
st->codec->channels = avio_rb32(pb);
|
|
st->codec->bits_per_coded_sample = avio_rb32(pb);
|
|
|
|
/* calculate bit rate for constant size packets */
|
|
if (caf->frames_per_packet > 0 && caf->bytes_per_packet > 0) {
|
|
st->codec->bit_rate = (uint64_t)st->codec->sample_rate * (uint64_t)caf->bytes_per_packet * 8
|
|
/ (uint64_t)caf->frames_per_packet;
|
|
} else {
|
|
st->codec->bit_rate = 0;
|
|
}
|
|
|
|
/* determine codec */
|
|
if (st->codec->codec_tag == MKBETAG('l','p','c','m'))
|
|
st->codec->codec_id = ff_mov_get_lpcm_codec_id(st->codec->bits_per_coded_sample, (flags ^ 0x2) | 0x4);
|
|
else
|
|
st->codec->codec_id = ff_codec_get_id(ff_codec_caf_tags, st->codec->codec_tag);
|
|
return 0;
|
|
}
|
|
|
|
/** Read magic cookie chunk */
|
|
static int read_kuki_chunk(AVFormatContext *s, int64_t size)
|
|
{
|
|
AVIOContext *pb = s->pb;
|
|
AVStream *st = s->streams[0];
|
|
|
|
if (size < 0 || size > INT_MAX - FF_INPUT_BUFFER_PADDING_SIZE)
|
|
return -1;
|
|
|
|
if (st->codec->codec_id == CODEC_ID_AAC) {
|
|
/* The magic cookie format for AAC is an mp4 esds atom.
|
|
The lavc AAC decoder requires the data from the codec specific
|
|
description as extradata input. */
|
|
int strt, skip;
|
|
MOVAtom atom;
|
|
|
|
strt = avio_tell(pb);
|
|
ff_mov_read_esds(s, pb, atom);
|
|
skip = size - (avio_tell(pb) - strt);
|
|
if (skip < 0 || !st->codec->extradata ||
|
|
st->codec->codec_id != CODEC_ID_AAC) {
|
|
av_log(s, AV_LOG_ERROR, "invalid AAC magic cookie\n");
|
|
return AVERROR_INVALIDDATA;
|
|
}
|
|
avio_skip(pb, skip);
|
|
} else if (st->codec->codec_id == CODEC_ID_ALAC) {
|
|
#define ALAC_PREAMBLE 12
|
|
#define ALAC_HEADER 36
|
|
if (size < ALAC_PREAMBLE + ALAC_HEADER) {
|
|
av_log(s, AV_LOG_ERROR, "invalid ALAC magic cookie\n");
|
|
avio_skip(pb, size);
|
|
return AVERROR_INVALIDDATA;
|
|
}
|
|
avio_skip(pb, ALAC_PREAMBLE);
|
|
st->codec->extradata = av_mallocz(ALAC_HEADER + FF_INPUT_BUFFER_PADDING_SIZE);
|
|
if (!st->codec->extradata)
|
|
return AVERROR(ENOMEM);
|
|
avio_read(pb, st->codec->extradata, ALAC_HEADER);
|
|
st->codec->extradata_size = ALAC_HEADER;
|
|
avio_skip(pb, size - ALAC_PREAMBLE - ALAC_HEADER);
|
|
} else {
|
|
st->codec->extradata = av_mallocz(size + FF_INPUT_BUFFER_PADDING_SIZE);
|
|
if (!st->codec->extradata)
|
|
return AVERROR(ENOMEM);
|
|
avio_read(pb, st->codec->extradata, size);
|
|
st->codec->extradata_size = size;
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
/** Read packet table chunk */
|
|
static int read_pakt_chunk(AVFormatContext *s, int64_t size)
|
|
{
|
|
AVIOContext *pb = s->pb;
|
|
AVStream *st = s->streams[0];
|
|
CaffContext *caf = s->priv_data;
|
|
int64_t pos = 0, ccount;
|
|
int num_packets, i;
|
|
|
|
ccount = avio_tell(pb);
|
|
|
|
num_packets = avio_rb64(pb);
|
|
if (num_packets < 0 || INT32_MAX / sizeof(AVIndexEntry) < num_packets)
|
|
return AVERROR_INVALIDDATA;
|
|
|
|
st->nb_frames = avio_rb64(pb); /* valid frames */
|
|
st->nb_frames += avio_rb32(pb); /* priming frames */
|
|
st->nb_frames += avio_rb32(pb); /* remainder frames */
|
|
|
|
st->duration = 0;
|
|
for (i = 0; i < num_packets; i++) {
|
|
av_add_index_entry(s->streams[0], pos, st->duration, 0, 0, AVINDEX_KEYFRAME);
|
|
pos += caf->bytes_per_packet ? caf->bytes_per_packet : ff_mp4_read_descr_len(pb);
|
|
