1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-12-12 19:18:44 +02:00
FFmpeg/libavformat/internal.h

1023 lines
32 KiB
C
Raw Normal View History

/*
* copyright (c) 2001 Fabrice Bellard
*
* 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
*/
#ifndef AVFORMAT_INTERNAL_H
#define AVFORMAT_INTERNAL_H
#include <stdint.h>
#include "libavcodec/avcodec.h"
#include "libavcodec/bsf.h"
#include "avformat.h"
#include "os_support.h"
#define MAX_URL_SIZE 4096
/** size of probe buffer, for guessing file type from file contents */
#define PROBE_BUF_MIN 2048
#define PROBE_BUF_MAX (1 << 20)
#ifdef DEBUG
# define hex_dump_debug(class, buf, size) av_hex_dump_log(class, AV_LOG_DEBUG, buf, size)
#else
# define hex_dump_debug(class, buf, size) do { if (0) av_hex_dump_log(class, AV_LOG_DEBUG, buf, size); } while(0)
#endif
/**
* For an AVInputFormat with this flag set read_close() needs to be called
* by the caller upon read_header() failure.
*/
#define FF_FMT_INIT_CLEANUP (1 << 0)
typedef struct AVCodecTag {
enum AVCodecID id;
unsigned int tag;
} AVCodecTag;
typedef struct CodecMime{
char str[32];
enum AVCodecID id;
} CodecMime;
/*************************************************/
/* fractional numbers for exact pts handling */
/**
* The exact value of the fractional number is: 'val + num / den'.
* num is assumed to be 0 <= num < den.
*/
typedef struct FFFrac {
int64_t val, num, den;
} FFFrac;
typedef struct FFFormatContext {
/**
* The public context.
*/
AVFormatContext pub;
/**
* Number of streams relevant for interleaving.
* Muxing only.
*/
int nb_interleaved_streams;
/**
* The interleavement function in use. Always set for muxers.
*/
int (*interleave_packet)(struct AVFormatContext *s, AVPacket *pkt,
int flush, int has_packet);
/**
* This buffer is only needed when packets were already buffered but
* not decoded, for example to get the codec parameters in MPEG
* streams.
*/
struct PacketList *packet_buffer;
struct PacketList *packet_buffer_end;
/* av_seek_frame() support */
int64_t data_offset; /**< offset of the first packet */
/**
* Raw packets from the demuxer, prior to parsing and decoding.
* This buffer is used for buffering packets until the codec can
* be identified, as parsing cannot be done without knowing the
* codec.
*/
struct PacketList *raw_packet_buffer;
struct PacketList *raw_packet_buffer_end;
/**
* Packets split by the parser get queued here.
*/
struct PacketList *parse_queue;
struct PacketList *parse_queue_end;
/**
* The generic code uses this as a temporary packet
* to parse packets or for muxing, especially flushing.
* For demuxers, it may also be used for other means
* for short periods that are guaranteed not to overlap
* with calls to av_read_frame() (or ff_read_packet())
* or with each other.
* It may be used by demuxers as a replacement for
* stack packets (unless they call one of the aforementioned
* functions with their own AVFormatContext).
* Every user has to ensure that this packet is blank
* after using it.
*/
AVPacket *parse_pkt;
/**
* Used to hold temporary packets for the generic demuxing code.
* When muxing, it may be used by muxers to hold packets (even
* permanent ones).
*/
AVPacket *pkt;
/**
* Sum of the size of packets in raw_packet_buffer, in bytes.
*/
int raw_packet_buffer_size;
/**
* Offset to remap timestamps to be non-negative.
* Expressed in timebase units.
* @see AVStream.mux_ts_offset
*/
int64_t offset;
/**
* Timebase for the timestamp offset.
*/
AVRational offset_timebase;
#if FF_API_COMPUTE_PKT_FIELDS2
int missing_ts_warning;
#endif
int inject_global_side_data;
int avoid_negative_ts_use_pts;
/**
* Timestamp of the end of the shortest stream.
*/
int64_t shortest_end;
/**
* Whether or not avformat_init_output has already been called
*/
int initialized;
/**
* Whether or not avformat_init_output fully initialized streams
*/
int streams_initialized;
/**
* ID3v2 tag useful for MP3 demuxing
*/
AVDictionary *id3v2_meta;
/*
* Prefer the codec framerate for avg_frame_rate computation.
*/
int prefer_codec_framerate;
/**
* Set if chapter ids are strictly monotonic.
*/
int chapter_ids_monotonic;
} FFFormatContext;
static av_always_inline FFFormatContext *ffformatcontext(AVFormatContext *s)
{
return (FFFormatContext*)s;
}
typedef struct FFStream {
/**
* The public context.
*/
AVStream pub;
/**
* Set to 1 if the codec allows reordering, so pts can be different
* from dts.
*/
int reorder;
/**
* bitstream filter to run on stream
* - encoding: Set by muxer using ff_stream_add_bitstream_filter
* - decoding: unused
*/
AVBSFContext *bsfc;
/**
* Whether or not check_bitstream should still be run on each packet
*/
int bitstream_checked;
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.
2014-06-18 21:42:52 +03:00
/**
* The codec context used by avformat_find_stream_info, the parser, etc.
