transcode函数调用结构

函数调用结构图：

大致函数说明：

1、avcodec_open2

函数原型：intavcodec_open2(AVCodecContext *avctx, const AVCodec *codec, AVDictionary**options);

函数介绍：

/**

 *Initialize the AVCodecContext to use the given AVCodec. Prior to using this

 *function the context has to be allocated with avcodec_alloc_context3().

 *

 *The functions avcodec_find_decoder_by_name(), avcodec_find_encoder_by_name(),

 *avcodec_find_decoder() and avcodec_find_encoder() provide an easy way for

 *retrieving a codec.

 *

 *@warning This function is not thread safe!

 *

 *@note Always call this function before using decoding routines (such as

 *@ref avcodec_decode_video2()).

 *

 *@code

 *avcodec_register_all();

 *av_dict_set(&opts, "b", "2.5M", 0);

 *codec = avcodec_find_decoder(AV_CODEC_ID_H264);

 * if(!codec)

 *    exit(1);

 *

 *context = avcodec_alloc_context3(codec);

 *

 * if(avcodec_open2(context, codec, opts) < 0)

 *    exit(1);

 *@endcode

 *

 *@param avctx The context to initialize.

 *@param codec The codec to open this context for. If a non-NULL codec has been

 *             previously passed to avcodec_alloc_context3() or

 *             avcodec_get_context_defaults3() for this context, then this

 *             parameter MUST be either NULL or equal to the previously passed

 *              codec.

 *@param options A dictionary filled with AVCodecContext and codec-privateoptions.

 *                On return this object will befilled with options that were not found.

 *

 *@return zero on success, a negative value on error

 *@see avcodec_alloc_context3(), avcodec_find_decoder(), avcodec_find_encoder(),

 *     av_dict_set(), av_opt_find().

 */

说明：为使用给定的AVCodec而初始化AVCodecContext，调用此函数之前AVCodecContext必须调用avcodec_alloc_context3()分配空间。

2、avcodec_copy_context

函数原型：int avcodec_copy_context(AVCodecContext*dest, const AVCodecContext *src);

函数介绍：

/**

 *Copy the settings of the source AVCodecContext into the destination

 *AVCodecContext. The resulting destination codec context will be

 *unopened, i.e. you are required to call avcodec_open2() before you

 *can use this AVCodecContext to decode/encode video/audio data.

 *

 *@param dest target codec context, should be initialized with

 *            avcodec_alloc_context3(NULL), but otherwise uninitialized

 *@param src source codec context

 *@return AVERROR() on error (e.g. memory allocation error), 0 on success

 */

说明：拷贝AVCodecContext，目标AVCodecContext的codec是未被打开的，使用codec之前请调用avcodec_open2()。

3、avformat_write_header

函数原型：int avformat_write_header(AVFormatContext*s, AVDictionary **options);

函数介绍：

/**

 *Allocate the stream private data and write the stream header to

 * anoutput media file.

 *

 * @params Media file handle, must be allocated with avformat_alloc_context().

 *         Its oformat field must be set to the desired output format;

 *         Its pb field must be set to an already opened AVIOContext.

 * @paramoptions  An AVDictionary filled withAVFormatContext and muxer-private options.

 *                 On return this parameter willbe destroyed and replaced with a dict containing

 *                 options that were not found.May be NULL.

 *

 *@return 0 on success, negative AVERROR on failure.

 *

 *@see av_opt_find, av_dict_set, avio_open, av_oformat_next.

 */

说明：为流特有的数据分配空间，并把流的头信息写入output file。

4、av_read_frame

函数原型：int av_read_frame(AVFormatContext *s,AVPacket *pkt)

函数介绍：

/**

 *Return the next frame of a stream.

 *This function returns what is stored in the file, and does not validate

 *that what is there are valid frames for the decoder. It will split what is

 *stored in the file into frames and return one for each call. It will not

 *omit invalid data between valid frames so as to give the decoder the maximum

 *information possible for decoding.

 *

 * Ifpkt->buf is NULL, then the packet is valid until the next

 *av_read_frame() or until avformat_close_input(). Otherwise the packet

 * isvalid indefinitely. In both cases the packet must be freed with

 *av_free_packet when it is no longer needed. For video, the packet contains

 *exactly one frame. For audio, it contains an integer number of frames if each

 *frame has a known fixed size (e.g. PCM or ADPCM data). If the audio frames

 *have a variable size (e.g. MPEG audio), then it contains one frame.