st->duration += caf->frames_per_packet ? caf->frames_per_packet : ff_mp4_read_descr_len(pb);
|
|
}
|
|
|
|
if (avio_tell(pb) - ccount != size) {
|
|
av_log(s, AV_LOG_ERROR, "error reading packet table\n");
|
|
return -1;
|
|
}
|
|
|
|
caf->num_bytes = pos;
|
|
return 0;
|
|
}
|
|
|
|
/** Read information chunk */
|
|
static void read_info_chunk(AVFormatContext *s, int64_t size)
|
|
{
|
|
AVIOContext *pb = s->pb;
|
|
unsigned int i;
|
|
unsigned int nb_entries = avio_rb32(pb);
|
|
for (i = 0; i < nb_entries; i++) {
|
|
char key[32];
|
|
char value[1024];
|
|
get_strz(pb, key, sizeof(key));
|
|
get_strz(pb, value, sizeof(value));
|
|
av_dict_set(&s->metadata, key, value, 0);
|
|
}
|
|
}
|
|
|
|
static int read_header(AVFormatContext *s,
|
|
AVFormatParameters *ap)
|
|
{
|
|
AVIOContext *pb = s->pb;
|
|
CaffContext *caf = s->priv_data;
|
|
AVStream *st;
|
|
uint32_t tag = 0;
|
|
int found_data, ret;
|
|
int64_t size;
|
|
|
|
avio_skip(pb, 8); /* magic, version, file flags */
|
|
|
|
/* audio description chunk */
|
|
if (avio_rb32(pb) != MKBETAG('d','e','s','c')) {
|
|
av_log(s, AV_LOG_ERROR, "desc chunk not present\n");
|
|
return AVERROR_INVALIDDATA;
|
|
}
|
|
size = avio_rb64(pb);
|
|
if (size != 32)
|
|
return AVERROR_INVALIDDATA;
|
|
|
|
ret = read_desc_chunk(s);
|
|
if (ret)
|
|
return ret;
|
|
st = s->streams[0];
|
|
|
|
/* parse each chunk */
|
|
found_data = 0;
|
|
while (!url_feof(pb)) {
|
|
|
|
/* stop at data chunk if seeking is not supported or
|
|
data chunk size is unknown */
|
|
if (found_data && (caf->data_size < 0 || !pb->seekable))
|
|
break;
|
|
|
|
tag = avio_rb32(pb);
|
|
size = avio_rb64(pb);
|
|
if (url_feof(pb))
|
|
break;
|
|
|
|
switch (tag) {
|
|
case MKBETAG('d','a','t','a'):
|
|
avio_skip(pb, 4); /* edit count */
|
|
caf->data_start = avio_tell(pb);
|
|
caf->data_size = size < 0 ? -1 : size - 4;
|
|
if (caf->data_size > 0 && pb->seekable)
|
|
avio_skip(pb, caf->data_size);
|
|
found_data = 1;
|
|
break;
|
|
|
|
/* magic cookie chunk */
|
|
case MKBETAG('k','u','k','i'):
|
|
if (read_kuki_chunk(s, size))
|
|
return AVERROR_INVALIDDATA;
|
|
break;
|
|
|
|
/* packet table chunk */
|
|
case MKBETAG('p','a','k','t'):
|
|
if (read_pakt_chunk(s, size))
|
|
return AVERROR_INVALIDDATA;
|
|
break;
|
|
|
|
case MKBETAG('i','n','f','o'):
|
|
read_info_chunk(s, size);
|
|
break;
|
|
|
|
case MKBETAG('c','h','a','n'):
|
|
if (size < 12)
|
|
return AVERROR_INVALIDDATA;
|
|
ff_mov_read_chan(s, size, st->codec);
|
|
break;
|
|
|
|
default:
|
|
#define _(x) ((x) >= ' ' ? (x) : ' ')
|
|
av_log(s, AV_LOG_WARNING, "skipping CAF chunk: %08X (%c%c%c%c), size %"PRId64"\n",
|
|
tag, _(tag>>24), _((tag>>16)&0xFF), _((tag>>8)&0xFF), _(tag&0xFF), size);
|
|
#undef _
|
|
case MKBETAG('f','r','e','e'):
|
|
if (size < 0)
|
|
return AVERROR_INVALIDDATA;
|
|
avio_skip(pb, size);
|
|
break;
|
|
}
|
|
}
|
|
|
|
if (!found_data)
|
|
return AVERROR_INVALIDDATA;
|
|
|
|
if (caf->bytes_per_packet > 0 && caf->frames_per_packet > 0) {
|
|
if (caf->data_size > 0)
|
|
st->nb_frames = (caf->data_size / caf->bytes_per_packet) * caf->frames_per_packet;
|
|
} else if (st->nb_index_entries) {
|
|
st->codec->bit_rate = st->codec->sample_rate * caf->data_size * 8 /
|
|
st->duration;
|
|
} else {
|
|
av_log(s, AV_LOG_ERROR, "Missing packet table. It is required when "
|
|
"block size or frame size are variable.\n");