*/
AVCodecContext *avctx;
/**
* 1 if avctx has been initialized with the values from the codec parameters
*/
int avctx_inited;
/* the context for extracting extradata in find_stream_info()
* inited=1/bsf=NULL signals that extracting is not possible (codec not
* supported) */
struct {
AVBSFContext *bsf;
int inited;
} extract_extradata;
/**
* Whether the internal avctx needs to be updated from codecpar (after a late change to codecpar)
*/
int need_context_update;
avformat/mux: Set AV_PKT_FLAG_KEY for is_intra_only packet The patch will make audio and subtitle packets be marked as AV_PKT_FLAG_KEY. For audio, it'll caused the audio sample to be sync sample. To verify ref/fate/movenc results: 1. Get the movenc test data [lmwang@vpn ffmpeg]$ libavformat/tests/movenc -w && mkdir -p audio_old && mv *.mp4 audio_old_ After applied the patch: [lmwang@vpn ffmpeg]$ make fate-movenc SAMPLES=../fate-suite [lmwang@vpn ffmpeg]$ libavformat/tests/movenc -w && mkdir -p audio_key && mv *.mp4 audio_key 2. Get l-smash and build boxdumper https://github.com/l-smash/l-smash.git 3. dump the box of crc change mp4 and diff -u [lmwang@vpn ffmpeg]$ ../l-smash/cli/boxdumper --box audio_key/non-empty-moov-no-elst.mp4 > audio_key/non-empty-moov-no-elst.log [lmwang@vpn ffmpeg]$ ../l-smash/cli/boxdumper --box audio_old/non-empty-moov-no-elst.mp4 > audio_old/non-empty-moov-no-elst.log [lmwang@vpn ffmpeg]$ diff -u audio_key/non-empty-moov-no-elst.log audio_old/non-empty-moov-no-elst.log - default_sample_flags = 0x02000000 - independent - sync sample + default_sample_flags = 0x01010000 + dependent + non-sync sample 4. have checked the change of crc are caused by default_sample_flags non-empty-moov.mp4, non-empty-moov-elst.mp4, non-empty-moov-no-elst.mp4, empty-moov.mp4, delay-moov-content.mp4, empty-moov-second-frag.mp4, empty-moov-second-frag-discont.mp4, delay-moov-second-frag-discont.mp4, delay-moov-elst-second-frag.mp4 etc 5 For subtitle, it'll effect for tests/ref/fate/binsub-movtextenc and tests/ref/fate/sub2video, that's expecting result for the subtitle is marked as keyframe. Below is the checking result of binsub-movtextenc: [lmwang@vpn ffmpeg]$ ./ffmpeg -i ../fate-suite/sub/MovText_capability_tester.mp4 -map 0 -scodec mov_text -f mp4 -flags +bitexact -fflags +bitexact -movflags frag_keyframe+empty_moov audio_key/binsub-movtextenc.mp4 [lmwang@vpn ffmpeg]$ ./ffmpeg -i ../fate-suite/sub/MovText_capability_tester.mp4 -map 0 -scodec mov_text -f mp4 -flags +bitexact -fflags +bitexact -movflags frag_keyframe+empty_moov audio_old/binsub-movtextenc.mp4 [lmwang@vpn ffmpeg]$../l-smash/cli/boxdumper audio_key/binsub-movtextenc.mp4 > audio_key/binsub-movtextenc.log [lmwang@vpn ffmpeg]$../l-smash/cli/boxdumper audio_old/binsub-movtextenc.mp4 > audio_old/binsub-movtextenc.log [lmwang@vpn ffmpeg]$ diff -u audio_key/binsub-movtextenc.log audio_old/binsub-movtextenc.log .... // the key difference is the flag for sync sample - flags = 0x000701 + flags = 0x000301 data-offset-present sample-duration-present sample-size-present - sample-flags-present sample_count = 6 - data_offset = 188 + data_offset = 164 sample[0] sample_duration = 1570000 sample_size = 21 - sample_flags = 0x02000000 - independent - sync sample - degradation_priority = 0 sample[1] sample_duration = 510000 sample_size = 2 - sample_flags = 0x01010000 - dependent - non-sync sample - degradation_priority = 0 sample[2] sample_duration = 1690000 sample_size = 9 - sample_flags = 0x02000000 - independent - sync sample - degradation_priority = 0 Suggested-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Suggested-by: Nicolas George <george@nsup.org> Signed-off-by: Limin Wang <lance.lmwang@gmail.com>
2020-04-21 17:35:24 +02:00
int is_intra_only;
FFFrac *priv_pts;
#define MAX_STD_TIMEBASES (30*12+30+3+6)
/**
* Stream information used internally by avformat_find_stream_info()
*/
struct {
int64_t last_dts;
int64_t duration_gcd;
int duration_count;
int64_t rfps_duration_sum;
double (*duration_error)[2][MAX_STD_TIMEBASES];
int64_t codec_info_duration;
int64_t codec_info_duration_fields;
int frame_delay_evidence;
/**
* 0 -> decoder has not been searched for yet.
* >0 -> decoder found
* <0 -> decoder with codec_id == -found_decoder has not been found
*/
int found_decoder;
int64_t last_duration;
/**
* Those are used for average framerate estimation.
