You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

852 lines
29 KiB

/*
* Ogg muxer
* Copyright (c) 2007 Baptiste Coudurier <baptiste dot coudurier at free dot fr>
*
* 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
*/
#include "config_components.h"
#include <stdint.h>
#include "libavutil/crc.h"
#include "libavutil/mathematics.h"
#include "libavutil/opt.h"
#include "libavutil/random_seed.h"
#include "libavcodec/xiph.h"
#include "libavcodec/bytestream.h"
#include "libavcodec/flac.h"
#include "avformat.h"
#include "avio_internal.h"
#include "internal.h"
#include "version.h"
#include "vorbiscomment.h"
#define MAX_PAGE_SIZE 65025
typedef struct OGGPage {
int64_t start_granule;
int64_t granule;
int stream_index;
uint8_t flags;
uint8_t segments_count;
uint8_t segments[255];
uint8_t data[MAX_PAGE_SIZE];
uint16_t size;
} OGGPage;
typedef struct OGGStreamContext {
unsigned page_counter;
uint8_t *header[3];
int header_len[3];
/** for theora granule */
int kfgshift;
int64_t last_kf_pts;
int vrev;
/* for VP8 granule */
int isvp8;
int eos;
unsigned page_count; ///< number of page buffered
OGGPage page; ///< current page
unsigned serial_num; ///< serial number
int64_t last_granule; ///< last packet granule
} OGGStreamContext;
typedef struct OGGPageList {
OGGPage page;
struct OGGPageList *next;
} OGGPageList;
typedef struct OGGContext {
const AVClass *class;
OGGPageList *page_list;
int pref_size; ///< preferred page size (0 => fill all segments)
int64_t pref_duration; ///< preferred page duration (0 => fill all segments)
int serial_offset;
} OGGContext;
#define OFFSET(x) offsetof(OGGContext, x)
#define PARAM AV_OPT_FLAG_ENCODING_PARAM
static const AVOption options[] = {
{ "serial_offset", "serial number offset",
OFFSET(serial_offset), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, INT_MAX, PARAM },
{ "oggpagesize", "Set preferred Ogg page size.",
OFFSET(pref_size), AV_OPT_TYPE_INT, {.i64 = 0}, 0, MAX_PAGE_SIZE, PARAM},
{ "pagesize", "preferred page size in bytes (deprecated)",
OFFSET(pref_size), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, MAX_PAGE_SIZE, PARAM },
{ "page_duration", "preferred page duration, in microseconds",
OFFSET(pref_duration), AV_OPT_TYPE_INT64, { .i64 = 1000000 }, 0, INT64_MAX, PARAM },
{ NULL },
};
static const AVClass ogg_muxer_class = {
.class_name = "Ogg (audio/video/Speex/Opus) muxer",
.item_name = av_default_item_name,
.option = options,
.version = LIBAVUTIL_VERSION_INT,
};
static void ogg_write_page(AVFormatContext *s, OGGPage *page, int extra_flags)
{
OGGStreamContext *oggstream = s->streams[page->stream_index]->priv_data;
uint8_t buf[4 + 1 + 1 + 8 + 4 + 4 + 4 + 1 + 255], *ptr = buf, *crc_pos;
const AVCRC *crc_table = av_crc_get_table(AV_CRC_32_IEEE);
uint32_t crc;
bytestream_put_le32(&ptr, MKTAG('O', 'g', 'g', 'S'));
bytestream_put_byte(&ptr, 0);
bytestream_put_byte(&ptr, page->flags | extra_flags);
bytestream_put_le64(&ptr, page->granule);