 *

 * pkt->pts,pkt->dts and pkt->duration are always set to correct

 *values in AVStream.time_base units (and guessed if the format cannot

 *provide them). pkt->pts can be AV_NOPTS_VALUE if the videoformat

 *has B-frames, so it is better to rely on pkt->dts if you donot

 * decompressthe payload.

 *

 *@return 0 if OK, < 0 on error or end of file

 */

说明：通过av_read_packet(***)，读取一个包，需要说明的是此函数必须是包含整数帧的，不存在半帧的情况，以ts流为例，是读取一个完整的PES包（一个完整pes包包含若干视频或音频es包），读取完毕后，通过av_parser_parse2(***)分析出视频一帧（或音频若干帧），返回，下次进入循环的时候，如果上次的数据没有完全取完，则st = s->cur_st;不会是NULL，即再此进入av_parser_parse2(***)流程，而不是下面的av_read_packet（**）流程，这样就保证了，如果读取一次包含了N帧视频数据（以视频为例），则调用av_read_frame（***）N次都不会去读数据，而是返回第一次读取的数据，直到全部解析完毕。

5、avcodec_encode_video2

函数原型：intavcodec_encode_video2(AVCodecContext *avctx, AVPacket *avpkt,

                          const AVFrame *frame, int *got_packet_ptr);

 

函数介绍：

/**

 *Encode a frame of video.

 *

 *Takes input raw video data from frame and writes the next output packet, if

 *available, to avpkt. The output packet does not necessarily contain data for

 *the most recent frame, as encoders can delay and reorder input frames

 *internally as needed.

 *

 *@param avctx     codec context

 *@param avpkt     output AVPacket.

 *                  The user can supply an outputbuffer by setting

 *                  avpkt->data andavpkt->size prior to calling the

 *                  function, but if the size ofthe user-provided data is not

 *                  large enough, encoding willfail. All other AVPacket fields

 *                  will be reset by the encoderusing av_init_packet(). If

 *                  avpkt->data is NULL, theencoder will allocate it.

 *                  The encoder will setavpkt->size to the size of the

 *                  output packet. The returneddata (if any) belongs to the

 *                  caller, he is responsible forfreeing it.

 *

 *                  If this function fails orproduces no output, avpkt will be

 *                  freed using av_free_packet()(i.e. avpkt->destruct will be

 *                  called to free the usersupplied buffer).

 *@param[in] frame AVFrame containing the raw video data to be encoded.

 *                  May be NULL when flushing anencoder that has the

 *                  CODEC_CAP_DELAY capabilityset.

 *@param[out] got_packet_ptr This field is set to 1 by libavcodec if the

 *                           output packet isnon-empty, and to 0 if it is

 *                            empty. If thefunction returns an error, the

 *                            packet can beassumed to be invalid, and the

 *                            value of got_packet_ptris undefined and should

 *                            not be used.

 *@return          0 on success, negativeerror code on failure

 */

说明：压缩一帧数据，如果允许，把数据写入avpkt里面。注意output 包不是每次都有最近的帧数据的，因为编码器有可能缓存起来。

6、av_packet_rescale_ts

函数原型：voidav_packet_rescale_ts(AVPacket *pkt, AVRational tb_src, AVRational tb_dst);

函数介绍：

/**

 *Convert valid timing fields (timestamps / durations) in a packet from one

 *timebase to another. Timestamps with unknown values (AV_NOPTS_VALUE) will be

 *ignored.

 *

 *@param pkt packet on which the conversion will be performed

 *@param tb_src source timebase, in which the timing fields in pkt are

 *              expressed

 *@param tb_dst destination timebase, to which the timing fields will be

 *              converted

 */

说明：转换时间戳。

7、av_interleaved_write_frame

函数原型：int av_interleaved_write_frame(AVFormatContext*s, AVPacket *pkt)

函数介绍：

/**

 *Write a packet to an output media file ensuring correct interleaving.

 *

 *This function will buffer the packets internally as needed to make sure the

 *packets in the output file are properly interleaved in the order of

 *increasing dts. Callers doing their own interleaving should call

 *av_write_frame() instead of this function.

 *

 * @params media file handle

 * @parampkt The packet containing the data to be written.