*/
int64_t fps_first_dts;
int fps_first_dts_idx;
int64_t fps_last_dts;
int fps_last_dts_idx;
} *info;
AVIndexEntry *index_entries; /**< Only used if the format does not
support seeking natively. */
int nb_index_entries;
unsigned int index_entries_allocated_size;
int64_t interleaver_chunk_size;
int64_t interleaver_chunk_duration;
/**
* stream probing state
* -1 -> probing finished
* 0 -> no probing requested
* rest -> perform probing with request_probe being the minimum score to accept.
*/
int request_probe;
/**
* Indicates that everything up to the next keyframe
* should be discarded.
*/
int skip_to_keyframe;
/**
* Number of samples to skip at the start of the frame decoded from the next packet.
*/
int skip_samples;
/**
* If not 0, the number of samples that should be skipped from the start of
* the stream (the samples are removed from packets with pts==0, which also
* assumes negative timestamps do not happen).
* Intended for use with formats such as mp3 with ad-hoc gapless audio
* support.
*/
int64_t start_skip_samples;
/**
* If not 0, the first audio sample that should be discarded from the stream.
* This is broken by design (needs global sample count), but can't be
* avoided for broken by design formats such as mp3 with ad-hoc gapless
* audio support.
*/
int64_t first_discard_sample;
/**
* The sample after last sample that is intended to be discarded after
* first_discard_sample. Works on frame boundaries only. Used to prevent
* early EOF if the gapless info is broken (considered concatenated mp3s).
*/
int64_t last_discard_sample;
/**
* Number of internally decoded frames, used internally in libavformat, do not access
* its lifetime differs from info which is why it is not in that structure.
*/
int nb_decoded_frames;
/**
* Timestamp offset added to timestamps before muxing
*/
int64_t mux_ts_offset;
/**
* Internal data to check for wrapping of the time stamp
*/
int64_t pts_wrap_reference;
/**
* Options for behavior, when a wrap is detected.
*
* Defined by AV_PTS_WRAP_ values.
*
* If correction is enabled, there are two possibilities:
* If the first time stamp is near the wrap point, the wrap offset
* will be subtracted, which will create negative time stamps.
* Otherwise the offset will be added.
*/
int pts_wrap_behavior;
/**
* Internal data to prevent doing update_initial_durations() twice
*/
int update_initial_durations_done;
#define MAX_REORDER_DELAY 16
/**
* Internal data to generate dts from pts
*/
int64_t pts_reorder_error[MAX_REORDER_DELAY+1];
uint8_t pts_reorder_error_count[MAX_REORDER_DELAY+1];
int64_t pts_buffer[MAX_REORDER_DELAY+1];
/**
* Internal data to analyze DTS and detect faulty mpeg streams
*/
int64_t last_dts_for_order_check;
uint8_t dts_ordered;
uint8_t dts_misordered;
/**
* Internal data to inject global side data
*/
int inject_global_side_data;
/**
* display aspect ratio (0 if unknown)
* - encoding: unused
* - decoding: Set by libavformat to calculate sample_aspect_ratio internally
*/
AVRational display_aspect_ratio;
AVProbeData probe_data;
/**
* last packet in packet_buffer for this stream when muxing.
*/
struct PacketList *last_in_packet_buffer;
int64_t last_IP_pts;
int last_IP_duration;
/**
* Number of packets to buffer for codec probing
*/
int probe_packets;
/* av_read_frame() support */
enum AVStreamParseType need_parsing;
struct AVCodecParserContext *parser;
/**
* Number of frames that have been demuxed during avformat_find_stream_info()
*/
int codec_info_nb_frames;
/**
* Stream Identifier
* This is the MPEG-TS stream identifier +1
* 0 means unknown
*/
int stream_identifier;
// Timestamp generation support:
/**
* Timestamp corresponding to the last dts sync point.
*
* Initialized when AVCodecParserContext.dts_sync_point >= 0 and
* a DTS is received from the underlying container. Otherwise set to
* AV_NOPTS_VALUE by default.
*/
int64_t first_dts;
int64_t cur_dts;
} FFStream;
static av_always_inline FFStream *ffstream(AVStream *st)
{
return (FFStream*)st;
}
static av_always_inline const FFStream *cffstream(const AVStream *st)
{
return (FFStream*)st;
}
void avpriv_stream_set_need_parsing(AVStream *st, enum AVStreamParseType type);
#ifdef __GNUC__
#define dynarray_add(tab, nb_ptr, elem)\
do {\
__typeof__(tab) _tab = (tab);\
__typeof__(elem) _elem = (elem);\
(void)sizeof(**_tab == _elem); /* check that types are compatible */\
av_dynarray_add(_tab, nb_ptr, _elem);\
} while(0)
#else
#define dynarray_add(tab, nb_ptr, elem)\
do {\
av_dynarray_add((tab), nb_ptr, (elem));\
} while(0)
#endif
#define RELATIVE_TS_BASE (INT64_MAX - (1LL << 48))
static av_always_inline int is_relative(int64_t ts)
{
return ts > (RELATIVE_TS_BASE - (1LL << 48));
}
/**
* Wrap a given time stamp, if there is an indication for an overflow
*
* @param st stream
* @param timestamp the time stamp to wrap
* @return resulting time stamp
*/
int64_t ff_wrap_timestamp(const AVStream *st, int64_t timestamp);
void ff_flush_packet_queue(AVFormatContext *s);
/**
* Automatically create sub-directories
*
* @param path will create sub-directories by path
* @return 0, or < 0 on error
*/
int ff_mkdir_p(const char *path);
/**
* Write hexadecimal string corresponding to given binary data. The string
* is zero-terminated.