bytestream_put_le32(&ptr, oggstream->serial_num);
bytestream_put_le32(&ptr, oggstream->page_counter++);
crc_pos = ptr;
bytestream_put_le32(&ptr, 0);
bytestream_put_byte(&ptr, page->segments_count);
bytestream_put_buffer(&ptr, page->segments, page->segments_count);
crc = av_crc(crc_table, 0, buf, ptr - buf);
crc = av_crc(crc_table, crc, page->data, page->size);
bytestream_put_be32(&crc_pos, crc);
avio_write(s->pb, buf, ptr - buf);
avio_write(s->pb, page->data, page->size);
avio_write_marker(s->pb, AV_NOPTS_VALUE, AVIO_DATA_MARKER_FLUSH_POINT);
oggstream->page_count--;
}
static int ogg_key_granule(OGGStreamContext *oggstream, int64_t granule)
{
return (oggstream->kfgshift && !(granule & ((1<<oggstream->kfgshift)-1))) ||
(oggstream->isvp8 && !((granule >> 3) & 0x07ffffff));
}
static int64_t ogg_granule_to_timestamp(OGGStreamContext *oggstream, int64_t granule)
{
if (oggstream->kfgshift)
return (granule>>oggstream->kfgshift) +
(granule & ((1<<oggstream->kfgshift)-1));
else if (oggstream->isvp8)
return granule >> 32;
else
return granule;
}
static int ogg_compare_granule(AVFormatContext *s, OGGPage *next, OGGPage *page)
{
AVStream *st2 = s->streams[next->stream_index];
AVStream *st = s->streams[page->stream_index];
int64_t next_granule, cur_granule;
if (next->granule == -1 || page->granule == -1)
return 0;
next_granule = av_rescale_q(ogg_granule_to_timestamp(st2->priv_data, next->granule),
st2->time_base, AV_TIME_BASE_Q);
cur_granule = av_rescale_q(ogg_granule_to_timestamp(st->priv_data, page->granule),
st ->time_base, AV_TIME_BASE_Q);
return next_granule > cur_granule;
}
static int ogg_reset_cur_page(OGGStreamContext *oggstream)
{
oggstream->page.granule = -1;
oggstream->page.flags = 0;
oggstream->page.segments_count = 0;
oggstream->page.size = 0;
return 0;
}
static int ogg_buffer_page(AVFormatContext *s, OGGStreamContext *oggstream)
{
OGGContext *ogg = s->priv_data;
OGGPageList **p = &ogg->page_list;
OGGPageList *l = av_mallocz(sizeof(*l));
if (!l)
return AVERROR(ENOMEM);
l->page = oggstream->page;
oggstream->page.start_granule = ogg_granule_to_timestamp(oggstream, oggstream->page.granule);
oggstream->page_count++;
ogg_reset_cur_page(oggstream);
while (*p) {
if (ogg_compare_granule(s, &(*p)->page, &l->page))
break;
p = &(*p)->next;
}
l->next = *p;
*p = l;
return 0;
}
static int ogg_buffer_data(AVFormatContext *s, AVStream *st,
uint8_t *data, unsigned size, int64_t granule,
int header)
{
OGGStreamContext *oggstream = st->priv_data;
OGGContext *ogg = s->priv_data;
int total_segments = size / 255 + 1;
uint8_t *p = data;
int i, segments, len, flush = 0;
// Handles VFR by flushing page because this frame needs to have a timestamp
// For theora and VP8, keyframes also need to have a timestamp to correctly mark
// them as such, otherwise seeking will not work correctly at the very
// least with old libogg versions.
// Do not try to flush header packets though, that will create broken files.
