mirror of
https://github.com/shaka-project/shaka-player.git
synced 2026-06-14 15:56:38 +03:00
theodab
4425dca283
The state engine mechanism, designed for the player class, was over-engineered. The structure of the class makes debugging player errors unnecessarily annoying, by obfuscating the code-path the error followed, and in general has created a significant amount of technical debt. This changes the player to use an async-await setup for the top-level operations, laying things out much more cleanly and linearly. --------- Co-authored-by: Álvaro Velad Galván <ladvan91@hotmail.com>
4010 lines
139 KiB
JavaScript
4010 lines
139 KiB
JavaScript
/*! @license
|
|
* Shaka Player
|
|
* Copyright 2016 Google LLC
|
|
* SPDX-License-Identifier: Apache-2.0
|
|
*/
|
|
|
|
|
|
goog.provide('shaka.hls.HlsParser');
|
|
|
|
goog.require('goog.Uri');
|
|
goog.require('goog.asserts');
|
|
goog.require('shaka.abr.Ewma');
|
|
goog.require('shaka.hls.ManifestTextParser');
|
|
goog.require('shaka.hls.Playlist');
|
|
goog.require('shaka.hls.PlaylistType');
|
|
goog.require('shaka.hls.Tag');
|
|
goog.require('shaka.hls.Utils');
|
|
goog.require('shaka.log');
|
|
goog.require('shaka.media.DrmEngine');
|
|
goog.require('shaka.media.InitSegmentReference');
|
|
goog.require('shaka.media.ManifestParser');
|
|
goog.require('shaka.media.PresentationTimeline');
|
|
goog.require('shaka.media.SegmentIndex');
|
|
goog.require('shaka.media.SegmentReference');
|
|
goog.require('shaka.net.DataUriPlugin');
|
|
goog.require('shaka.net.NetworkingEngine');
|
|
goog.require('shaka.util.ArrayUtils');
|
|
goog.require('shaka.util.BufferUtils');
|
|
goog.require('shaka.util.Error');
|
|
goog.require('shaka.util.FakeEvent');
|
|
goog.require('shaka.util.Functional');
|
|
goog.require('shaka.util.LanguageUtils');
|
|
goog.require('shaka.util.ManifestParserUtils');
|
|
goog.require('shaka.util.MimeUtils');
|
|
goog.require('shaka.util.OperationManager');
|
|
goog.require('shaka.util.Pssh');
|
|
goog.require('shaka.media.SegmentUtils');
|
|
goog.require('shaka.util.Timer');
|
|
goog.require('shaka.util.Platform');
|
|
goog.require('shaka.util.Uint8ArrayUtils');
|
|
goog.require('shaka.util.XmlUtils');
|
|
goog.requireType('shaka.hls.Segment');
|
|
|
|
|
|
/**
|
|
* HLS parser.
|
|
*
|
|
* @implements {shaka.extern.ManifestParser}
|
|
* @export
|
|
*/
|
|
shaka.hls.HlsParser = class {
|
|
/**
|
|
* Creates an Hls Parser object.
|
|
*/
|
|
constructor() {
|
|
/** @private {?shaka.extern.ManifestParser.PlayerInterface} */
|
|
this.playerInterface_ = null;
|
|
|
|
/** @private {?shaka.extern.ManifestConfiguration} */
|
|
this.config_ = null;
|
|
|
|
/** @private {number} */
|
|
this.globalId_ = 1;
|
|
|
|
/** @private {!Map.<string, string>} */
|
|
this.globalVariables_ = new Map();
|
|
|
|
/**
|
|
* A map from group id to stream infos created from the media tags.
|
|
* @private {!Map.<string, !Array.<?shaka.hls.HlsParser.StreamInfo>>}
|
|
*/
|
|
this.groupIdToStreamInfosMap_ = new Map();
|
|
|
|
/**
|
|
* For media playlist lazy-loading to work in livestreams, we have to assume
|
|
* that each stream of a type (video, audio, etc) has the same mappings of
|
|
* sequence number to start time.
|
|
* This map stores those relationships.
|
|
* Only used during livestreams; we do not assume that VOD content is
|
|
* aligned in that way.
|
|
* @private {!Map.<string, !Map.<number, number>>}
|
|
*/
|
|
this.mediaSequenceToStartTimeByType_ = new Map();
|
|
|
|
// Set initial maps.
|
|
const ContentType = shaka.util.ManifestParserUtils.ContentType;
|
|
this.mediaSequenceToStartTimeByType_.set(ContentType.VIDEO, new Map());
|
|
this.mediaSequenceToStartTimeByType_.set(ContentType.AUDIO, new Map());
|
|
this.mediaSequenceToStartTimeByType_.set(ContentType.TEXT, new Map());
|
|
this.mediaSequenceToStartTimeByType_.set(ContentType.IMAGE, new Map());
|
|
|
|
/**
|
|
* The values are strings of the form "<VIDEO URI> - <AUDIO URI>",
|
|
* where the URIs are the verbatim media playlist URIs as they appeared in
|
|
* the master playlist.
|
|
*
|
|
* Used to avoid duplicates that vary only in their text stream.
|
|
*
|
|
* @private {!Set.<string>}
|
|
*/
|
|
this.variantUriSet_ = new Set();
|
|
|
|
/**
|
|
* A map from (verbatim) media playlist URI to stream infos representing the
|
|
* playlists.
|
|
*
|
|
* On update, used to iterate through and update from media playlists.
|
|
*
|
|
* On initial parse, used to iterate through and determine minimum
|
|
* timestamps, offsets, and to handle TS rollover.
|
|
*
|
|
* During parsing, used to avoid duplicates in the async methods
|
|
* createStreamInfoFromMediaTag_, createStreamInfoFromImageTag_ and
|
|
* createStreamInfoFromVariantTag_.
|
|
*
|
|
* @private {!Map.<string, shaka.hls.HlsParser.StreamInfo>}
|
|
*/
|
|
this.uriToStreamInfosMap_ = new Map();
|
|
|
|
/** @private {?shaka.media.PresentationTimeline} */
|
|
this.presentationTimeline_ = null;
|
|
|
|
/**
|
|
* The master playlist URI, after redirects.
|
|
*
|
|
* @private {string}
|
|
*/
|
|
this.masterPlaylistUri_ = '';
|
|
|
|
/** @private {shaka.hls.ManifestTextParser} */
|
|
this.manifestTextParser_ = new shaka.hls.ManifestTextParser();
|
|
|
|
/**
|
|
* The minimum sequence number for generated segments, when ignoring
|
|
* EXT-X-PROGRAM-DATE-TIME.
|
|
*
|
|
* @private {number}
|
|
*/
|
|
this.minSequenceNumber_ = -1;
|
|
|
|
/**
|
|
* The lowest time value for any of the streams, as defined by the
|
|
* EXT-X-PROGRAM-DATE-TIME value. Measured in seconds since January 1, 1970.
|
|
*
|
|
* @private {number}
|
|
*/
|
|
this.lowestSyncTime_ = Infinity;
|
|
|
|
/**
|
|
* Whether the streams have previously been "finalized"; that is to say,
|
|
* whether we have loaded enough streams to know information about the asset
|
|
* such as timing information, live status, etc.
|
|
*
|
|
* @private {boolean}
|
|
*/
|
|
this.streamsFinalized_ = false;
|
|
|
|
/**
|
|
* Whether the manifest informs about the codec to use.
|
|
*
|
|
* @private
|
|
*/
|
|
this.codecInfoInManifest_ = false;
|
|
|
|
/**
|
|
* This timer is used to trigger the start of a manifest update. A manifest
|
|
* update is async. Once the update is finished, the timer will be restarted
|
|
* to trigger the next update. The timer will only be started if the content
|
|
* is live content.
|
|
*
|
|
* @private {shaka.util.Timer}
|
|
*/
|
|
this.updatePlaylistTimer_ = new shaka.util.Timer(() => {
|
|
this.onUpdate_();
|
|
});
|
|
|
|
/** @private {shaka.hls.HlsParser.PresentationType_} */
|
|
this.presentationType_ = shaka.hls.HlsParser.PresentationType_.VOD;
|
|
|
|
/** @private {?shaka.extern.Manifest} */
|
|
this.manifest_ = null;
|
|
|
|
/** @private {number} */
|
|
this.maxTargetDuration_ = 0;
|
|
|
|
/** @private {number} */
|
|
this.lastTargetDuration_ = Infinity;
|
|
|
|
/** Partial segments target duration.
|
|
* @private {number}
|
|
*/
|
|
this.partialTargetDuration_ = 0;
|
|
|
|
/** @private {number} */
|
|
this.presentationDelay_ = 0;
|
|
|
|
/** @private {number} */
|
|
this.lowLatencyPresentationDelay_ = 0;
|
|
|
|
/** @private {shaka.util.OperationManager} */
|
|
this.operationManager_ = new shaka.util.OperationManager();
|
|
|
|
/** A map from closed captions' group id, to a map of closed captions info.
|
|
* {group id -> {closed captions channel id -> language}}
|
|
* @private {Map.<string, Map.<string, string>>}
|
|
*/
|
|
this.groupIdToClosedCaptionsMap_ = new Map();
|
|
|
|
/** @private {Map.<string, string>} */
|
|
this.groupIdToCodecsMap_ = new Map();
|
|
|
|
/** A cache mapping EXT-X-MAP tag info to the InitSegmentReference created
|
|
* from the tag.
|
|
* The key is a string combining the EXT-X-MAP tag's absolute uri, and
|
|
* its BYTERANGE if available.