*
* @param buf the output string is written here;
* needs to be at least 2 * size + 1 bytes long.
* @param src the input data to be transformed.
* @param size the size (in byte) of src.
* @param lowercase determines whether to use the range [0-9a-f] or [0-9A-F].
* @return buf.
*/
char *ff_data_to_hex(char *buf, const uint8_t *src, int size, int lowercase);
/**
* Parse a string of hexadecimal strings. Any space between the hexadecimal
* digits is ignored.
*
* @param data if non-null, the parsed data is written to this pointer
* @param p the string to parse
* @return the number of bytes written (or to be written, if data is null)
*/
int ff_hex_to_data(uint8_t *data, const char *p);
/**
* Add packet to an AVFormatContext's packet_buffer list, determining its
* interleaved position using compare() function argument.
* @return 0 on success, < 0 on error. pkt will always be blank on return.
*/
int ff_interleave_add_packet(AVFormatContext *s, AVPacket *pkt,
int (*compare)(AVFormatContext *, const AVPacket *, const AVPacket *));
void ff_read_frame_flush(AVFormatContext *s);
#define NTP_OFFSET 2208988800ULL
#define NTP_OFFSET_US (NTP_OFFSET * 1000000ULL)
/** Get the current time since NTP epoch in microseconds. */
uint64_t ff_ntp_time(void);
/**
* Get the NTP time stamp formatted as per the RFC-5905.
*
* @param ntp_time NTP time in micro seconds (since NTP epoch)
* @return the formatted NTP time stamp
*/
uint64_t ff_get_formatted_ntp_time(uint64_t ntp_time_us);
/**
* Parse the NTP time in micro seconds (since NTP epoch).
*
* @param ntp_ts NTP time stamp formatted as per the RFC-5905.
* @return the time in micro seconds (since NTP epoch)
*/
uint64_t ff_parse_ntp_time(uint64_t ntp_ts);
/**
* Append the media-specific SDP fragment for the media stream c
* to the buffer buff.
*
* Note, the buffer needs to be initialized, since it is appended to
* existing content.
*
* @param buff the buffer to append the SDP fragment to
* @param size the size of the buff buffer
* @param st the AVStream of the media to describe
* @param idx the global stream index
* @param dest_addr the destination address of the media stream, may be NULL
* @param dest_type the destination address type, may be NULL
* @param port the destination port of the media stream, 0 if unknown
* @param ttl the time to live of the stream, 0 if not multicast
* @param fmt the AVFormatContext, which might contain options modifying
* the generated SDP
* @return 0 on success, a negative error code on failure
*/
int ff_sdp_write_media(char *buff, int size, const AVStream *st, int idx,
const char *dest_addr, const char *dest_type,
int port, int ttl, AVFormatContext *fmt);
/**
* Write a packet to another muxer than the one the user originally
* intended. Useful when chaining muxers, where one muxer internally
* writes a received packet to another muxer.
*
* @param dst the muxer to write the packet to
* @param dst_stream the stream index within dst to write the packet to
avformat/mux: Fix double-free when using AVPacket.opaque_ref Up until now, ff_write_chained() copied the packet (manually, not with av_packet_move_ref()) from a packet given to it to a stack packet whose timing and stream_index is then modified before being sent to another muxer via av_(interleaved_)write_frame(). Afterwards it is intended to sync the fields of the packet relevant to freeing again; yet this only encompasses buf, side_data and side_data_elems and not the newly added opaque_ref. The other fields are not synced so that the returned packet can have a size > 0 and data != NULL despite its buf being NULL (this always happens in the interleaved codepath; before commit fe251f77c80b0512ab8907902e1dbed3f4fe1aad it could also happen in the noninterleaved one). This leads to double-frees if the interleaved codepath is used and opaque_ref is set. This commit therefore changes this by directly reusing the packet instead of a spare packet. Given that av_write_frame() does not change the packet given to it, one only needs to restore the timing information to return it as it was; for the interleaved codepath it is not possible to do likewise*, because av_interleaved_write_frame() takes ownership of the packets given to it and returns blank packets. But precisely because of this users of the interleaved codepath have no legitimate expectation that their packet will be returned unchanged. In line with av_interleaved_write_frame() ff_write_chained() therefore returns blank packets when using the interleaved codepath. Making the only user of said codepath compatible with this was trivial. *: Unless one wanted to create a full new reference. Reviewed-by: Lynne <dev@lynne.ee> Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2021-09-01 21:14:41 +02:00
* @param pkt the packet to be written. It will be returned blank when
* av_interleaved_write_frame() is used, unchanged otherwise.