if ((st->codecpar->codec_id == AV_CODEC_ID_THEORA || st->codecpar->codec_id == AV_CODEC_ID_VP8) && !header &&
(ogg_granule_to_timestamp(oggstream, granule) >
ogg_granule_to_timestamp(oggstream, oggstream->last_granule) + 1 ||
ogg_key_granule(oggstream, granule))) {
if (oggstream->page.granule != -1)
ogg_buffer_page(s, oggstream);
flush = 1;
}
// avoid a continued page
if (!header && oggstream->page.size > 0 &&
MAX_PAGE_SIZE - oggstream->page.size < size) {
ogg_buffer_page(s, oggstream);
}
for (i = 0; i < total_segments; ) {
OGGPage *page = &oggstream->page;
segments = FFMIN(total_segments - i, 255 - page->segments_count);
if (i && !page->segments_count)
page->flags |= 1; // continued packet
memset(page->segments+page->segments_count, 255, segments - 1);
page->segments_count += segments - 1;
len = FFMIN(size, segments*255);
page->segments[page->segments_count++] = len - (segments-1)*255;
memcpy(page->data+page->size, p, len);
p += len;
size -= len;
i += segments;
page->size += len;
if (i == total_segments)
page->granule = granule;
{
AVStream *st = s->streams[page->stream_index];
int64_t start = av_rescale_q(page->start_granule, st->time_base,
AV_TIME_BASE_Q);
int64_t next = av_rescale_q(ogg_granule_to_timestamp(oggstream, page->granule),
st->time_base, AV_TIME_BASE_Q);
if (page->segments_count == 255) {
ogg_buffer_page(s, oggstream);
} else if (!header) {
if ((ogg->pref_size > 0 && page->size >= ogg->pref_size) ||
(ogg->pref_duration > 0 && next - start >= ogg->pref_duration)) {
ogg_buffer_page(s, oggstream);
}
}
}
}
if (flush && oggstream->page.granule != -1)
ogg_buffer_page(s, oggstream);
return 0;
}
static uint8_t *ogg_write_vorbiscomment(int64_t offset, int bitexact,
int *header_len, AVDictionary **m, int framing_bit,
AVChapter **chapters, unsigned int nb_chapters)
{
const char *vendor = bitexact ? "ffmpeg" : LIBAVFORMAT_IDENT;
FFIOContext pb;
int64_t size;
uint8_t *p;
ff_metadata_conv(m, ff_vorbiscomment_metadata_conv, NULL);
size = offset + ff_vorbiscomment_length(*m, vendor, chapters, nb_chapters) + framing_bit;
if (size > INT_MAX)
return NULL;
p = av_mallocz(size);
if (!p)
return NULL;
ffio_init_context(&pb, p + offset, size - offset, 1, NULL, NULL, NULL, NULL);
ff_vorbiscomment_write(&pb.pub, *m, vendor, chapters, nb_chapters);
if (framing_bit)
avio_w8(&pb.pub, 1);
*header_len = size;
return p;
}
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
static int ogg_build_flac_headers(AVCodecParameters *par,
OGGStreamContext *oggstream, int bitexact,
AVDictionary **m)
{
uint8_t *p;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (par->extradata_size < FLAC_STREAMINFO_SIZE)
return AVERROR(EINVAL);
// first packet: STREAMINFO
oggstream->header_len[0] = 51;
oggstream->header[0] = av_mallocz(51); // per ogg flac specs
p = oggstream->header[0];
if (!p)
return AVERROR(ENOMEM);
bytestream_put_byte(&p, 0x7F);
bytestream_put_buffer(&p, "FLAC", 4);
bytestream_put_byte(&p, 1); // major version
bytestream_put_byte(&p, 0); // minor version
bytestream_put_be16(&p, 1); // headers packets without this one
bytestream_put_buffer(&p, "fLaC", 4);
bytestream_put_byte(&p, 0x00); // streaminfo
bytestream_put_be24(&p, 34);