|
|
* {!Map.<string, !shaka.media.InitSegmentReference>} */
|
|
this.mapTagToInitSegmentRefMap_ = new Map();
|
|
|
|
/** @private {boolean} */
|
|
this.lowLatencyMode_ = false;
|
|
|
|
/** @private {boolean} */
|
|
this.lowLatencyByterangeOptimization_ = false;
|
|
|
|
/**
|
|
* An ewma that tracks how long updates take.
|
|
* This is to mitigate issues caused by slow parsing on embedded devices.
|
|
* @private {!shaka.abr.Ewma}
|
|
*/
|
|
this.averageUpdateDuration_ = new shaka.abr.Ewma(5);
|
|
}
|
|
|
|
|
|
/**
|
|
* @override
|
|
* @exportInterface
|
|
*/
|
|
configure(config) {
|
|
this.config_ = config;
|
|
}
|
|
|
|
/**
|
|
* @override
|
|
* @exportInterface
|
|
*/
|
|
async start(uri, playerInterface) {
|
|
goog.asserts.assert(this.config_, 'Must call configure() before start()!');
|
|
this.playerInterface_ = playerInterface;
|
|
this.lowLatencyMode_ = playerInterface.isLowLatencyMode();
|
|
|
|
const response = await this.requestManifest_([uri]);
|
|
|
|
// Record the master playlist URI after redirects.
|
|
this.masterPlaylistUri_ = response.uri;
|
|
|
|
goog.asserts.assert(response.data, 'Response data should be non-null!');
|
|
await this.parseManifest_(response.data, uri);
|
|
|
|
goog.asserts.assert(this.manifest_, 'Manifest should be non-null');
|
|
return this.manifest_;
|
|
}
|
|
|
|
/**
|
|
* @override
|
|
* @exportInterface
|
|
*/
|
|
stop() {
|
|
// Make sure we don't update the manifest again. Even if the timer is not
|
|
// running, this is safe to call.
|
|
if (this.updatePlaylistTimer_) {
|
|
this.updatePlaylistTimer_.stop();
|
|
this.updatePlaylistTimer_ = null;
|
|
}
|
|
|
|
/** @type {!Array.<!Promise>} */
|
|
const pending = [];
|
|
|
|
if (this.operationManager_) {
|
|
pending.push(this.operationManager_.destroy());
|
|
this.operationManager_ = null;
|
|
}
|
|
|
|
this.playerInterface_ = null;
|
|
this.config_ = null;
|
|
this.variantUriSet_.clear();
|
|
this.manifest_ = null;
|
|
this.uriToStreamInfosMap_.clear();
|
|
this.groupIdToStreamInfosMap_.clear();
|
|
this.groupIdToCodecsMap_.clear();
|
|
this.globalVariables_.clear();
|
|
|
|
return Promise.all(pending);
|
|
}
|
|
|
|
/**
|
|
* @override
|
|
* @exportInterface
|
|
*/
|
|
async update() {
|
|
if (!this.isLive_()) {
|
|
return;
|
|
}
|
|
|
|
/** @type {!Array.<!Promise>} */
|
|
const updates = [];
|
|
const streamInfos = Array.from(this.uriToStreamInfosMap_.values());
|
|
|
|
// This is necessary to calculate correctly the update time.
|
|
this.lastTargetDuration_ = Infinity;
|
|
|
|
// Only update active streams.
|
|
const activeStreamInfos = streamInfos.filter((s) => s.stream.segmentIndex);
|
|
for (const streamInfo of activeStreamInfos) {
|
|
updates.push(this.updateStream_(streamInfo));
|
|
}
|
|
await Promise.all(updates);
|
|
|
|
// Now that streams have been updated, notify the presentation timeline.
|
|
this.notifySegmentsForStreams_(activeStreamInfos.map((s) => s.stream));
|
|
|
|
// If any hasEndList is false, the stream is still live.
|
|
const stillLive = activeStreamInfos.some((s) => s.hasEndList == false);
|
|
if (activeStreamInfos.length && !stillLive) {
|
|
// Convert the presentation to VOD and set the duration.
|
|
const PresentationType = shaka.hls.HlsParser.PresentationType_;
|
|
this.setPresentationType_(PresentationType.VOD);
|
|
|
|
// The duration is the minimum of the end times of all active streams.
|
|
// Non-active streams are not guaranteed to have useful maxTimestamp
|
|
// values, due to the lazy-loading system, so they are ignored.
|
|
const maxTimestamps = activeStreamInfos.map((s) => s.maxTimestamp);
|
|
// The duration is the minimum of the end times of all streams.
|
|
this.presentationTimeline_.setDuration(Math.min(...maxTimestamps));
|
|
this.playerInterface_.updateDuration();
|
|
}
|
|
if (stillLive) {
|
|
this.determineDuration_();
|
|
}
|
|
}
|
|
|
|
/**
|
|
* @param {!shaka.hls.HlsParser.StreamInfo} streamInfo
|
|
* @return {!Map.<number, number>}
|
|
* @private
|
|
*/
|
|
getMediaSequenceToStartTimeFor_(streamInfo) {
|
|
if (this.isLive_()) {
|
|
return this.mediaSequenceToStartTimeByType_.get(streamInfo.type);
|
|
} else {
|
|
return streamInfo.mediaSequenceToStartTime;
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Updates a stream.
|
|
*
|
|
* @param {!shaka.hls.HlsParser.StreamInfo} streamInfo
|
|
* @return {!Promise}
|
|
* @private
|
|
*/
|
|
async updateStream_(streamInfo) {
|
|
const manifestUris = [];
|
|
for (const uri of streamInfo.absoluteMediaPlaylistUris) {
|
|
const uriObj = new goog.Uri(uri);
|
|
const queryData = uriObj.getQueryData();
|
|
if (streamInfo.canBlockReload) {
|
|
if (streamInfo.nextMediaSequence >= 0) {
|
|
// Indicates that the server must hold the request until a Playlist
|
|
// contains a Media Segment with Media Sequence
|
|
queryData.add('_HLS_msn', String(streamInfo.nextMediaSequence));
|
|
}
|
|
if (streamInfo.nextPart >= 0) {
|
|
// Indicates, in combination with _HLS_msn, that the server must hold
|
|
// the request until a Playlist contains Partial Segment N of Media
|
|
// Sequence Number M or later.
|
|
queryData.add('_HLS_part', String(streamInfo.nextPart));
|
|
}
|
|
}
|
|
if (streamInfo.canSkipSegments) {
|
|
// Enable delta updates. This will replace older segments with
|
|
// 'EXT-X-SKIP' tag in the media playlist.
|
|
queryData.add('_HLS_skip', 'YES');
|
|
}
|
|
if (queryData.getCount()) {
|
|
uriObj.setQueryData(queryData);
|
|
}
|
|
manifestUris.push(uriObj.toString());
|
|
}
|
|
const response =
|
|
await this.requestManifest_(manifestUris, /* isPlaylist= */ true);
|
|
if (!streamInfo.stream.segmentIndex) {
|
|
// The stream was closed since the update was first requested.
|
|
return;
|
|
}
|
|
|
|
/** @type {shaka.hls.Playlist} */
|
|
const playlist = this.manifestTextParser_.parsePlaylist(
|
|
response.data, response.uri);
|
|
|
|
if (playlist.type != shaka.hls.PlaylistType.MEDIA) {
|
|
throw new shaka.util.Error(
|
|
shaka.util.Error.Severity.CRITICAL,
|
|
shaka.util.Error.Category.MANIFEST,
|
|
shaka.util.Error.Code.HLS_INVALID_PLAYLIST_HIERARCHY);
|
|
}
|
|
|
|
/** @type {!Array.<!shaka.hls.Tag>} */
|
|
const variablesTags = shaka.hls.Utils.filterTagsByName(playlist.tags,
|
|
'EXT-X-DEFINE');
|
|
|
|
const mediaVariables = this.parseMediaVariables_(
|
|
variablesTags, response.uri);
|
|
|
|
const stream = streamInfo.stream;
|
|
|
|
const mediaSequenceToStartTime =
|
|
this.getMediaSequenceToStartTimeFor_(streamInfo);
|
|
const {keyIds, drmInfos} = this.parseDrmInfo_(playlist, stream.mimeType);
|
|
|
|
const keysAreEqual =
|
|
(a, b) => a.size === b.size && [...a].every((value) => b.has(value));
|
|
|
|
if (!keysAreEqual(stream.keyIds, keyIds)) {
|
|
stream.keyIds = keyIds;
|
|
stream.drmInfos = drmInfos;
|
|
this.playerInterface_.newDrmInfo(stream);
|
|
}
|
|
|
|
const {segments, bandwidth} = this.createSegments_(
|
|
playlist, stream, mediaSequenceToStartTime, mediaVariables);
|
|
|
|
stream.bandwidth = bandwidth;
|
|
|
|
stream.segmentIndex.mergeAndEvict(
|
|
segments, this.presentationTimeline_.getSegmentAvailabilityStart());
|
|
if (segments.length) {
|
|
const mediaSequenceNumber = shaka.hls.Utils.getFirstTagWithNameAsNumber(
|
|
playlist.tags, 'EXT-X-MEDIA-SEQUENCE', 0);
|
|
const skipTag = shaka.hls.Utils.getFirstTagWithName(
|
|
playlist.tags, 'EXT-X-SKIP');
|
|
const skippedSegments =
|
|
skipTag ? Number(skipTag.getAttributeValue('SKIPPED-SEGMENTS')) : 0;
|
|
const {nextMediaSequence, nextPart} =
|
|
this.getNextMediaSequenceAndPart_(mediaSequenceNumber, segments);
|
|
streamInfo.nextMediaSequence = nextMediaSequence + skippedSegments;
|
|
streamInfo.nextPart = nextPart;
|
|
const playlistStartTime = mediaSequenceToStartTime.get(
|
|
mediaSequenceNumber);
|
|
stream.segmentIndex.evict(playlistStartTime);
|
|
}
|
|
const oldSegment = segments[0];
|
|
goog.asserts.assert(oldSegment, 'Should have segments!');
|
|
|
|
streamInfo.minTimestamp = oldSegment.startTime;
|
|
|
|
const newestSegment = segments[segments.length - 1];
|
|
goog.asserts.assert(newestSegment, 'Should have segments!');
|
|
|
|
streamInfo.maxTimestamp = newestSegment.endTime;
|
|
|
|
// Once the last segment has been added to the playlist,
|
|
// #EXT-X-ENDLIST tag will be appended.