* @param src the muxer the packet originally was intended for
* @param interleave 0->use av_write_frame, 1->av_interleaved_write_frame
* @return the value av_write_frame returned
*/
int ff_write_chained(AVFormatContext *dst, int dst_stream, AVPacket *pkt,
AVFormatContext *src, int interleave);
/**
* Read a whole line of text from AVIOContext. Stop reading after reaching
* either a \\n, a \\0 or EOF. The returned string is always \\0-terminated,
* and may be truncated if the buffer is too small.
*
* @param s the read-only AVIOContext
* @param buf buffer to store the read line
* @param maxlen size of the buffer
* @return the length of the string written in the buffer, not including the
* final \\0
*/
int ff_get_line(AVIOContext *s, char *buf, int maxlen);
/**
* Same as ff_get_line but strip the white-space characters in the text tail
*
* @param s the read-only AVIOContext
* @param buf buffer to store the read line
* @param maxlen size of the buffer
* @return the length of the string written in the buffer
*/
int ff_get_chomp_line(AVIOContext *s, char *buf, int maxlen);
#define SPACE_CHARS " \t\r\n"
/**
* Callback function type for ff_parse_key_value.
*
* @param key a pointer to the key
* @param key_len the number of bytes that belong to the key, including the '='
* char
* @param dest return the destination pointer for the value in *dest, may
* be null to ignore the value
* @param dest_len the length of the *dest buffer
*/
typedef void (*ff_parse_key_val_cb)(void *context, const char *key,
int key_len, char **dest, int *dest_len);
/**
* Parse a string with comma-separated key=value pairs. The value strings
* may be quoted and may contain escaped characters within quoted strings.
*
* @param str the string to parse
* @param callback_get_buf function that returns where to store the
* unescaped value string.
* @param context the opaque context pointer to pass to callback_get_buf
*/
void ff_parse_key_value(const char *str, ff_parse_key_val_cb callback_get_buf,
void *context);
/**
* Find stream index based on format-specific stream ID
* @return stream index, or < 0 on error
*/
int ff_find_stream_index(const AVFormatContext *s, int id);
/**
* Internal version of av_index_search_timestamp
*/
int ff_index_search_timestamp(const AVIndexEntry *entries, int nb_entries,
int64_t wanted_timestamp, int flags);
/**
* Internal version of av_add_index_entry
*/
int ff_add_index_entry(AVIndexEntry **index_entries,
int *nb_index_entries,
unsigned int *index_entries_allocated_size,
int64_t pos, int64_t timestamp, int size, int distance, int flags);
void ff_configure_buffers_for_index(AVFormatContext *s, int64_t time_tolerance);
/**
* Add a new chapter.
*
* @param s media file handle
* @param id unique ID for this chapter
* @param start chapter start time in time_base units
* @param end chapter end time in time_base units
* @param title chapter title
*
* @return AVChapter or NULL on error
*/
AVChapter *avpriv_new_chapter(AVFormatContext *s, int64_t id, AVRational time_base,
int64_t start, int64_t end, const char *title);
/**
* Ensure the index uses less memory than the maximum specified in
* AVFormatContext.max_index_size by discarding entries if it grows
* too large.
*/
void ff_reduce_index(AVFormatContext *s, int stream_index);
enum AVCodecID ff_guess_image2_codec(const char *filename);
const AVCodec *ff_find_decoder(AVFormatContext *s, const AVStream *st,
enum AVCodecID codec_id);
/**
* Perform a binary search using av_index_search_timestamp() and
* AVInputFormat.read_timestamp().
*
* @param target_ts target timestamp in the time base of the given stream
* @param stream_index stream number
*/
int ff_seek_frame_binary(AVFormatContext *s, int stream_index,
int64_t target_ts, int flags);
/**
* Update cur_dts of all streams based on the given timestamp and AVStream.
*
* Stream ref_st unchanged, others set cur_dts in their native time base.
* Only needed for timestamp wrapping or if (dts not set and pts!=dts).
* @param timestamp new dts expressed in time_base of param ref_st
* @param ref_st reference stream giving time_base of param timestamp
*/
void avpriv_update_cur_dts(AVFormatContext *s, AVStream *ref_st, int64_t timestamp);
int ff_find_last_ts(AVFormatContext *s, int stream_index, int64_t *ts, int64_t *pos,
int64_t (*read_timestamp)(struct AVFormatContext *, int , int64_t *, int64_t ));
/**
* Perform a binary search using read_timestamp().
*
* @param target_ts target timestamp in the time base of the given stream
* @param stream_index stream number
*/
int64_t ff_gen_search(AVFormatContext *s, int stream_index,
int64_t target_ts, int64_t pos_min,
int64_t pos_max, int64_t pos_limit,
int64_t ts_min, int64_t ts_max,
int flags, int64_t *ts_ret,
int64_t (*read_timestamp)(struct AVFormatContext *, int , int64_t *, int64_t ));
/**
* Set the time base and wrapping info for a given stream. This will be used
* to interpret the stream's timestamps. If the new time base is invalid
* (numerator or denominator are non-positive), it leaves the stream
* unchanged.
*
* @param st stream
* @param pts_wrap_bits number of bits effectively used by the pts
* (used for wrap control)
* @param pts_num time base numerator
* @param pts_den time base denominator
*/
void avpriv_set_pts_info(AVStream *st, int pts_wrap_bits,
unsigned int pts_num, unsigned int pts_den);
/**
* Add side data to a packet for changing parameters to the given values.