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
bytestream_put_buffer(&p, par->extradata, FLAC_STREAMINFO_SIZE);
// second packet: VorbisComment
p = ogg_write_vorbiscomment(4, bitexact, &oggstream->header_len[1], m, 0, NULL, 0);
if (!p)
return AVERROR(ENOMEM);
oggstream->header[1] = p;
bytestream_put_byte(&p, 0x84); // last metadata block and vorbis comment
bytestream_put_be24(&p, oggstream->header_len[1] - 4);
return 0;
}
#define SPEEX_HEADER_SIZE 80
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
static int ogg_build_speex_headers(AVCodecParameters *par,
OGGStreamContext *oggstream, int bitexact,
AVDictionary **m)
{
uint8_t *p;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (par->extradata_size < SPEEX_HEADER_SIZE)
return AVERROR_INVALIDDATA;
// first packet: Speex header
p = av_mallocz(SPEEX_HEADER_SIZE);
if (!p)
return AVERROR(ENOMEM);
oggstream->header[0] = p;
oggstream->header_len[0] = SPEEX_HEADER_SIZE;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
bytestream_put_buffer(&p, par->extradata, SPEEX_HEADER_SIZE);
AV_WL32(&oggstream->header[0][68], 0); // set extra_headers to 0
// second packet: VorbisComment
p = ogg_write_vorbiscomment(0, bitexact, &oggstream->header_len[1], m, 0, NULL, 0);
if (!p)
return AVERROR(ENOMEM);
oggstream->header[1] = p;
return 0;
}
#define OPUS_HEADER_SIZE 19
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
static int ogg_build_opus_headers(AVCodecParameters *par,
OGGStreamContext *oggstream, int bitexact,
AVDictionary **m, AVChapter **chapters,
unsigned int nb_chapters)
{
uint8_t *p;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (par->extradata_size < OPUS_HEADER_SIZE)
return AVERROR_INVALIDDATA;
/* first packet: Opus header */
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
p = av_mallocz(par->extradata_size);
if (!p)
return AVERROR(ENOMEM);
oggstream->header[0] = p;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
oggstream->header_len[0] = par->extradata_size;
bytestream_put_buffer(&p, par->extradata, par->extradata_size);
/* second packet: VorbisComment */
p = ogg_write_vorbiscomment(8, bitexact, &oggstream->header_len[1], m, 0, chapters, nb_chapters);
if (!p)
return AVERROR(ENOMEM);
oggstream->header[1] = p;
bytestream_put_buffer(&p, "OpusTags", 8);
return 0;
}
#define VP8_HEADER_SIZE 26
static int ogg_build_vp8_headers(AVFormatContext *s, AVStream *st,
OGGStreamContext *oggstream, int bitexact)
{
AVCodecParameters *par = st->codecpar;
uint8_t *p;
/* first packet: VP8 header */
p = av_mallocz(VP8_HEADER_SIZE);
if (!p)
return AVERROR(ENOMEM);
oggstream->header[0] = p;
oggstream->header_len[0] = VP8_HEADER_SIZE;
bytestream_put_byte(&p, 0x4f); // HDRID
bytestream_put_buffer(&p, "VP80", 4); // Identifier
bytestream_put_byte(&p, 1); // HDRTYP
bytestream_put_byte(&p, 1); // VMAJ
bytestream_put_byte(&p, 0); // VMIN
bytestream_put_be16(&p, par->width);
bytestream_put_be16(&p, par->height);
bytestream_put_be24(&p, par->sample_aspect_ratio.num);
bytestream_put_be24(&p, par->sample_aspect_ratio.den);
if (st->r_frame_rate.num > 0 && st->r_frame_rate.den > 0) {
// OggVP8 requires pts to increase by 1 per visible frame, so use the least common
// multiple framerate if available.