|
|
// If that happened, treat the rest of the EVENT presentation as VOD.
|
|
const endListTag =
|
|
shaka.hls.Utils.getFirstTagWithName(playlist.tags, 'EXT-X-ENDLIST');
|
|
|
|
if (endListTag) {
|
|
// Flag this for later. We don't convert the whole presentation into VOD
|
|
// until we've seen the ENDLIST tag for all active playlists.
|
|
streamInfo.hasEndList = true;
|
|
}
|
|
|
|
this.determineLastTargetDuration_(playlist);
|
|
}
|
|
|
|
|
|
/**
|
|
* @override
|
|
* @exportInterface
|
|
*/
|
|
onExpirationUpdated(sessionId, expiration) {
|
|
// No-op
|
|
}
|
|
|
|
/**
|
|
* @override
|
|
* @exportInterface
|
|
*/
|
|
onInitialVariantChosen(variant) {
|
|
// No-op
|
|
}
|
|
|
|
/**
|
|
* Align the streams by sequence number by dropping early segments. Then
|
|
* offset the streams to begin at presentation time 0.
|
|
* @param {!Array.<!shaka.hls.HlsParser.StreamInfo>} streamInfos
|
|
* @private
|
|
*/
|
|
syncStreamsWithSequenceNumber_(streamInfos) {
|
|
// We assume that, when this is first called, we have enough info to
|
|
// determine how to use the program date times (e.g. we have both a video
|
|
// and an audio, and all other videos and audios match those).
|
|
// Thus, we only need to calculate this once.
|
|
const updateMinSequenceNumber = this.minSequenceNumber_ == -1;
|
|
// Sync using media sequence number. Find the highest starting sequence
|
|
// number among all streams. Later, we will drop any references to
|
|
// earlier segments in other streams, then offset everything back to 0.
|
|
for (const streamInfo of streamInfos) {
|
|
const segmentIndex = streamInfo.stream.segmentIndex;
|
|
goog.asserts.assert(segmentIndex,
|
|
'Only loaded streams should be synced');
|
|
const mediaSequenceToStartTime =
|
|
this.getMediaSequenceToStartTimeFor_(streamInfo);
|
|
const segment0 = segmentIndex.earliestReference();
|
|
if (segment0) {
|
|
// This looks inefficient, but iteration order is insertion order.
|
|
// So the very first entry should be the one we want.
|
|
// We assert that this holds true so that we are alerted by debug
|
|
// builds and tests if it changes. We still do a loop, though, so
|
|
// that the code functions correctly in production no matter what.
|
|
if (goog.DEBUG) {
|
|
const firstSequenceStartTime =
|
|
mediaSequenceToStartTime.values().next().value;
|
|
goog.asserts.assert(
|
|
firstSequenceStartTime == segment0.startTime,
|
|
'Sequence number map is not ordered as expected!');
|
|
}
|
|
for (const [sequence, start] of mediaSequenceToStartTime) {
|
|
if (start == segment0.startTime) {
|
|
if (updateMinSequenceNumber) {
|
|
this.minSequenceNumber_ = Math.max(
|
|
this.minSequenceNumber_, sequence);
|
|
}
|
|
// Even if we already have decided on a value for
|
|
// |this.minSequenceNumber_|, we still need to determine the first
|
|
// sequence number for the stream, to offset it in the code below.
|
|
streamInfo.firstSequenceNumber = sequence;
|
|
break;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
if (this.minSequenceNumber_ < 0) {
|
|
// Nothing to sync.
|
|
return;
|
|
}
|
|
|
|
shaka.log.debug('Syncing HLS streams against base sequence number:',
|
|
this.minSequenceNumber_);
|
|
|
|
for (const streamInfo of streamInfos) {
|
|
const segmentIndex = streamInfo.stream.segmentIndex;
|
|
if (segmentIndex) {
|
|
// Drop any earlier references.
|
|
const numSegmentsToDrop =
|
|
this.minSequenceNumber_ - streamInfo.firstSequenceNumber;
|
|
segmentIndex.dropFirstReferences(numSegmentsToDrop);
|
|
|
|
// Now adjust timestamps back to begin at 0.
|
|
const segmentN = segmentIndex.earliestReference();
|
|
if (segmentN) {
|
|
const streamOffset = -segmentN.startTime;
|
|
// Modify all SegmentReferences equally.
|
|
streamInfo.stream.segmentIndex.offset(streamOffset);
|
|
// Update other parts of streamInfo the same way.
|
|
this.offsetStreamInfo_(streamInfo, streamOffset);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Synchronize streams by the EXT-X-PROGRAM-DATE-TIME tags attached to their
|
|
* segments. Also normalizes segment times so that the earliest segment in
|
|
* any stream is at time 0.
|
|
* @param {!Array.<!shaka.hls.HlsParser.StreamInfo>} streamInfos
|
|
* @private
|
|
*/
|
|
syncStreamsWithProgramDateTime_(streamInfos) {
|
|
// We assume that, when this is first called, we have enough info to
|
|
// determine how to use the program date times (e.g. we have both a video
|
|
// and an audio, and all other videos and audios match those).
|
|
// Thus, we only need to calculate this once.
|
|
if (this.lowestSyncTime_ == Infinity) {
|
|
for (const streamInfo of streamInfos) {
|
|
const segmentIndex = streamInfo.stream.segmentIndex;
|
|
goog.asserts.assert(segmentIndex,
|
|
'Only loaded streams should be synced');
|
|
const segment0 = segmentIndex.earliestReference();
|
|
if (segment0 != null && segment0.syncTime != null) {
|
|
this.lowestSyncTime_ =
|
|
Math.min(this.lowestSyncTime_, segment0.syncTime);
|
|
}
|
|
}
|
|
}
|
|
|
|
const lowestSyncTime = this.lowestSyncTime_;
|
|
if (lowestSyncTime == Infinity) {
|
|
// Nothing to sync.
|
|
return;
|
|
}
|
|
|
|
shaka.log.debug('Syncing HLS streams against base time:', lowestSyncTime);
|
|
|
|
for (const streamInfo of this.uriToStreamInfosMap_.values()) {
|
|
const segmentIndex = streamInfo.stream.segmentIndex;
|
|
if (segmentIndex != null) {
|
|
// A segment's startTime should be based on its syncTime vs the lowest
|
|
// syncTime across all streams. The earliest segment sync time from
|
|
// any stream will become presentation time 0. If two streams start
|
|
// e.g. 6 seconds apart in syncTime, then their first segments will
|
|
// also start 6 seconds apart in presentation time.
|
|
|
|
const segment0 = segmentIndex.earliestReference();
|
|
if (segment0.syncTime == null) {
|
|
shaka.log.alwaysError('Missing EXT-X-PROGRAM-DATE-TIME for stream',
|
|
streamInfo.verbatimMediaPlaylistUri,
|
|
'Expect AV sync issues!');
|
|
} else {
|
|
// Stream metadata are offset by a fixed amount based on the
|
|
// first segment.
|
|
const segment0TargetTime = segment0.syncTime - lowestSyncTime;
|
|
const streamOffset = segment0TargetTime - segment0.startTime;
|
|
this.offsetStreamInfo_(streamInfo, streamOffset);
|
|
|
|
// This is computed across all segments separately to manage
|
|
// accumulated drift in durations.
|
|
for (const segment of segmentIndex) {
|
|
segment.syncAgainst(lowestSyncTime);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
/**
|
|
* @param {!shaka.hls.HlsParser.StreamInfo} streamInfo
|
|
* @param {number} offset
|
|
* @private
|
|
*/
|
|
offsetStreamInfo_(streamInfo, offset) {
|
|
// Adjust our accounting of the minimum timestamp.
|
|
streamInfo.minTimestamp += offset;
|
|
|
|
// Adjust our accounting of the maximum timestamp.
|
|
streamInfo.maxTimestamp += offset;
|
|
goog.asserts.assert(streamInfo.maxTimestamp >= 0,
|
|
'Negative maxTimestamp after adjustment!');
|
|
|
|
// Update our map from sequence number to start time.
|
|
const mediaSequenceToStartTime =
|
|
this.getMediaSequenceToStartTimeFor_(streamInfo);
|
|
for (const [key, value] of mediaSequenceToStartTime) {
|
|
mediaSequenceToStartTime.set(key, value + offset);
|
|
}
|
|
|
|
shaka.log.debug('Offset', offset, 'applied to',
|
|
streamInfo.verbatimMediaPlaylistUri);
|
|
}
|
|
|
|
/**
|
|
* Parses the manifest.
