AVFormatContext

AVFormatContext is a central structure in FFmpeg's libavformat that represents a media container context. It is used for both demuxing (reading) and muxing (writing) operations and encapsulates information about the container format, the I/O layer, and the streams it contains.

The context references the format in use (input or output), the I/O context (AVIOContext*) used for data input/output, the target URL or file path, and an array of AVStream pointers describing each stream inside the container (such as video, audio, and subtitles). It also stores metadata, duration, bit rate, and various format-specific flags. Additional fields cover stream mappings, disposition, and format-specific data.

Typical lifecycle and usage: allocate a context with avformat_alloc_context. For input, call avformat_open_input to open the file or stream, optionally supply a guessed format, then avformat_find_stream_info to read headers and populate the streams array. For output, allocate the context and use avformat_alloc_output_context2 to create a context for a given format and URL, add streams with avformat_new_stream, then write headers with avformat_write_header, write packets with av_write_frame, and finally write trailer with avformat_write_trailer.

Associated utilities include av_dump_format to print a summary of the container, which is helpful for debugging, and avformat_seek_file to seek within the stream.

Common fields conceptually include the number of streams (nb_streams), the streams[] array, the URL, duration, and metadata. Cleanup involves avformat_close_input for inputs or avformat_free_context for freeing the context when done.