|
|
return AVERROR_INVALIDDATA;
|
|
}
|
|
s->file_size = avio_size(pb);
|
|
s->file_size = FFMAX(0, s->file_size);
|
|
|
|
av_set_pts_info(st, 64, 1, st->codec->sample_rate);
|
|
st->start_time = 0;
|
|
|
|
/* position the stream at the start of data */
|
|
if (caf->data_size >= 0)
|
|
avio_seek(pb, caf->data_start, SEEK_SET);
|
|
|
|
return 0;
|
|
}
|
|
|
|
#define CAF_MAX_PKT_SIZE 4096
|
|
|
|
static int read_packet(AVFormatContext *s, AVPacket *pkt)
|
|
{
|
|
AVIOContext *pb = s->pb;
|
|
AVStream *st = s->streams[0];
|
|
CaffContext *caf = s->priv_data;
|
|
int res, pkt_size = 0, pkt_frames = 0;
|
|
int64_t left = CAF_MAX_PKT_SIZE;
|
|
|
|
if (url_feof(pb))
|
|
return AVERROR(EIO);
|
|
|
|
/* don't read past end of data chunk */
|
|
if (caf->data_size > 0) {
|
|
left = (caf->data_start + caf->data_size) - avio_tell(pb);
|
|
if (left <= 0)
|
|
return AVERROR(EIO);
|
|
}
|
|
|
|
pkt_frames = caf->frames_per_packet;
|
|
pkt_size = caf->bytes_per_packet;
|
|
|
|
if (pkt_size > 0 && pkt_frames == 1) {
|
|
pkt_size = (CAF_MAX_PKT_SIZE / pkt_size) * pkt_size;
|
|
pkt_size = FFMIN(pkt_size, left);
|
|
pkt_frames = pkt_size / caf->bytes_per_packet;
|
|
} else if (st->nb_index_entries) {
|
|
if (caf->packet_cnt < st->nb_index_entries - 1) {
|
|
pkt_size = st->index_entries[caf->packet_cnt + 1].pos - st->index_entries[caf->packet_cnt].pos;
|
|
pkt_frames = st->index_entries[caf->packet_cnt + 1].timestamp - st->index_entries[caf->packet_cnt].timestamp;
|
|
} else if (caf->packet_cnt == st->nb_index_entries - 1) {
|
|
pkt_size = caf->num_bytes - st->index_entries[caf->packet_cnt].pos;
|
|
pkt_frames = st->duration - st->index_entries[caf->packet_cnt].timestamp;
|
|
} else {
|
|
return AVERROR(EIO);
|
|
}
|
|
}
|
|
|
|
if (pkt_size == 0 || pkt_frames == 0 || pkt_size > left)
|
|
return AVERROR(EIO);
|
|
|
|
res = av_get_packet(pb, pkt, pkt_size);
|
|
if (res < 0)
|
|
return res;
|
|
|
|
pkt->size = res;
|
|
pkt->stream_index = 0;
|
|
pkt->dts = pkt->pts = caf->frame_cnt;
|
|
|
|
caf->packet_cnt++;
|
|
caf->frame_cnt += pkt_frames;
|
|
|
|
return 0;
|
|
}
|
|
|
|
static int read_seek(AVFormatContext *s, int stream_index,
|
|
int64_t timestamp, int flags)
|
|
{
|
|
AVStream *st = s->streams[0];
|
|
CaffContext *caf = s->priv_data;
|
|
int64_t pos;
|
|
|
|
timestamp = FFMAX(timestamp, 0);
|
|
|
|
if (caf->frames_per_packet > 0 && caf->bytes_per_packet > 0) {
|
|
/* calculate new byte position based on target frame position */
|
|
pos = caf->bytes_per_packet * timestamp / caf->frames_per_packet;
|
|
if (caf->data_size > 0)
|
|
pos = FFMIN(pos, caf->data_size);
|
|
caf->packet_cnt = pos / caf->bytes_per_packet;
|
|
caf->frame_cnt = caf->frames_per_packet * caf->packet_cnt;
|
|
} else if (st->nb_index_entries) {
|
|
caf->packet_cnt = av_index_search_timestamp(st, timestamp, flags);
|
|
caf->frame_cnt = st->index_entries[caf->packet_cnt].timestamp;
|
|
pos = st->index_entries[caf->packet_cnt].pos;
|
|
} else {
|
|
return -1;
|
|
}
|
|
|
|
avio_seek(s->pb, pos + caf->data_start, SEEK_SET);
|
|
return 0;
|
|
}
|
|
|
|
AVInputFormat ff_caf_demuxer = {
|
|
"caf",
|
|
NULL_IF_CONFIG_SMALL("Apple Core Audio Format"),
|
|
sizeof(CaffContext),
|
|
probe,
|
|
read_header,
|
|
read_packet,
|
|
NULL,
|
|
read_seek,
|
|
.codec_tag = (const AVCodecTag*[]){ff_codec_caf_tags, 0},
|
|
};
|