av_log(s, AV_LOG_DEBUG, "Changing time base from %d/%d to %d/%d\n",
st->time_base.num, st->time_base.den,
st->r_frame_rate.den, st->r_frame_rate.num);
avpriv_set_pts_info(st, 64, st->r_frame_rate.den, st->r_frame_rate.num);
}
bytestream_put_be32(&p, st->time_base.den);
bytestream_put_be32(&p, st->time_base.num);
/* optional second packet: VorbisComment */
if (av_dict_get(st->metadata, "", NULL, AV_DICT_IGNORE_SUFFIX)) {
p = ogg_write_vorbiscomment(7, bitexact, &oggstream->header_len[1], &st->metadata, 0, NULL, 0);
if (!p)
return AVERROR(ENOMEM);
oggstream->header[1] = p;
bytestream_put_byte(&p, 0x4f); // HDRID
bytestream_put_buffer(&p, "VP80", 4); // Identifier
bytestream_put_byte(&p, 2); // HDRTYP
bytestream_put_byte(&p, 0x20);
}
oggstream->isvp8 = 1;
return 0;
}
static void ogg_write_pages(AVFormatContext *s, int flush)
{
OGGContext *ogg = s->priv_data;
OGGPageList *next, *p;
if (!ogg->page_list)
return;
for (p = ogg->page_list; p; ) {
OGGStreamContext *oggstream =
s->streams[p->page.stream_index]->priv_data;
if (oggstream->page_count < 2 && !flush)
break;
ogg_write_page(s, &p->page,
flush == 1 && oggstream->page_count == 1 ? 4 : 0); // eos
next = p->next;
av_freep(&p);
p = next;
}
ogg->page_list = p;
}
static int ogg_init(AVFormatContext *s)
{
OGGContext *ogg = s->priv_data;
OGGStreamContext *oggstream = NULL;
int i, j;
if (ogg->pref_size)
av_log(s, AV_LOG_WARNING, "The pagesize option is deprecated\n");
for (i = 0; i < s->nb_streams; i++) {
AVStream *st = s->streams[i];
unsigned serial_num = i + ogg->serial_offset;
if (st->codecpar->codec_type == AVMEDIA_TYPE_AUDIO) {
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (st->codecpar->codec_id == AV_CODEC_ID_OPUS)
/* Opus requires a fixed 48kHz clock */
avpriv_set_pts_info(st, 64, 1, 48000);
else
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
avpriv_set_pts_info(st, 64, 1, st->codecpar->sample_rate);
}
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (st->codecpar->codec_id != AV_CODEC_ID_VORBIS &&
st->codecpar->codec_id != AV_CODEC_ID_THEORA &&
st->codecpar->codec_id != AV_CODEC_ID_SPEEX &&
st->codecpar->codec_id != AV_CODEC_ID_FLAC &&
st->codecpar->codec_id != AV_CODEC_ID_OPUS &&
st->codecpar->codec_id != AV_CODEC_ID_VP8) {
av_log(s, AV_LOG_ERROR, "Unsupported codec id in stream %d\n", i);
return AVERROR(EINVAL);
}
if ((!st->codecpar->extradata || !st->codecpar->extradata_size) &&
st->codecpar->codec_id != AV_CODEC_ID_VP8) {
av_log(s, AV_LOG_ERROR, "No extradata present\n");
return AVERROR_INVALIDDATA;
}
oggstream = av_mallocz(sizeof(*oggstream));
if (!oggstream)
return AVERROR(ENOMEM);
oggstream->page.stream_index = i;
if (!(s->flags & AVFMT_FLAG_BITEXACT))
do {
serial_num = av_get_random_seed();
for (j = 0; j < i; j++) {
OGGStreamContext *sc = s->streams[j]->priv_data;
if (serial_num == sc->serial_num)
break;
}
} while (j < i);
oggstream->serial_num = serial_num;
av_dict_copy(&st->metadata, s->metadata, AV_DICT_DONT_OVERWRITE);
st->priv_data = oggstream;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (st->codecpar->codec_id == AV_CODEC_ID_FLAC) {
int err = ogg_build_flac_headers(st->codecpar, oggstream,
s->flags & AVFMT_FLAG_BITEXACT,
&st->metadata);
if (err) {
av_log(s, AV_LOG_ERROR, "Error writing FLAC headers\n");
return err;
}
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