|
|
*
|
|
* @param {BufferSource} data
|
|
* @param {string} uri
|
|
* @return {!Promise}
|
|
* @private
|
|
*/
|
|
async parseManifest_(data, uri) {
|
|
const Utils = shaka.hls.Utils;
|
|
|
|
goog.asserts.assert(this.masterPlaylistUri_,
|
|
'Master playlist URI must be set before calling parseManifest_!');
|
|
|
|
const playlist = this.manifestTextParser_.parsePlaylist(
|
|
data, this.masterPlaylistUri_);
|
|
|
|
/** @type {!Array.<!shaka.hls.Tag>} */
|
|
const variablesTags = Utils.filterTagsByName(playlist.tags, 'EXT-X-DEFINE');
|
|
|
|
/** @type {!Array.<!shaka.extern.Variant>} */
|
|
let variants = [];
|
|
/** @type {!Array.<!shaka.extern.Stream>} */
|
|
let textStreams = [];
|
|
/** @type {!Array.<!shaka.extern.Stream>} */
|
|
let imageStreams = [];
|
|
|
|
// Parsing a media playlist results in a single-variant stream.
|
|
if (playlist.type == shaka.hls.PlaylistType.MEDIA) {
|
|
// Get necessary info for this stream. These are things we would normally
|
|
// find from the master playlist (e.g. from values on EXT-X-MEDIA tags).
|
|
const basicInfo = await this.getMediaPlaylistBasicInfo_(playlist);
|
|
const type = basicInfo.type;
|
|
const mimeType = basicInfo.mimeType;
|
|
const codecs = basicInfo.codecs;
|
|
const languageValue = basicInfo.language;
|
|
const height = basicInfo.height;
|
|
const width = basicInfo.width;
|
|
const channelsCount = basicInfo.channelCount;
|
|
const sampleRate = basicInfo.sampleRate;
|
|
const closedCaptions = basicInfo.closedCaptions;
|
|
|
|
// Some values we cannot figure out, and aren't important enough to ask
|
|
// the user to provide through config values. A lot of these are only
|
|
// relevant to ABR, which isn't necessary if there's only one variant.
|
|
// So these unknowns should be set to false or null, largely.
|
|
const spatialAudio = false;
|
|
const characteristics = null;
|
|
const forced = false; // Only relevant for text.
|
|
const primary = true; // This is the only stream!
|
|
const name = 'Media Playlist';
|
|
|
|
// Make the stream info, with those values.
|
|
const streamInfo = await this.convertParsedPlaylistIntoStreamInfo_(
|
|
playlist, uri, uri, codecs, type, languageValue, primary, name,
|
|
channelsCount, closedCaptions, characteristics, forced, sampleRate,
|
|
spatialAudio, mimeType);
|
|
this.uriToStreamInfosMap_.set(uri, streamInfo);
|
|
|
|
if (type == 'video') {
|
|
this.addVideoAttributes_(streamInfo.stream, width, height,
|
|
/* frameRate= */ null, /* videoRange= */ null,
|
|
/* videoLayout= */ null);
|
|
}
|
|
|
|
// Wrap the stream from that stream info with a variant.
|
|
variants.push({
|
|
id: 0,
|
|
language: this.getLanguage_(languageValue),
|
|
disabledUntilTime: 0,
|
|
primary: true,
|
|
audio: type == 'audio' ? streamInfo.stream : null,
|
|
video: type == 'video' ? streamInfo.stream : null,
|
|
bandwidth: streamInfo.stream.bandwidth || 0,
|
|
allowedByApplication: true,
|
|
allowedByKeySystem: true,
|
|
decodingInfos: [],
|
|
});
|
|
} else {
|
|
this.parseMasterVariables_(variablesTags);
|
|
|
|
/** @type {!Array.<!shaka.hls.Tag>} */
|
|
const mediaTags = Utils.filterTagsByName(
|
|
playlist.tags, 'EXT-X-MEDIA');
|
|
/** @type {!Array.<!shaka.hls.Tag>} */
|
|
const variantTags = Utils.filterTagsByName(
|
|
playlist.tags, 'EXT-X-STREAM-INF');
|
|
/** @type {!Array.<!shaka.hls.Tag>} */
|
|
const imageTags = Utils.filterTagsByName(
|
|
playlist.tags, 'EXT-X-IMAGE-STREAM-INF');
|
|
/** @type {!Array.<!shaka.hls.Tag>} */
|
|
const sessionKeyTags = Utils.filterTagsByName(
|
|
playlist.tags, 'EXT-X-SESSION-KEY');
|
|
/** @type {!Array.<!shaka.hls.Tag>} */
|
|
const sessionDataTags = Utils.filterTagsByName(
|
|
playlist.tags, 'EXT-X-SESSION-DATA');
|
|
|
|
this.processSessionData_(sessionDataTags);
|
|
|
|
this.parseCodecs_(variantTags);
|
|
|
|
// Parse audio and video media tags first, so that we can extract segment
|
|
// start time from audio/video streams and reuse for text streams.
|
|
this.createStreamInfosFromMediaTags_(mediaTags);
|
|
this.parseClosedCaptions_(mediaTags);
|
|
variants = this.createVariantsForTags_(variantTags, sessionKeyTags);
|
|
textStreams = this.parseTexts_(mediaTags);
|
|
imageStreams = await this.parseImages_(imageTags);
|
|
}
|
|
|
|
// Make sure that the parser has not been destroyed.
|
|
if (!this.playerInterface_) {
|
|
throw new shaka.util.Error(
|
|
shaka.util.Error.Severity.CRITICAL,
|
|
shaka.util.Error.Category.PLAYER,
|
|
shaka.util.Error.Code.OPERATION_ABORTED);
|
|
}
|
|
|
|
// This assert is our own sanity check.
|
|
goog.asserts.assert(this.presentationTimeline_ == null,
|
|
'Presentation timeline created early!');
|
|
// We don't know if the presentation is VOD or live until we parse at least
|
|
// one media playlist, so make a VOD-style presentation timeline for now
|
|
// and change the type later if we discover this is live.
|
|
// Since the player will load the first variant chosen early in the process,
|
|
// there isn't a window during playback where the live-ness is unknown.
|
|
this.presentationTimeline_ = new shaka.media.PresentationTimeline(
|
|
/* presentationStartTime= */ null, /* delay= */ 0);
|
|
this.presentationTimeline_.setStatic(true);
|
|
|
|
// Single-variant streams aren't lazy-loaded, so for them we already have
|
|
// enough info here to determine the presentation type and duration.
|
|
if (playlist.type == shaka.hls.PlaylistType.MEDIA) {
|
|
if (this.isLive_()) {
|
|
this.changePresentationTimelineToLive_(playlist);
|
|
const delay = this.getUpdatePlaylistDelay_();
|
|
this.updatePlaylistTimer_.tickAfter(/* seconds= */ delay);
|
|
}
|
|
const streamInfos = Array.from(this.uriToStreamInfosMap_.values());
|
|
this.finalizeStreams_(streamInfos);
|
|
this.determineDuration_();
|
|
}
|
|
|
|
this.manifest_ = {
|
|
presentationTimeline: this.presentationTimeline_,
|
|
variants,
|
|
textStreams,
|
|
imageStreams,
|
|
offlineSessionIds: [],
|
|
minBufferTime: 0,
|
|
sequenceMode: this.config_.hls.sequenceMode,
|
|
ignoreManifestTimestampsInSegmentsMode:
|
|
this.config_.hls.ignoreManifestTimestampsInSegmentsMode,
|
|
type: shaka.media.ManifestParser.HLS,
|
|
serviceDescription: null,
|
|
};
|
|
|
|
// If there is no 'CODECS' attribute in the manifest and codec guessing is
|
|
// disabled, we need to create the segment indexes now so that missing info
|
|
// can be parsed from the media data and added to the stream objects.
|
|
if (!this.codecInfoInManifest_ && this.config_.hls.disableCodecGuessing) {
|
|
const createIndexes = [];
|
|
for (const variant of this.manifest_.variants) {
|
|
if (variant.audio && variant.audio.codecs === '') {
|
|
createIndexes.push(variant.audio.createSegmentIndex());
|
|
}
|
|
if (variant.video && variant.video.codecs === '') {
|
|
createIndexes.push(variant.video.createSegmentIndex());
|
|
}
|
|
}
|
|
|
|
await Promise.all(createIndexes);
|
|
}
|
|
|
|
this.playerInterface_.makeTextStreamsForClosedCaptions(this.manifest_);
|
|
}
|
|
|
|
/**
|
|
* @param {shaka.hls.Playlist} playlist
|
|
* @return {!Promise.<shaka.media.SegmentUtils.BasicInfo>}
|
|
* @private
|
|
*/
|
|
async getMediaPlaylistBasicInfo_(playlist) {
|
|
const HlsParser = shaka.hls.HlsParser;
|
|
const defaultBasicInfo = shaka.media.SegmentUtils.getBasicInfoFromMimeType(
|
|
this.config_.hls.mediaPlaylistFullMimeType);
|
|
if (!playlist.segments.length) {
|
|
return defaultBasicInfo;
|
|
}
|
|
const middleSegmentIdx = Math.trunc((playlist.segments.length - 1) / 2);
|
|
const middleSegment = playlist.segments[middleSegmentIdx];
|
|
const middleSegmentUris = shaka.hls.Utils.constructUris(
|
|
[playlist.absoluteUri],
|
|
playlist.segments[0].verbatimSegmentUri);
|
|
const middleSegmentUri = middleSegmentUris[0];
|
|
const parsedUri = new goog.Uri(middleSegmentUri);
|
|
const extension = parsedUri.getPath().split('.').pop();
|
|
const rawMimeType = HlsParser.RAW_FORMATS_TO_MIME_TYPES_[extension];
|
|
if (rawMimeType) {
|
|
return shaka.media.SegmentUtils.getBasicInfoFromMimeType(
|
|
rawMimeType);
|
|
}
|
|
|
|
const requestType = shaka.net.NetworkingEngine.RequestType.SEGMENT;
|
|
|
|
let initData = null;
|
|
const initSegmentRef = this.getInitSegmentReference_(
|
|
playlist, middleSegment.tags);
|
|
if (initSegmentRef) {
|
|
const initSegmentRequest = shaka.net.NetworkingEngine.makeRequest(
|
|
initSegmentRef.getUris(), this.config_.retryParameters);
|
|
const initType =
|
|
shaka.net.NetworkingEngine.AdvancedRequestType.INIT_SEGMENT;
|
|
const initResponse = await this.makeNetworkRequest_(
|
|
initSegmentRequest, requestType, {type: initType});
|
|
initData = initResponse.data;
|
|
}
|
|
|
|
const segmentRequest = shaka.net.NetworkingEngine.makeRequest(
|
|
middleSegmentUris, this.config_.retryParameters);
|
|
const type = shaka.net.NetworkingEngine.AdvancedRequestType.MEDIA_SEGMENT;
|
|
const response = await this.makeNetworkRequest_(
|
|
segmentRequest, requestType, {type});
|
|
|
|
let contentMimeType = response.headers['content-type'];
|
|
if (contentMimeType) {
|
|
// Split the MIME type in case the server sent additional parameters.