* Parameters set to 0 aren't included in the change.
*/
int ff_add_param_change(AVPacket *pkt, int32_t channels,
uint64_t channel_layout, int32_t sample_rate,
int32_t width, int32_t height);
/**
* Set the timebase for each stream from the corresponding codec timebase and
* print it.
*/
int ff_framehash_write_header(AVFormatContext *s);
/**
* Read a transport packet from a media file.
*
* @param s media file handle
* @param pkt is filled
* @return 0 if OK, AVERROR_xxx on error
*/
int ff_read_packet(AVFormatContext *s, AVPacket *pkt);
/**
* Add an attached pic to an AVStream.
*
* @param st if set, the stream to add the attached pic to;
* if unset, a new stream will be added to s.
* @param pb AVIOContext to read data from if buf is unset.
* @param buf if set, it contains the data and size information to be used
* for the attached pic; if unset, data is read from pb.
* @param size the size of the data to read if buf is unset.
*
* @return 0 on success, < 0 on error. On error, this function removes
* the stream it has added (if any).
*/
int ff_add_attached_pic(AVFormatContext *s, AVStream *st, AVIOContext *pb,
AVBufferRef **buf, int size);
/**
* Interleave an AVPacket per dts so it can be muxed.
* See the documentation of AVOutputFormat.interleave_packet for details.
*/
int ff_interleave_packet_per_dts(AVFormatContext *s, AVPacket *pkt,
int flush, int has_packet);
/**
* Interleave packets directly in the order in which they arrive
* without any sort of buffering.
*/
int ff_interleave_packet_passthrough(AVFormatContext *s, AVPacket *pkt,
int flush, int has_packet);
void ff_free_stream(AVFormatContext *s, AVStream *st);
unsigned int ff_codec_get_tag(const AVCodecTag *tags, enum AVCodecID id);
enum AVCodecID ff_codec_get_id(const AVCodecTag *tags, unsigned int tag);
int ff_is_intra_only(enum AVCodecID id);
/**
* Select a PCM codec based on the given parameters.
*
* @param bps bits-per-sample
* @param flt floating-point
* @param be big-endian
* @param sflags signed flags. each bit corresponds to one byte of bit depth.
* e.g. the 1st bit indicates if 8-bit should be signed or
* unsigned, the 2nd bit indicates if 16-bit should be signed or
* unsigned, etc... This is useful for formats such as WAVE where
* only 8-bit is unsigned and all other bit depths are signed.
* @return a PCM codec id or AV_CODEC_ID_NONE
*/
enum AVCodecID ff_get_pcm_codec_id(int bps, int flt, int be, int sflags);
/**
* Chooses a timebase for muxing the specified stream.
*
* The chosen timebase allows sample accurate timestamps based
* on the framerate or sample rate for audio streams. It also is
* at least as precise as 1/min_precision would be.
*/
AVRational ff_choose_timebase(AVFormatContext *s, AVStream *st, int min_precision);
/**
* Chooses a timebase for muxing the specified stream.
*/
enum AVChromaLocation ff_choose_chroma_location(AVFormatContext *s, AVStream *st);
/**
* Generate standard extradata for AVC-Intra based on width/height and field
* order.
*/
int ff_generate_avci_extradata(AVStream *st);
/**
* Add a bitstream filter to a stream.
*
* @param st output stream to add a filter to
* @param name the name of the filter to add
* @param args filter-specific argument string
* @return >0 on success;
* AVERROR code on failure
*/
int ff_stream_add_bitstream_filter(AVStream *st, const char *name, const char *args);
/**
* Copy encoding parameters from source to destination stream
*
* @param dst pointer to destination AVStream
* @param src pointer to source AVStream
* @return >=0 on success, AVERROR code on error
*/
int ff_stream_encode_params_copy(AVStream *dst, const AVStream *src);
/**
* Copy side data from source to destination stream
*
* @param dst pointer to destination AVStream
* @param src pointer to source AVStream
* @return >=0 on success, AVERROR code on error
*/
int ff_stream_side_data_copy(AVStream *dst, const AVStream *src);
/**
* Wrap ffurl_move() and log if error happens.