} else if (st->codecpar->codec_id == AV_CODEC_ID_SPEEX) {
int err = ogg_build_speex_headers(st->codecpar, oggstream,
s->flags & AVFMT_FLAG_BITEXACT,
&st->metadata);
if (err) {
av_log(s, AV_LOG_ERROR, "Error writing Speex headers\n");
return err;
}
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
} else if (st->codecpar->codec_id == AV_CODEC_ID_OPUS) {
int err = ogg_build_opus_headers(st->codecpar, oggstream,
s->flags & AVFMT_FLAG_BITEXACT,
&st->metadata, s->chapters, s->nb_chapters);
if (err) {
av_log(s, AV_LOG_ERROR, "Error writing Opus headers\n");
return err;
}
} else if (st->codecpar->codec_id == AV_CODEC_ID_VP8) {
int err = ogg_build_vp8_headers(s, st, oggstream,
s->flags & AVFMT_FLAG_BITEXACT);
if (err) {
av_log(s, AV_LOG_ERROR, "Error writing VP8 headers\n");
return err;
}
} else {
uint8_t *p;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
const char *cstr = st->codecpar->codec_id == AV_CODEC_ID_VORBIS ? "vorbis" : "theora";
int header_type = st->codecpar->codec_id == AV_CODEC_ID_VORBIS ? 3 : 0x81;
int framing_bit = st->codecpar->codec_id == AV_CODEC_ID_VORBIS ? 1 : 0;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (avpriv_split_xiph_headers(st->codecpar->extradata, st->codecpar->extradata_size,
st->codecpar->codec_id == AV_CODEC_ID_VORBIS ? 30 : 42,
(const uint8_t**)oggstream->header, oggstream->header_len) < 0) {
av_log(s, AV_LOG_ERROR, "Extradata corrupted\n");
oggstream->header[1] = NULL;
return AVERROR_INVALIDDATA;
}
p = ogg_write_vorbiscomment(7, s->flags & AVFMT_FLAG_BITEXACT,
&oggstream->header_len[1], &st->metadata,
framing_bit, NULL, 0);
oggstream->header[1] = p;
if (!p)
return AVERROR(ENOMEM);
bytestream_put_byte(&p, header_type);
bytestream_put_buffer(&p, cstr, 6);
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (st->codecpar->codec_id == AV_CODEC_ID_THEORA) {
int den = AV_RB32(oggstream->header[0] + 22), num = AV_RB32(oggstream->header[0] + 26);
/* Make sure to use time base stored in the Theora stream header to write
correct timestamps */
if (st->time_base.num != num || st->time_base.den != den) {
av_log(s, AV_LOG_DEBUG, "Changing time base from %d/%d to %d/%d\n",
st->time_base.num, st->time_base.den, num, den);
avpriv_set_pts_info(st, 64, num, den);
}
/** KFGSHIFT is the width of the less significant section of the granule position
The less significant section is the frame count since the last keyframe */
oggstream->kfgshift = ((oggstream->header[0][40]&3)<<3)|(oggstream->header[0][41]>>5);
oggstream->vrev = oggstream->header[0][9];
av_log(s, AV_LOG_DEBUG, "theora kfgshift %d, vrev %d\n",
oggstream->kfgshift, oggstream->vrev);
}
}
}
return 0;
}
static int ogg_write_header(AVFormatContext *s)
{
OGGStreamContext *oggstream = NULL;
int i, j;
for (j = 0; j < s->nb_streams; j++) {
oggstream = s->streams[j]->priv_data;
ogg_buffer_data(s, s->streams[j], oggstream->header[0],
oggstream->header_len[0], 0, 1);
oggstream->page.flags |= 2; // bos
ogg_buffer_page(s, oggstream);
}
for (j = 0; j < s->nb_streams; j++) {
AVStream *st = s->streams[j];
oggstream = st->priv_data;
for (i = 1; i < 3; i++) {
if (oggstream->header_len[i])
ogg_buffer_data(s, st, oggstream->header[i],
oggstream->header_len[i], 0, 1);
}
ogg_buffer_page(s, oggstream);
}
oggstream->page.start_granule = AV_NOPTS_VALUE;
ogg_write_pages(s, 2);
return 0;
}
