Mp4Extractor

const val FACTORY: ExtractorsFactory!

const val FLAG_EMIT_RAW_SUBTITLE_DATA = 16: Int

const val FLAG_MARK_FIRST_VIDEO_TRACK_WITH_MAIN_ROLE = 8: Int

Flag to mark the first video track encountered as ROLE_FLAG_MAIN and all subsequent video tracks as ROLE_FLAG_ALTERNATE.

const val FLAG_READ_EDITABLE_VIDEO_TRACKS = 64: Int

Flag to extract the editable video tracks.

Either primary video tracks or editable video tracks (but not both) will be extracted based on the flag.

If the flag is set but the editable video tracks are not present, then it fallbacks to extract primary tracks instead.

const val FLAG_READ_MOTION_PHOTO_METADATA = 2: Int

Flag to extract MotionPhotoMetadata from HEIC motion photos following the Google Photos Motion Photo File Format V1.1.

As playback is not supported for motion photos, this flag should only be used for metadata retrieval use cases.

const val FLAG_READ_SEF_DATA = 4: Int

Flag to extract SlowMotionData metadata from Samsung Extension Format (SEF) slow motion videos.

const val FLAG_READ_WITHIN_GOP_SAMPLE_DEPENDENCIES = 32: Int

Flag to extract additional sample dependency information, and mark output buffers with BUFFER_FLAG_NOT_DEPENDED_ON.

This class always marks the samples at the start of each group of picture (GOP) with BUFFER_FLAG_KEY_FRAME. Usually, key frames can be decoded independently, without depending on other samples.

Setting this flag enables elementary stream parsing to identify disposable samples that are not depended on by other samples. Any disposable sample can be safely omitted, and the rest of the track will remain valid.

Supported formats are:

• H.264

const val FLAG_WORKAROUND_IGNORE_EDIT_LISTS = 1: Int

Flag to ignore any edit lists in the stream.

Mp4Extractor()

Mp4Extractor(@Mp4Extractor.Flags flags: Int)

Mp4Extractor(subtitleParserFactory: SubtitleParser.Factory!)

Creates a new extractor for unfragmented MP4 streams.

Mp4Extractor(
    subtitleParserFactory: SubtitleParser.Factory!,
    @Mp4Extractor.Flags flags: Int
)

Creates a new extractor for unfragmented MP4 streams, using the specified flags to control the extractor's behavior.

fun getSampleTimestampsUs(trackId: Int): LongArray<Long>!

Returns the list of sample timestamps of a trackId, in microseconds.

fun getSeekPoints(timeUs: Long): SeekMap.SeekPoints!

Obtains seek points for the specified seek time in microseconds. The returned will contain one or two distinct seek points.

Two seek points [A, B] are returned in the case that seeking can only be performed to discrete points in time, there does not exist a seek point at exactly the requested time, and there exist seek points on both sides of it. In this case A and B are the closest seek points before and after the requested time. A single seek point is returned in all other cases.

fun getSeekPoints(timeUs: Long, trackId: Int): SeekMap.SeekPoints!

Equivalent to getSeekPoints, except it adds the trackId parameter.

fun getSniffFailureDetails(): ImmutableList<SniffFailure!>!

Returns additional details about the last call to sniff. The returned list may be empty if no additional details are available, or the last sniff call returned true.

This only contains details that were discovered before sniff returned false, it is not an exhaustive list of issues which, if resolved, would cause the file to be successfully sniffed.

fun init(output: ExtractorOutput!): Unit

Initializes the extractor with an ExtractorOutput. Called at most once.

fun isSeekable(): Boolean

Returns whether seeking is supported.

java-static fun newFactory(subtitleParserFactory: SubtitleParser.Factory!): ExtractorsFactory!

Creates a factory for Mp4Extractor instances with the provided .

fun read(input: ExtractorInput!, seekPosition: PositionHolder!): Int

Extracts data read from a provided ExtractorInput. Must not be called before init.

A single call to this method will block until some progress has been made, but will not block for longer than this. Hence each call will consume only a small amount of input data.

In the common case, RESULT_CONTINUE is returned to indicate that the passed to the next read is required to provide data continuing from the position in the stream reached by the returning call. If the extractor requires data to be provided from a different position, then that position is set in seekPosition and RESULT_SEEK is returned. If the extractor reached the end of the data provided by the ExtractorInput, then RESULT_END_OF_INPUT is returned.

When this method throws an IOException, extraction may continue by providing an ExtractorInput with an unchanged read position to a subsequent call to this method.

fun release(): Unit

Releases all kept resources.

fun seek(position: Long, timeUs: Long): Unit

Notifies the extractor that a seek has occurred.

Following a call to this method, the ExtractorInput passed to the next invocation of read is required to provide data starting from position in the stream. Valid random access positions are the start of the stream and positions that can be obtained from any SeekMap passed to the ExtractorOutput.

fun sniff(input: ExtractorInput!): Boolean

Returns whether this extractor can extract samples from the ExtractorInput, which must provide data from the start of the stream.

If true is returned, the input's reading position may have been modified. Otherwise, only its peek position may have been modified.

val durationUs: Long