|
|
contentMimeType = contentMimeType.split(';')[0].toLowerCase();
|
|
}
|
|
|
|
if (extension == 'ts' || contentMimeType == 'video/mp2t') {
|
|
const basicInfo =
|
|
shaka.media.SegmentUtils.getBasicInfoFromTs(response.data);
|
|
if (basicInfo) {
|
|
return basicInfo;
|
|
}
|
|
} else if (extension == 'mp4' || extension == 'cmfv' ||
|
|
contentMimeType == 'video/mp4' || contentMimeType == 'audio/mp4') {
|
|
const basicInfo = shaka.media.SegmentUtils.getBasicInfoFromMp4(
|
|
initData, response.data);
|
|
if (basicInfo) {
|
|
return basicInfo;
|
|
}
|
|
}
|
|
return defaultBasicInfo;
|
|
}
|
|
|
|
/** @private */
|
|
determineDuration_() {
|
|
goog.asserts.assert(this.presentationTimeline_,
|
|
'Presentation timeline not created!');
|
|
|
|
if (this.isLive_()) {
|
|
// The spec says nothing much about seeking in live content, but Safari's
|
|
// built-in HLS implementation does not allow it. Therefore we will set
|
|
// the availability window equal to the presentation delay. The player
|
|
// will be able to buffer ahead three segments, but the seek window will
|
|
// be zero-sized.
|
|
const PresentationType = shaka.hls.HlsParser.PresentationType_;
|
|
|
|
if (this.presentationType_ == PresentationType.LIVE) {
|
|
let segmentAvailabilityDuration = this.getLiveDuration_();
|
|
|
|
// This defaults to the presentation delay, which has the effect of
|
|
// making the live stream unseekable. This is consistent with Apple's
|
|
// HLS implementation.
|
|
if (this.config_.hls.useSafariBehaviorForLive) {
|
|
segmentAvailabilityDuration = this.presentationTimeline_.getDelay();
|
|
}
|
|
|
|
// The app can override that with a longer duration, to allow seeking.
|
|
if (!isNaN(this.config_.availabilityWindowOverride)) {
|
|
segmentAvailabilityDuration = this.config_.availabilityWindowOverride;
|
|
}
|
|
|
|
this.presentationTimeline_.setSegmentAvailabilityDuration(
|
|
segmentAvailabilityDuration);
|
|
}
|
|
} else {
|
|
// Use the minimum duration as the presentation duration.
|
|
this.presentationTimeline_.setDuration(this.getMinDuration_());
|
|
}
|
|
|
|
// This is the first point where we have a meaningful presentation start
|
|
// time, and we need to tell PresentationTimeline that so that it can
|
|
// maintain consistency from here on.
|
|
this.presentationTimeline_.lockStartTime();
|
|
|
|
// This asserts that the live edge is being calculated from segment times.
|
|
// For VOD and event streams, this check should still pass.
|
|
goog.asserts.assert(
|
|
!this.presentationTimeline_.usingPresentationStartTime(),
|
|
'We should not be using the presentation start time in HLS!');
|
|
}
|
|
|
|
/**
|
|
* Get the variables of each variant tag, and store in a map.
|
|
* @param {!Array.<!shaka.hls.Tag>} tags Variant tags from the playlist.
|
|
* @private
|
|
*/
|
|
parseMasterVariables_(tags) {
|
|
const queryParams = new goog.Uri(this.masterPlaylistUri_).getQueryData();
|
|
for (const variableTag of tags) {
|
|
const name = variableTag.getAttributeValue('NAME');
|
|
const value = variableTag.getAttributeValue('VALUE');
|
|
const queryParam = variableTag.getAttributeValue('QUERYPARAM');
|
|
if (name && value) {
|
|
if (!this.globalVariables_.has(name)) {
|
|
this.globalVariables_.set(name, value);
|
|
}
|
|
}
|
|
if (queryParam) {
|
|
const queryParamValue = queryParams.get(queryParam)[0];
|
|
if (queryParamValue && !this.globalVariables_.has(queryParamValue)) {
|
|
this.globalVariables_.set(queryParam, queryParamValue);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Get the variables of each variant tag, and store in a map.
|
|
* @param {!Array.<!shaka.hls.Tag>} tags Variant tags from the playlist.
|
|
* @param {string} uri Media playlist URI.
|
|
* @return {!Map.<string, string>}
|
|
* @private
|
|
*/
|
|
parseMediaVariables_(tags, uri) {
|
|
const queryParams = new goog.Uri(uri).getQueryData();
|
|
const mediaVariables = new Map();
|
|
for (const variableTag of tags) {
|
|
const name = variableTag.getAttributeValue('NAME');
|
|
const value = variableTag.getAttributeValue('VALUE');
|
|
const queryParam = variableTag.getAttributeValue('QUERYPARAM');
|
|
const mediaImport = variableTag.getAttributeValue('IMPORT');
|
|
if (name && value) {
|
|
if (!mediaVariables.has(name)) {
|
|
mediaVariables.set(name, value);
|
|
}
|
|
}
|
|
if (queryParam) {
|
|
const queryParamValue = queryParams.get(queryParam)[0];
|
|
if (queryParamValue && !mediaVariables.has(queryParamValue)) {
|
|
mediaVariables.set(queryParam, queryParamValue);
|
|
}
|
|
}
|
|
if (mediaImport) {
|
|
const globalValue = this.globalVariables_.get(mediaImport);
|
|
if (globalValue) {
|
|
mediaVariables.set(mediaImport, globalValue);
|
|
}
|
|
}
|
|
}
|
|
return mediaVariables;
|
|
}
|
|
|
|
/**
|
|
* Get the codecs of each variant tag, and store in a map from
|
|
* audio/video/subtitle group id to the codecs arraylist.
|
|
* @param {!Array.<!shaka.hls.Tag>} tags Variant tags from the playlist.
|
|
* @private
|
|
*/
|
|
parseCodecs_(tags) {
|
|
const ContentType = shaka.util.ManifestParserUtils.ContentType;
|
|
|
|
for (const variantTag of tags) {
|
|
const audioGroupId = variantTag.getAttributeValue('AUDIO');
|
|
const videoGroupId = variantTag.getAttributeValue('VIDEO');
|
|
const subGroupId = variantTag.getAttributeValue('SUBTITLES');
|
|
const allCodecs = this.getCodecsForVariantTag_(variantTag);
|
|
|
|
if (subGroupId) {
|
|
const textCodecs = shaka.util.ManifestParserUtils.guessCodecsSafe(
|
|
ContentType.TEXT, allCodecs);
|
|
goog.asserts.assert(textCodecs != null, 'Text codecs should be valid.');
|
|
this.groupIdToCodecsMap_.set(subGroupId, textCodecs);
|
|
shaka.util.ArrayUtils.remove(allCodecs, textCodecs);
|
|
}
|
|
if (audioGroupId) {
|
|
let codecs = shaka.util.ManifestParserUtils.guessCodecsSafe(
|
|
ContentType.AUDIO, allCodecs);
|
|
if (!codecs) {
|
|
codecs = this.config_.hls.defaultAudioCodec;
|
|
}
|
|
this.groupIdToCodecsMap_.set(audioGroupId, codecs);
|
|
}
|
|
if (videoGroupId) {
|
|
let codecs = shaka.util.ManifestParserUtils.guessCodecsSafe(
|
|
ContentType.VIDEO, allCodecs);
|
|
if (!codecs) {
|
|
codecs = this.config_.hls.defaultVideoCodec;
|
|
}
|
|
this.groupIdToCodecsMap_.set(videoGroupId, codecs);
|
|
}
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Process EXT-X-SESSION-DATA tags.
|
|
*
|
|
* @param {!Array.<!shaka.hls.Tag>} tags
|
|
* @private
|
|
*/
|
|
processSessionData_(tags) {
|
|
for (const tag of tags) {
|
|
const id = tag.getAttributeValue('DATA-ID');
|
|
const uri = tag.getAttributeValue('URI');
|
|
const language = tag.getAttributeValue('LANGUAGE');
|
|
const value = tag.getAttributeValue('VALUE');
|
|
const data = (new Map()).set('id', id);
|
|
if (uri) {
|
|
data.set('uri', shaka.hls.Utils.constructUris(
|
|
[this.masterPlaylistUri_], uri, this.globalVariables_)[0]);
|
|
}
|
|
if (language) {
|
|
data.set('language', language);
|
|
}
|
|
if (value) {
|
|
data.set('value', value);
|
|
}
|
|
const event = new shaka.util.FakeEvent('sessiondata', data);
|
|
if (this.playerInterface_) {
|
|
this.playerInterface_.onEvent(event);
|
|
}
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Parse Subtitles and Closed Captions from 'EXT-X-MEDIA' tags.