static int ogg_write_packet_internal(AVFormatContext *s, AVPacket *pkt)
{
AVStream *st = s->streams[pkt->stream_index];
OGGStreamContext *oggstream = st->priv_data;
int ret;
int64_t granule;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (st->codecpar->codec_id == AV_CODEC_ID_THEORA) {
int64_t pts = oggstream->vrev < 1 ? pkt->pts : pkt->pts + pkt->duration;
int pframe_count;
if (pkt->flags & AV_PKT_FLAG_KEY)
oggstream->last_kf_pts = pts;
pframe_count = pts - oggstream->last_kf_pts;
// prevent frame count from overflow if key frame flag is not set
if (pframe_count >= (1<<oggstream->kfgshift)) {
oggstream->last_kf_pts += pframe_count;
pframe_count = 0;
}
granule = (oggstream->last_kf_pts<<oggstream->kfgshift) | pframe_count;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
} else if (st->codecpar->codec_id == AV_CODEC_ID_OPUS)
granule = pkt->pts + pkt->duration +
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
av_rescale_q(st->codecpar->initial_padding,
(AVRational){ 1, st->codecpar->sample_rate },
st->time_base);
else if (st->codecpar->codec_id == AV_CODEC_ID_VP8) {
int64_t pts, invcnt, dist;
int visible;
visible = (pkt->data[0] >> 4) & 1;
pts = pkt->pts + pkt->duration;
invcnt = (oggstream->last_granule >> 30) & 3;
invcnt = visible ? 3 : (invcnt == 3 ? 0 : invcnt + 1);
dist = (pkt->flags & AV_PKT_FLAG_KEY) ? 0 : ((oggstream->last_granule >> 3) & 0x07ffffff) + 1;
granule = (pts << 32) | (invcnt << 30) | (dist << 3);
} else
granule = pkt->pts + pkt->duration;
if (oggstream->page.start_granule == AV_NOPTS_VALUE)
oggstream->page.start_granule = pkt->pts;
ret = ogg_buffer_data(s, st, pkt->data, pkt->size, granule, 0);
if (ret < 0)
return ret;
ogg_write_pages(s, 0);
oggstream->last_granule = granule;
return 0;
}
static int ogg_write_packet(AVFormatContext *s, AVPacket *pkt)
{
int i;
if (pkt)
return ogg_write_packet_internal(s, pkt);
for (i = 0; i < s->nb_streams; i++) {
OGGStreamContext *oggstream = s->streams[i]->priv_data;
if (oggstream->page.segments_count)
ogg_buffer_page(s, oggstream);
}
ogg_write_pages(s, 2);
return 1;
}
static int ogg_write_trailer(AVFormatContext *s)
{
int i;
/* flush current page if needed */
for (i = 0; i < s->nb_streams; i++) {
OGGStreamContext *oggstream = s->streams[i]->priv_data;
if (oggstream->page.size > 0)
ogg_buffer_page(s, oggstream);
}
ogg_write_pages(s, 1);
return 0;
}
static void ogg_free(AVFormatContext *s)
{
OGGContext *ogg = s->priv_data;
OGGPageList *p = ogg->page_list;
int i;
for (i = 0; i < s->nb_streams; i++) {
AVStream *st = s->streams[i];
OGGStreamContext *oggstream = st->priv_data;
if (!oggstream)
continue;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (st->codecpar->codec_id == AV_CODEC_ID_FLAC ||
st->codecpar->codec_id == AV_CODEC_ID_SPEEX ||
st->codecpar->codec_id == AV_CODEC_ID_OPUS ||
st->codecpar->codec_id == AV_CODEC_ID_VP8) {
av_freep(&oggstream->header[0]);
}
av_freep(&oggstream->header[1]);
}
while (p) {
OGGPageList *next = p->next;
av_free(p);
p = next;
}
ogg->page_list = NULL;
}
#if CONFIG_OGG_MUXER
const AVOutputFormat ff_ogg_muxer = {
.name = "ogg",
.long_name = NULL_IF_CONFIG_SMALL("Ogg"),
.mime_type = "application/ogg",
.extensions = "ogg"
#if !CONFIG_OGV_MUXER
",ogv"
#endif
#if !CONFIG_SPX_MUXER
",spx"
#endif
#if !CONFIG_OPUS_MUXER
",opus"
#endif
,
.priv_data_size = sizeof(OGGContext),
.audio_codec = CONFIG_LIBVORBIS_ENCODER ?