*
* @param url_src source path
* @param url_dst destination path
* @return 0 or AVERROR on failure
*/
int ff_rename(const char *url_src, const char *url_dst, void *logctx);
/**
Merge commit '059a934806d61f7af9ab3fd9f74994b838ea5eba' * commit '059a934806d61f7af9ab3fd9f74994b838ea5eba': lavc: Consistently prefix input buffer defines Conflicts: doc/examples/decoding_encoding.c libavcodec/4xm.c libavcodec/aac_adtstoasc_bsf.c libavcodec/aacdec.c libavcodec/aacenc.c libavcodec/ac3dec.h libavcodec/asvenc.c libavcodec/avcodec.h libavcodec/avpacket.c libavcodec/dvdec.c libavcodec/ffv1enc.c libavcodec/g2meet.c libavcodec/gif.c libavcodec/h264.c libavcodec/h264_mp4toannexb_bsf.c libavcodec/huffyuvdec.c libavcodec/huffyuvenc.c libavcodec/jpeglsenc.c libavcodec/libxvid.c libavcodec/mdec.c libavcodec/motionpixels.c libavcodec/mpeg4videodec.c libavcodec/mpegvideo.c libavcodec/noise_bsf.c libavcodec/nuv.c libavcodec/nvenc.c libavcodec/options.c libavcodec/parser.c libavcodec/pngenc.c libavcodec/proresenc_kostya.c libavcodec/qsvdec.c libavcodec/svq1enc.c libavcodec/tiffenc.c libavcodec/truemotion2.c libavcodec/utils.c libavcodec/utvideoenc.c libavcodec/vc1dec.c libavcodec/wmalosslessdec.c libavformat/adxdec.c libavformat/aiffdec.c libavformat/apc.c libavformat/apetag.c libavformat/avidec.c libavformat/bink.c libavformat/cafdec.c libavformat/flvdec.c libavformat/id3v2.c libavformat/isom.c libavformat/matroskadec.c libavformat/mov.c libavformat/mpc.c libavformat/mpc8.c libavformat/mpegts.c libavformat/mvi.c libavformat/mxfdec.c libavformat/mxg.c libavformat/nutdec.c libavformat/oggdec.c libavformat/oggparsecelt.c libavformat/oggparseflac.c libavformat/oggparseopus.c libavformat/oggparsespeex.c libavformat/omadec.c libavformat/rawdec.c libavformat/riffdec.c libavformat/rl2.c libavformat/rmdec.c libavformat/rtpdec_latm.c libavformat/rtpdec_mpeg4.c libavformat/rtpdec_qdm2.c libavformat/rtpdec_svq3.c libavformat/sierravmd.c libavformat/smacker.c libavformat/smush.c libavformat/spdifenc.c libavformat/takdec.c libavformat/tta.c libavformat/utils.c libavformat/vqf.c libavformat/westwood_vqa.c libavformat/xmv.c libavformat/xwma.c libavformat/yop.c Merged-by: Michael Niedermayer <michael@niedermayer.cc>
2015-07-27 22:53:16 +02:00
* Allocate extradata with additional AV_INPUT_BUFFER_PADDING_SIZE at end
* which is always set to 0.
*
* Previously allocated extradata in par will be freed.
*
* @param size size of extradata
* @return 0 if OK, AVERROR_xxx on error
*/
int ff_alloc_extradata(AVCodecParameters *par, int size);
/**
Merge commit '059a934806d61f7af9ab3fd9f74994b838ea5eba' * commit '059a934806d61f7af9ab3fd9f74994b838ea5eba': lavc: Consistently prefix input buffer defines Conflicts: doc/examples/decoding_encoding.c libavcodec/4xm.c libavcodec/aac_adtstoasc_bsf.c libavcodec/aacdec.c libavcodec/aacenc.c libavcodec/ac3dec.h libavcodec/asvenc.c libavcodec/avcodec.h libavcodec/avpacket.c libavcodec/dvdec.c libavcodec/ffv1enc.c libavcodec/g2meet.c libavcodec/gif.c libavcodec/h264.c libavcodec/h264_mp4toannexb_bsf.c libavcodec/huffyuvdec.c libavcodec/huffyuvenc.c libavcodec/jpeglsenc.c libavcodec/libxvid.c libavcodec/mdec.c libavcodec/motionpixels.c libavcodec/mpeg4videodec.c libavcodec/mpegvideo.c libavcodec/noise_bsf.c libavcodec/nuv.c libavcodec/nvenc.c libavcodec/options.c libavcodec/parser.c libavcodec/pngenc.c libavcodec/proresenc_kostya.c libavcodec/qsvdec.c libavcodec/svq1enc.c libavcodec/tiffenc.c libavcodec/truemotion2.c libavcodec/utils.c libavcodec/utvideoenc.c libavcodec/vc1dec.c libavcodec/wmalosslessdec.c libavformat/adxdec.c libavformat/aiffdec.c libavformat/apc.c libavformat/apetag.c libavformat/avidec.c libavformat/bink.c libavformat/cafdec.c libavformat/flvdec.c libavformat/id3v2.c libavformat/isom.c libavformat/matroskadec.c libavformat/mov.c libavformat/mpc.c libavformat/mpc8.c libavformat/mpegts.c libavformat/mvi.c libavformat/mxfdec.c libavformat/mxg.c libavformat/nutdec.c libavformat/oggdec.c libavformat/oggparsecelt.c libavformat/oggparseflac.c libavformat/oggparseopus.c libavformat/oggparsespeex.c libavformat/omadec.c libavformat/rawdec.c libavformat/riffdec.c libavformat/rl2.c libavformat/rmdec.c libavformat/rtpdec_latm.c libavformat/rtpdec_mpeg4.c libavformat/rtpdec_qdm2.c libavformat/rtpdec_svq3.c libavformat/sierravmd.c libavformat/smacker.c libavformat/smush.c libavformat/spdifenc.c libavformat/takdec.c libavformat/tta.c libavformat/utils.c libavformat/vqf.c libavformat/westwood_vqa.c libavformat/xmv.c libavformat/xwma.c libavformat/yop.c Merged-by: Michael Niedermayer <michael@niedermayer.cc>
2015-07-27 22:53:16 +02:00
* Allocate extradata with additional AV_INPUT_BUFFER_PADDING_SIZE at end
* which is always set to 0 and fill it from pb.