|
|
* Create text streams for Subtitles, but not Closed Captions.
|
|
*
|
|
* @param {!Array.<!shaka.hls.Tag>} mediaTags Media tags from the playlist.
|
|
* @return {!Array.<!shaka.extern.Stream>}
|
|
* @private
|
|
*/
|
|
parseTexts_(mediaTags) {
|
|
// Create text stream for each Subtitle media tag.
|
|
const subtitleTags =
|
|
shaka.hls.Utils.filterTagsByType(mediaTags, 'SUBTITLES');
|
|
const textStreams = subtitleTags.map((tag) => {
|
|
const disableText = this.config_.disableText;
|
|
if (disableText) {
|
|
return null;
|
|
}
|
|
try {
|
|
return this.createStreamInfoFromMediaTag_(tag).stream;
|
|
} catch (e) {
|
|
if (this.config_.hls.ignoreTextStreamFailures) {
|
|
return null;
|
|
}
|
|
throw e;
|
|
}
|
|
});
|
|
|
|
const type = shaka.util.ManifestParserUtils.ContentType.TEXT;
|
|
|
|
// Set the codecs for text streams.
|
|
for (const tag of subtitleTags) {
|
|
const groupId = tag.getRequiredAttrValue('GROUP-ID');
|
|
const codecs = this.groupIdToCodecsMap_.get(groupId);
|
|
if (codecs) {
|
|
const textStreamInfos = this.groupIdToStreamInfosMap_.get(groupId);
|
|
if (textStreamInfos) {
|
|
for (const textStreamInfo of textStreamInfos) {
|
|
textStreamInfo.stream.codecs = codecs;
|
|
textStreamInfo.stream.mimeType =
|
|
this.guessMimeTypeBeforeLoading_(type, codecs) ||
|
|
this.guessMimeTypeFallback_(type);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
// Do not create text streams for Closed captions.
|
|
return textStreams.filter((s) => s);
|
|
}
|
|
|
|
/**
|
|
* @param {!Array.<!shaka.hls.Tag>} imageTags from the playlist.
|
|
* @return {!Promise.<!Array.<!shaka.extern.Stream>>}
|
|
* @private
|
|
*/
|
|
async parseImages_(imageTags) {
|
|
// Create image stream for each image tag.
|
|
const imageStreamPromises = imageTags.map(async (tag) => {
|
|
const disableThumbnails = this.config_.disableThumbnails;
|
|
if (disableThumbnails) {
|
|
return null;
|
|
}
|
|
try {
|
|
const streamInfo = await this.createStreamInfoFromImageTag_(tag);
|
|
return streamInfo.stream;
|
|
} catch (e) {
|
|
if (this.config_.hls.ignoreImageStreamFailures) {
|
|
return null;
|
|
}
|
|
throw e;
|
|
}
|
|
});
|
|
const imageStreams = await Promise.all(imageStreamPromises);
|
|
return imageStreams.filter((s) => s);
|
|
}
|
|
|
|
/**
|
|
* @param {!Array.<!shaka.hls.Tag>} mediaTags Media tags from the playlist.
|
|
* @private
|
|
*/
|
|
createStreamInfosFromMediaTags_(mediaTags) {
|
|
// Filter out subtitles and media tags without uri.
|
|
mediaTags = mediaTags.filter((tag) => {
|
|
const uri = tag.getAttributeValue('URI') || '';
|
|
const type = tag.getAttributeValue('TYPE');
|
|
return type != 'SUBTITLES' && uri != '';
|
|
});
|
|
|
|
// Create stream info for each audio / video media tag.
|
|
for (const tag of mediaTags) {
|
|
this.createStreamInfoFromMediaTag_(tag);
|
|
}
|
|
}
|
|
|
|
/**
|
|
* @param {!Array.<!shaka.hls.Tag>} tags Variant tags from the playlist.
|
|
* @param {!Array.<!shaka.hls.Tag>} sessionKeyTags EXT-X-SESSION-KEY tags
|
|
* from the playlist.
|
|
* @return {!Array.<!shaka.extern.Variant>}
|
|
* @private
|
|
*/
|
|
createVariantsForTags_(tags, sessionKeyTags) {
|
|
// EXT-X-SESSION-KEY processing
|
|
const drmInfos = [];
|
|
const keyIds = new Set();
|
|
if (sessionKeyTags.length > 0) {
|
|
for (const drmTag of sessionKeyTags) {
|
|
const method = drmTag.getRequiredAttrValue('METHOD');
|
|
if (method != 'NONE' && method != 'AES-128') {
|
|
// According to the HLS spec, KEYFORMAT is optional and implicitly
|
|
// defaults to "identity".
|
|
// https://datatracker.ietf.org/doc/html/draft-pantos-hls-rfc8216bis-11#section-4.4.4.4
|
|
const keyFormat =
|
|
drmTag.getAttributeValue('KEYFORMAT') || 'identity';
|
|
const drmParser =
|
|
shaka.hls.HlsParser.KEYFORMATS_TO_DRM_PARSERS_[keyFormat];
|
|
|
|
const drmInfo = drmParser ?
|
|
drmParser(drmTag, /* mimeType= */ '') : null;
|
|
if (drmInfo) {
|
|
if (drmInfo.keyIds) {
|
|
for (const keyId of drmInfo.keyIds) {
|
|
keyIds.add(keyId);
|
|
}
|
|
}
|
|
drmInfos.push(drmInfo);
|
|
} else {
|
|
shaka.log.warning('Unsupported HLS KEYFORMAT', keyFormat);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
// Create variants for each variant tag.
|
|
const allVariants = tags.map((tag) => {
|
|
const frameRate = tag.getAttributeValue('FRAME-RATE');
|
|
const bandwidth = Number(tag.getAttributeValue('AVERAGE-BANDWIDTH')) ||
|
|
Number(tag.getRequiredAttrValue('BANDWIDTH'));
|
|
|
|
const resolution = tag.getAttributeValue('RESOLUTION');
|
|
const [width, height] = resolution ? resolution.split('x') : [null, null];
|
|
|
|
const videoRange = tag.getAttributeValue('VIDEO-RANGE');
|
|
|
|
let videoLayout = tag.getAttributeValue('REQ-VIDEO-LAYOUT');
|
|
if (videoLayout && videoLayout.includes(',')) {
|
|
// If multiple video layout strings are present, pick the first valid
|
|
// one.
|
|
const layoutStrings = videoLayout.split(',').filter((layoutString) => {
|
|
return layoutString == 'CH-STEREO' || layoutString == 'CH-MONO';
|
|
});
|
|
videoLayout = layoutStrings[0];
|
|
}
|
|
// According to the HLS spec:
|
|
// By default a video variant is monoscopic, so an attribute
|
|
// consisting entirely of REQ-VIDEO-LAYOUT="CH-MONO" is unnecessary
|
|
// and SHOULD NOT be present.
|
|
videoLayout = videoLayout || 'CH-MONO';
|
|
|
|
const streamInfos = this.createStreamInfosForVariantTag_(tag,
|
|
resolution, frameRate);
|
|
|
|
goog.asserts.assert(streamInfos.audio.length ||
|
|
streamInfos.video.length, 'We should have created a stream!');
|
|
|
|
return this.createVariants_(
|
|
streamInfos.audio,
|
|
streamInfos.video,
|
|
bandwidth,
|
|
width,
|
|
height,
|
|
frameRate,
|
|
videoRange,
|
|
videoLayout,
|
|
drmInfos,
|
|
keyIds);
|
|
});
|
|
let variants = allVariants.reduce(shaka.util.Functional.collapseArrays, []);
|
|
// Filter out null variants.
|
|
variants = variants.filter((variant) => variant != null);
|
|
return variants;
|
|
}
|
|
|
|
/**
|
|
* Create audio and video streamInfos from an 'EXT-X-STREAM-INF' tag and its
|
|
* related media tags.
|
|
*
|
|
* @param {!shaka.hls.Tag} tag
|
|
* @param {?string} resolution
|
|
* @param {?string} frameRate
|
|
* @return {!shaka.hls.HlsParser.StreamInfos}
|
|
* @private
|
|
*/
|
|
createStreamInfosForVariantTag_(tag, resolution, frameRate) {
|
|
const ContentType = shaka.util.ManifestParserUtils.ContentType;
|
|
/** @type {!Array.<string>} */
|
|
let allCodecs = this.getCodecsForVariantTag_(tag);
|
|
const audioGroupId = tag.getAttributeValue('AUDIO');
|
|
const videoGroupId = tag.getAttributeValue('VIDEO');
|
|
goog.asserts.assert(audioGroupId == null || videoGroupId == null,
|
|
'Unexpected: both video and audio described by media tags!');
|
|
|
|
const groupId = audioGroupId || videoGroupId;
|
|
const streamInfos =
|
|
(groupId && this.groupIdToStreamInfosMap_.has(groupId)) ?
|
|
this.groupIdToStreamInfosMap_.get(groupId) : [];
|
|
|
|
/** @type {shaka.hls.HlsParser.StreamInfos} */
|
|
const res = {
|
|
audio: audioGroupId ? streamInfos : [],
|
|
video: videoGroupId ? streamInfos : [],
|
|
};
|
|
|
|
// Make an educated guess about the stream type.
|
|
shaka.log.debug('Guessing stream type for', tag.toString());
|
|
let type;
|
|
let ignoreStream = false;
|
|
|
|
// The Microsoft HLS manifest generators will make audio-only variants
|
|
// that link to their URI both directly and through an audio tag.
|
|
// In that case, ignore the local URI and use the version in the
|
|
// AUDIO tag, so you inherit its language.