AV_CODEC_ID_VORBIS : AV_CODEC_ID_FLAC,
.video_codec = AV_CODEC_ID_THEORA,
.init = ogg_init,
.write_header = ogg_write_header,
.write_packet = ogg_write_packet,
.write_trailer = ogg_write_trailer,
.deinit = ogg_free,
.flags = AVFMT_TS_NEGATIVE | AVFMT_TS_NONSTRICT | AVFMT_ALLOW_FLUSH,
.priv_class = &ogg_muxer_class,
};
#endif
#if CONFIG_OGA_MUXER
const AVOutputFormat ff_oga_muxer = {
.name = "oga",
.long_name = NULL_IF_CONFIG_SMALL("Ogg Audio"),
.mime_type = "audio/ogg",
.extensions = "oga",
.priv_data_size = sizeof(OGGContext),
.audio_codec = AV_CODEC_ID_FLAC,
.init = ogg_init,
.write_header = ogg_write_header,
.write_packet = ogg_write_packet,
.write_trailer = ogg_write_trailer,
.deinit = ogg_free,
.flags = AVFMT_TS_NEGATIVE | AVFMT_ALLOW_FLUSH,
.priv_class = &ogg_muxer_class,
};
#endif
#if CONFIG_OGV_MUXER
const AVOutputFormat ff_ogv_muxer = {
.name = "ogv",
.long_name = NULL_IF_CONFIG_SMALL("Ogg Video"),
.mime_type = "video/ogg",
.extensions = "ogv",
.priv_data_size = sizeof(OGGContext),
.audio_codec = CONFIG_LIBVORBIS_ENCODER ?
AV_CODEC_ID_VORBIS : AV_CODEC_ID_FLAC,
.video_codec = CONFIG_LIBTHEORA_ENCODER ?
AV_CODEC_ID_THEORA : AV_CODEC_ID_VP8,
.init = ogg_init,
.write_header = ogg_write_header,
.write_packet = ogg_write_packet,
.write_trailer = ogg_write_trailer,
.deinit = ogg_free,
.flags = AVFMT_TS_NEGATIVE | AVFMT_TS_NONSTRICT | AVFMT_ALLOW_FLUSH,
.priv_class = &ogg_muxer_class,
};
#endif
#if CONFIG_SPX_MUXER
const AVOutputFormat ff_spx_muxer = {
.name = "spx",
.long_name = NULL_IF_CONFIG_SMALL("Ogg Speex"),
.mime_type = "audio/ogg",
.extensions = "spx",
.priv_data_size = sizeof(OGGContext),
.audio_codec = AV_CODEC_ID_SPEEX,
.init = ogg_init,
.write_header = ogg_write_header,
.write_packet = ogg_write_packet,
.write_trailer = ogg_write_trailer,
.deinit = ogg_free,
.flags = AVFMT_TS_NEGATIVE | AVFMT_ALLOW_FLUSH,
.priv_class = &ogg_muxer_class,
};
#endif
#if CONFIG_OPUS_MUXER
const AVOutputFormat ff_opus_muxer = {
.name = "opus",
.long_name = NULL_IF_CONFIG_SMALL("Ogg Opus"),
.mime_type = "audio/ogg",
.extensions = "opus",
.priv_data_size = sizeof(OGGContext),
.audio_codec = AV_CODEC_ID_OPUS,
.init = ogg_init,
.write_header = ogg_write_header,
.write_packet = ogg_write_packet,
.write_trailer = ogg_write_trailer,
.deinit = ogg_free,
.flags = AVFMT_TS_NEGATIVE | AVFMT_ALLOW_FLUSH,
.priv_class = &ogg_muxer_class,
};
#endif