 *           <br>

 *           If the packet is reference-counted, this function will take

 *           ownership of this reference and unreference it later when it sees

 *           fit.

 *           The caller must not access the data through this reference after

 *           this function returns. If the packet is not reference-counted,

 *           libavformat will make a copy.

 *           <br>

 *           This parameter can be NULL (at any time, not just at the end), to

 *           flush the interleaving queues.

 *           <br>

 *           Packet's @ref AVPacket.stream_index "stream_index"field must be

 *           set to the index of the corresponding stream in @ref

 *            AVFormatContext.streams"s->streams". It is very strongly

 *           recommended that timing information (@ref AVPacket.pts "pts",@ref

 *           AVPacket.dts "dts", @ref AVPacket.duration"duration") is set to

 *           correct values.

 *

 *@return 0 on success, a negative AVERROR on error. Libavformat willalways

 *        take care of freeing the packet, even if this function fails.

 *

 *@see av_write_frame(), AVFormatContext.max_interleave_delta

 */

说明：把输出数据交错写入输出文件。

8、av_frame_unref

函数原型：void av_frame_unref(AVFrame *frame);

函数介绍：

/**

 * Unreferenceall the buffers referenced by frame and reset the frame fields.

 */

说明：解除数据绑定。

9、av_frame_ref

函数原型：int av_frame_ref(AVFrame *dst, const AVFrame *src);

函数介绍：

/**

 *Set up a new reference to the data described by the source frame.

 *

 *Copy frame properties from src to dst and create a new referencefor each

 *AVBufferRef from src.

 *

 * Ifsrc is not reference counted, new buffers are allocated and the data is

 *copied.

 *

 *@return 0 on success, a negative AVERROR on error

 */

说明：从src拷贝到dst，并从src建立新的buffers引用。

10、av_init_packet

函数原型：voidav_init_packet(AVPacket *pkt);

函数介绍：

/**

 *Initialize optional fields of a packet with default values.

 *

 *Note, this does not touch the data and size members, which have to be

 *initialized separately.

 *

 *@param pkt packet

 */

说明：用缺省值初始化pkt。并不会为data申请空间，和为size设置有效值，这些要单独设置值。

11、av_write_trailer

函数原型：int av_write_trailer(AVFormatContext *s);

函数介绍：

/**

 *Write the stream trailer to an output media file and free the

 *file private data.

 *

 *May only be called after a successful call to avformat_write_header.

 *

 * @params media file handle

 *@return 0 if OK, AVERROR_xxx on error

 */

说明：循环调用interleave_packet()以及write_packet()，将还未输出的AVPacket输出出来。调用AVOutputFormat的write_trailer()，输出文件尾。最后做一些释放动作。

12、interleave_packet

函数原型：static int interleave_packet(AVFormatContext *s, AVPacket *out, AVPacket *in,int flush);

函数介绍：

/**

 *Interleave an AVPacket correctly so it can be muxed.

 * @paramout the interleaved packet will be output here

 * @paramin the input packet

 * @paramflush 1 if no further packets are available as input and all

 *             remaining packets should be output

 *@return 1 if a packet was output, 0 if no packet could be output,

 *        < 0 if an error occurred

 */

说明：把缓冲中的数据输出出来。

13、write_packet

函数原型：static int write_packet(AVFormatContext *s, AVPacket *pkt)

函数介绍：

/**

 *Make timestamps non negative, move side data from payload tointernal struct, call muxer, and restore

 * sidedata.

 *

 * FIXME: this function should NEVER getundefined pts/dts beside when the

 *AVFMT_NOTIMESTAMPS is set.

 *Those additional safety checks should be dropped once the correct checks

 *are set in the callers.

 */

说明：给时间戳赋值，调用av_packet_split_side_data处理packet,然后调用s->oformat的write_uncoded_frame或者write_packet写packet。

14、avcodec_close

函数原型：int avcodec_close(AVCodecContext *avctx);

函数介绍：

/**

 *Close a given AVCodecContext and free all the data associated with it

 *(but not the AVCodecContext itself).

 *

 *Calling this function on an AVCodecContext that hasn't been opened will free

 *the codec-specific data allocated in avcodec_alloc_context3() /

 *avcodec_get_context_defaults3() with a non-NULL codec. Subsequent calls will

 * donothing.

 */

说明：关闭AVCodecContext并释放与之相关的数据（除了AVCodecContext自己）。