|
|
// As an example, see the manifest linked in issue #860.
|
|
const streamURI = tag.getRequiredAttrValue('URI');
|
|
const hasSameUri = res.audio.find((audio) => {
|
|
return audio && audio.verbatimMediaPlaylistUri == streamURI;
|
|
});
|
|
|
|
const videoCodecs = shaka.util.ManifestParserUtils.guessCodecsSafe(
|
|
ContentType.VIDEO, allCodecs);
|
|
const audioCodecs = shaka.util.ManifestParserUtils.guessCodecsSafe(
|
|
ContentType.AUDIO, allCodecs);
|
|
|
|
if (audioCodecs && !videoCodecs) {
|
|
// There are no associated media tags, and there's only audio codec,
|
|
// and no video codec, so it should be audio.
|
|
type = ContentType.AUDIO;
|
|
shaka.log.debug('Guessing audio-only.');
|
|
ignoreStream = res.audio.length > 0;
|
|
} else if (!streamInfos.length && audioCodecs && videoCodecs) {
|
|
// There are both audio and video codecs, so assume multiplexed content.
|
|
// Note that the default used when CODECS is missing assumes multiple
|
|
// (and therefore multiplexed).
|
|
// Recombine the codec strings into one so that MediaSource isn't
|
|
// lied to later. (That would trigger an error in Chrome.)
|
|
shaka.log.debug('Guessing multiplexed audio+video.');
|
|
type = ContentType.VIDEO;
|
|
allCodecs = [[videoCodecs, audioCodecs].join(',')];
|
|
} else if (res.audio.length && hasSameUri) {
|
|
shaka.log.debug('Guessing audio-only.');
|
|
type = ContentType.AUDIO;
|
|
ignoreStream = true;
|
|
} else if (res.video.length && !res.audio.length) {
|
|
// There are associated video streams. Assume this is audio.
|
|
shaka.log.debug('Guessing audio-only.');
|
|
type = ContentType.AUDIO;
|
|
} else {
|
|
shaka.log.debug('Guessing video-only.');
|
|
type = ContentType.VIDEO;
|
|
}
|
|
|
|
if (!ignoreStream) {
|
|
const streamInfo =
|
|
this.createStreamInfoFromVariantTag_(tag, allCodecs, type);
|
|
streamInfo.stream.groupId = groupId;
|
|
res[streamInfo.stream.type] = [streamInfo];
|
|
}
|
|
return res;
|
|
}
|
|
|
|
|
|
/**
|
|
* Get the codecs from the 'EXT-X-STREAM-INF' tag.
|
|
*
|
|
* @param {!shaka.hls.Tag} tag
|
|
* @return {!Array.<string>} codecs
|
|
* @private
|
|
*/
|
|
getCodecsForVariantTag_(tag) {
|
|
let codecsString = tag.getAttributeValue('CODECS') || '';
|
|
|
|
this.codecInfoInManifest_ = codecsString.length > 0;
|
|
|
|
if (!this.codecInfoInManifest_ && !this.config_.hls.disableCodecGuessing) {
|
|
// These are the default codecs to assume if none are specified.
|
|
const defaultCodecsArray = [];
|
|
|
|
if (!this.config_.disableVideo) {
|
|
defaultCodecsArray.push(this.config_.hls.defaultVideoCodec);
|
|
}
|
|
if (!this.config_.disableAudio) {
|
|
defaultCodecsArray.push(this.config_.hls.defaultAudioCodec);
|
|
}
|
|
|
|
codecsString = defaultCodecsArray.join(',');
|
|
}
|
|
|
|
// Strip out internal whitespace while splitting on commas:
|
|
/** @type {!Array.<string>} */
|
|
const codecs = codecsString.split(/\s*,\s*/);
|
|
|
|
return this.filterDuplicateCodecs_(codecs);
|
|
}
|
|
|
|
|
|
/**
|
|
* @param {!Array.<string>} codecs
|
|
* @return {!Array.<string>} codecs
|
|
* @private
|
|
*/
|
|
filterDuplicateCodecs_(codecs) {
|
|
// Filter out duplicate codecs.
|
|
const seen = new Set();
|
|
const ret = [];
|
|
for (const codec of codecs) {
|
|
// HLS says the CODECS field needs to include all codecs that appear in
|
|
// the content. This means that if the content changes profiles, it should
|
|
// include both. Since all known browsers support changing profiles
|
|
// without any other work, just ignore them. See also:
|
|
// https://github.com/shaka-project/shaka-player/issues/1817
|
|
const shortCodec = shaka.util.MimeUtils.getCodecBase(codec);
|
|
if (!seen.has(shortCodec)) {
|
|
ret.push(codec);
|
|
seen.add(shortCodec);
|
|
} else {
|
|
shaka.log.debug('Ignoring duplicate codec');
|
|
}
|
|
}
|
|
return ret;
|
|
}
|
|
|
|
/**
|
|
* Get the channel count information for an HLS audio track.
|
|
* CHANNELS specifies an ordered, "/" separated list of parameters.
|
|
* If the type is audio, the first parameter will be a decimal integer
|
|
* specifying the number of independent, simultaneous audio channels.
|
|
* No other channels parameters are currently defined.
|
|
*
|
|
* @param {!shaka.hls.Tag} tag
|
|
* @return {?number}
|
|
* @private
|
|
*/
|
|
getChannelsCount_(tag) {
|
|
const channels = tag.getAttributeValue('CHANNELS');
|
|
if (!channels) {
|
|
return null;
|
|
}
|
|
const channelcountstring = channels.split('/')[0];
|
|
const count = parseInt(channelcountstring, 10);
|
|
return count;
|
|
}
|
|
|
|
/**
|
|
* Get the sample rate information for an HLS audio track.
|
|
*
|
|
* @param {!shaka.hls.Tag} tag
|
|
* @return {?number}
|
|
* @private
|
|
*/
|
|
getSampleRate_(tag) {
|
|
const sampleRate = tag.getAttributeValue('SAMPLE-RATE');
|
|
if (!sampleRate) {
|
|
return null;
|
|
}
|
|
return parseInt(sampleRate, 10);
|
|
}
|
|
|
|
/**
|
|
* Get the spatial audio information for an HLS audio track.
|
|
* In HLS the channels field indicates the number of audio channels that the
|
|
* stream has (eg: 2). In the case of Dolby Atmos, the complexity is
|
|
* expressed with the number of channels followed by the word JOC
|
|
* (eg: 16/JOC), so 16 would be the number of channels (eg: 7.3.6 layout),
|
|
* and JOC indicates that the stream has spatial audio.
|
|
* @see https://developer.apple.com/documentation/http_live_streaming/hls_authoring_specification_for_apple_devices/hls_authoring_specification_for_apple_devices_appendixes
|
|
*
|
|
* @param {!shaka.hls.Tag} tag
|
|
* @return {boolean}
|
|
* @private
|
|
*/
|
|
isSpatialAudio_(tag) {
|
|
const channels = tag.getAttributeValue('CHANNELS');
|
|
if (!channels) {
|
|
return false;
|
|
}
|
|
return channels.includes('/JOC');
|
|
}
|
|
|
|
/**
|
|
* Get the closed captions map information for the EXT-X-STREAM-INF tag, to
|
|
* create the stream info.
|
|
* @param {!shaka.hls.Tag} tag
|
|
* @param {string} type
|
|
* @return {Map.<string, string>} closedCaptions
|
|
* @private
|
|
*/
|
|
getClosedCaptions_(tag, type) {
|
|
const ContentType = shaka.util.ManifestParserUtils.ContentType;
|
|
// The attribute of closed captions is optional, and the value may be
|
|
// 'NONE'.
|
|
const closedCaptionsAttr = tag.getAttributeValue('CLOSED-CAPTIONS');
|
|
|
|
// EXT-X-STREAM-INF tags may have CLOSED-CAPTIONS attributes.
|
|
// The value can be either a quoted-string or an enumerated-string with
|
|
// the value NONE. If the value is a quoted-string, it MUST match the
|
|
// value of the GROUP-ID attribute of an EXT-X-MEDIA tag elsewhere in the
|
|
// Playlist whose TYPE attribute is CLOSED-CAPTIONS.
|
|
if (type == ContentType.VIDEO && closedCaptionsAttr &&
|
|
closedCaptionsAttr != 'NONE') {
|
|
return this.groupIdToClosedCaptionsMap_.get(closedCaptionsAttr);
|
|
}
|
|
return null;
|
|
}
|
|
|
|
/**
|
|
* Get the normalized language value.
|
|
*
|
|
* @param {?string} languageValue
|
|
* @return {string}
|
|
* @private
|
|
*/
|
|
getLanguage_(languageValue) {
|
|
const LanguageUtils = shaka.util.LanguageUtils;
|
|
return LanguageUtils.normalize(languageValue || 'und');
|
|
}
|
|
|
|
/**
|
|
* Get the type value.
|
|
* Shaka recognizes the content types 'audio', 'video', 'text', and 'image'.
|
|
* The HLS 'subtitles' type needs to be mapped to 'text'.