*
* @param size size of extradata
* @return >= 0 if OK, AVERROR_xxx on error
*/
int ff_get_extradata(AVFormatContext *s, AVCodecParameters *par, AVIOContext *pb, int size);
/**
* add frame for rfps calculation.
*
* @param dts timestamp of the i-th frame
* @return 0 if OK, AVERROR_xxx on error
*/
int ff_rfps_add_frame(AVFormatContext *ic, AVStream *st, int64_t dts);
void ff_rfps_calculate(AVFormatContext *ic);
2013-12-31 16:09:48 +03:00
/**
* Flags for AVFormatContext.write_uncoded_frame()
*/
enum AVWriteUncodedFrameFlags {
/**
* Query whether the feature is possible on this stream.
* The frame argument is ignored.
*/
AV_WRITE_UNCODED_FRAME_QUERY = 0x0001,
};
/**
* Copies the whilelists from one context to the other
*/
int ff_copy_whiteblacklists(AVFormatContext *dst, const AVFormatContext *src);
2013-12-31 16:09:48 +03:00
/**
* Returned by demuxers to indicate that data was consumed but discarded
* (ignored streams or junk data). The framework will re-call the demuxer.
*/
#define FFERROR_REDO FFERRTAG('R','E','D','O')
/**
* Utility function to open IO stream of output format.
*
* @param s AVFormatContext
* @param url URL or file name to open for writing
* @options optional options which will be passed to io_open callback
* @return >=0 on success, negative AVERROR in case of failure
*/
int ff_format_output_open(AVFormatContext *s, const char *url, AVDictionary **options);
/*
* A wrapper around AVFormatContext.io_close that should be used
* instead of calling the pointer directly.
*
* @param s AVFormatContext
* @param *pb the AVIOContext to be closed and freed. Can be NULL.
* @return >=0 on success, negative AVERROR in case of failure
*/
int ff_format_io_close(AVFormatContext *s, AVIOContext **pb);
/* Default io_close callback, not to be used directly, use ff_format_io_close
* instead. */
void ff_format_io_close_default(AVFormatContext *s, AVIOContext *pb);
/**
* Utility function to check if the file uses http or https protocol
*
* @param s AVFormatContext
* @param filename URL or file name to open for writing
*/
int ff_is_http_proto(const char *filename);
/**
* Parse creation_time in AVFormatContext metadata if exists and warn if the
* parsing fails.
*
* @param s AVFormatContext
* @param timestamp parsed timestamp in microseconds, only set on successful parsing
* @param return_seconds set this to get the number of seconds in timestamp instead of microseconds
* @return 1 if OK, 0 if the metadata was not present, AVERROR(EINVAL) on parse error
*/
int ff_parse_creation_time_metadata(AVFormatContext *s, int64_t *timestamp, int return_seconds);
/**
* Standardize creation_time metadata in AVFormatContext to an ISO-8601
* timestamp string.
*
* @param s AVFormatContext
* @return <0 on error
*/
int ff_standardize_creation_time(AVFormatContext *s);
#define CONTAINS_PAL 2
/**
* Reshuffles the lines to use the user specified stride.
*
* @param ppkt input and output packet
* @return negative error code or
* 0 if no new packet was allocated
* non-zero if a new packet was allocated and ppkt has to be freed
* CONTAINS_PAL if in addition to a new packet the old contained a palette
*/
int ff_reshuffle_raw_rgb(AVFormatContext *s, AVPacket **ppkt, AVCodecParameters *par, int expected_stride);
/**
* Retrieves the palette from a packet, either from side data, or
* appended to the video data in the packet itself (raw video only).
* It is commonly used after a call to ff_reshuffle_raw_rgb().
*
* Use 0 for the ret parameter to check for side data only.
*
* @param pkt pointer to packet before calling ff_reshuffle_raw_rgb()
* @param ret return value from ff_reshuffle_raw_rgb(), or 0
* @param palette pointer to palette buffer
* @return negative error code or
* 1 if the packet has a palette, else 0
*/
int ff_get_packet_palette(AVFormatContext *s, AVPacket *pkt, int ret, uint32_t *palette);
struct AVBPrint;
/**
* Finalize buf into extradata and set its size appropriately.
*/
int ff_bprint_to_codecpar_extradata(AVCodecParameters *par, struct AVBPrint *buf);
/**
* Find the next packet in the interleaving queue for the given stream.
*
* @return a pointer to a packet if one was found, NULL otherwise.
*/
const AVPacket *ff_interleaved_peek(AVFormatContext *s, int stream);
int ff_get_muxer_ts_offset(AVFormatContext *s, int stream_index, int64_t *offset);
int ff_lock_avformat(void);
int ff_unlock_avformat(void);
/**
* Set AVFormatContext url field to the provided pointer. The pointer must
* point to a valid string. The existing url field is freed if necessary. Also
* set the legacy filename field to the same string which was provided in url.
*/
void ff_format_set_url(AVFormatContext *s, char *url);
void avpriv_register_devices(const AVOutputFormat * const o[], const AVInputFormat * const i[]);
#endif /* AVFORMAT_INTERNAL_H */