|
|
* @param {!shaka.hls.Tag} tag
|
|
* @return {string}
|
|
* @private
|
|
*/
|
|
getType_(tag) {
|
|
let type = tag.getRequiredAttrValue('TYPE').toLowerCase();
|
|
if (type == 'subtitles') {
|
|
type = shaka.util.ManifestParserUtils.ContentType.TEXT;
|
|
}
|
|
return type;
|
|
}
|
|
|
|
/**
|
|
* @param {!Array.<shaka.hls.HlsParser.StreamInfo>} audioInfos
|
|
* @param {!Array.<shaka.hls.HlsParser.StreamInfo>} videoInfos
|
|
* @param {number} bandwidth
|
|
* @param {?string} width
|
|
* @param {?string} height
|
|
* @param {?string} frameRate
|
|
* @param {?string} videoRange
|
|
* @param {?string} videoLayout
|
|
* @param {!Array.<shaka.extern.DrmInfo>} drmInfos
|
|
* @param {!Set.<string>} keyIds
|
|
* @return {!Array.<!shaka.extern.Variant>}
|
|
* @private
|
|
*/
|
|
createVariants_(
|
|
audioInfos, videoInfos, bandwidth, width, height, frameRate, videoRange,
|
|
videoLayout, drmInfos, keyIds) {
|
|
const ContentType = shaka.util.ManifestParserUtils.ContentType;
|
|
const DrmEngine = shaka.media.DrmEngine;
|
|
|
|
for (const info of videoInfos) {
|
|
this.addVideoAttributes_(
|
|
info.stream, width, height, frameRate, videoRange, videoLayout);
|
|
}
|
|
|
|
// In case of audio-only or video-only content or the audio/video is
|
|
// disabled by the config, we create an array of one item containing
|
|
// a null. This way, the double-loop works for all kinds of content.
|
|
// NOTE: we currently don't have support for audio-only content.
|
|
const disableAudio = this.config_.disableAudio;
|
|
if (!audioInfos.length || disableAudio) {
|
|
audioInfos = [null];
|
|
}
|
|
const disableVideo = this.config_.disableVideo;
|
|
if (!videoInfos.length || disableVideo) {
|
|
videoInfos = [null];
|
|
}
|
|
|
|
const variants = [];
|
|
for (const audioInfo of audioInfos) {
|
|
for (const videoInfo of videoInfos) {
|
|
const audioStream = audioInfo ? audioInfo.stream : null;
|
|
if (audioStream) {
|
|
audioStream.drmInfos = drmInfos;
|
|
audioStream.keyIds = keyIds;
|
|
}
|
|
const videoStream = videoInfo ? videoInfo.stream : null;
|
|
if (videoStream) {
|
|
videoStream.drmInfos = drmInfos;
|
|
videoStream.keyIds = keyIds;
|
|
}
|
|
const audioDrmInfos = audioInfo ? audioInfo.stream.drmInfos : null;
|
|
const videoDrmInfos = videoInfo ? videoInfo.stream.drmInfos : null;
|
|
const videoStreamUri =
|
|
videoInfo ? videoInfo.verbatimMediaPlaylistUri : '';
|
|
const audioStreamUri =
|
|
audioInfo ? audioInfo.verbatimMediaPlaylistUri : '';
|
|
const variantUriKey = videoStreamUri + ' - ' + audioStreamUri;
|
|
|
|
if (audioStream && videoStream) {
|
|
if (!DrmEngine.areDrmCompatible(audioDrmInfos, videoDrmInfos)) {
|
|
shaka.log.warning(
|
|
'Incompatible DRM info in HLS variant. Skipping.');
|
|
continue;
|
|
}
|
|
}
|
|
|
|
if (this.variantUriSet_.has(variantUriKey)) {
|
|
// This happens when two variants only differ in their text streams.
|
|
shaka.log.debug(
|
|
'Skipping variant which only differs in text streams.');
|
|
continue;
|
|
}
|
|
|
|
// Since both audio and video are of the same type, this assertion will
|
|
// catch certain mistakes at runtime that the compiler would miss.
|
|
goog.asserts.assert(!audioStream ||
|
|
audioStream.type == ContentType.AUDIO, 'Audio parameter mismatch!');
|
|
goog.asserts.assert(!videoStream ||
|
|
videoStream.type == ContentType.VIDEO, 'Video parameter mismatch!');
|
|
|
|
const variant = {
|
|
id: this.globalId_++,
|
|
language: audioStream ? audioStream.language : 'und',
|
|
disabledUntilTime: 0,
|
|
primary: (!!audioStream && audioStream.primary) ||
|
|
(!!videoStream && videoStream.primary),
|
|
audio: audioStream,
|
|
video: videoStream,
|
|
bandwidth,
|
|
allowedByApplication: true,
|
|
allowedByKeySystem: true,
|
|
decodingInfos: [],
|
|
};
|
|
|
|
variants.push(variant);
|
|
this.variantUriSet_.add(variantUriKey);
|
|
}
|
|
}
|
|
return variants;
|
|
}
|
|
|
|
/**
|
|
* Parses an array of EXT-X-MEDIA tags, then stores the values of all tags
|
|
* with TYPE="CLOSED-CAPTIONS" into a map of group id to closed captions.
|
|
*
|
|
* @param {!Array.<!shaka.hls.Tag>} mediaTags
|
|
* @private
|
|
*/
|
|
parseClosedCaptions_(mediaTags) {
|
|
const closedCaptionsTags =
|
|
shaka.hls.Utils.filterTagsByType(mediaTags, 'CLOSED-CAPTIONS');
|
|
for (const tag of closedCaptionsTags) {
|
|
goog.asserts.assert(tag.name == 'EXT-X-MEDIA',
|
|
'Should only be called on media tags!');
|
|
const languageValue = tag.getAttributeValue('LANGUAGE');
|
|
let language = this.getLanguage_(languageValue);
|
|
if (!languageValue) {
|
|
const nameValue = tag.getAttributeValue('NAME');
|
|
if (nameValue) {
|
|
language = nameValue;
|
|
}
|
|
}
|
|
|
|
// The GROUP-ID value is a quoted-string that specifies the group to which
|
|
// the Rendition belongs.
|
|
const groupId = tag.getRequiredAttrValue('GROUP-ID');
|
|
|
|
// The value of INSTREAM-ID is a quoted-string that specifies a Rendition
|
|
// within the segments in the Media Playlist. This attribute is REQUIRED
|
|
// if the TYPE attribute is CLOSED-CAPTIONS.
|
|
// We need replace SERVICE string by our internal svc string.
|
|
const instreamId = tag.getRequiredAttrValue('INSTREAM-ID')
|
|
.replace('SERVICE', 'svc');
|
|
if (!this.groupIdToClosedCaptionsMap_.get(groupId)) {
|
|
this.groupIdToClosedCaptionsMap_.set(groupId, new Map());
|
|
}
|
|
this.groupIdToClosedCaptionsMap_.get(groupId).set(instreamId, language);
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Parse EXT-X-MEDIA media tag into a Stream object.
|
|
*
|
|
* @param {shaka.hls.Tag} tag
|
|
* @return {!shaka.hls.HlsParser.StreamInfo}
|
|
* @private
|
|
*/
|
|
createStreamInfoFromMediaTag_(tag) {
|
|
goog.asserts.assert(tag.name == 'EXT-X-MEDIA',
|
|
'Should only be called on media tags!');
|
|
const groupId = tag.getRequiredAttrValue('GROUP-ID');
|
|
let codecs = '';
|
|
/** @type {string} */
|
|
const type = this.getType_(tag);
|
|
if (type == shaka.util.ManifestParserUtils.ContentType.TEXT) {
|
|
codecs = tag.getAttributeValue('CODECS') || '';
|
|
} else if (groupId && this.groupIdToCodecsMap_.has(groupId)) {
|
|
codecs = this.groupIdToCodecsMap_.get(groupId);
|
|
}
|
|
|
|
const verbatimMediaPlaylistUri = tag.getRequiredAttrValue('URI');
|
|
|
|
// Check if the stream has already been created as part of another Variant
|
|
// and return it if it has.
|
|
if (this.uriToStreamInfosMap_.has(verbatimMediaPlaylistUri)) {
|
|
return this.uriToStreamInfosMap_.get(verbatimMediaPlaylistUri);
|
|
}
|
|
|
|
const language = tag.getAttributeValue('LANGUAGE');
|
|
const name = tag.getAttributeValue('NAME');
|
|
|
|
// NOTE: According to the HLS spec, "DEFAULT=YES" requires "AUTOSELECT=YES".
|
|
// However, we don't bother to validate "AUTOSELECT", since we don't
|
|
// actually use it in our streaming model, and we treat everything as
|
|
// "AUTOSELECT=YES". A value of "AUTOSELECT=NO" would imply that it may
|
|
// only be selected explicitly by the user, and we don't have a way to
|
|
// represent that in our model.
|
|
const defaultAttrValue = tag.getAttributeValue('DEFAULT');
|
|
const primary = defaultAttrValue == 'YES';
|
|
|
|
const channelsCount = type == 'audio' ? this.getChannelsCount_(tag) : null;
|
|
const spatialAudio = type == 'audio' ? this.isSpatialAudio_(tag) : false;
|
|
const characteristics = tag.getAttributeValue('CHARACTERISTICS');
|
|
|
|
const forcedAttrValue = tag.getAttributeValue('FORCED');
|
|
const forced = forcedAttrValue == 'YES';
|
|
const sampleRate = type == 'audio' ? this.getSampleRate_(tag) : null;
|
|
// TODO: Should we take into account some of the currently ignored
|
|
// attributes: INSTREAM-ID, Attribute descriptions: https://bit.ly/2lpjOhj
|
|
const streamInfo = this.createStreamInfo_(
|
|
verbatimMediaPlaylistUri, codecs, type, language, primary, name,
|
|
channelsCount, /* closedCaptions= */ null, characteristics, forced,
|
|
sampleRate, spatialAudio);
|
|
if (streamInfo.stream) {
|
|
streamInfo.stream.groupId = groupId;
|
|
}
|
|
if (this.groupIdToStreamInfosMap_.has(groupId)) {
|
|
this.groupIdToStreamInfosMap_.get(groupId).push(streamInfo);
|
|
} else {
|
|
this.groupIdToStreamInfosMap_.set(groupId, [streamInfo]);
|
|
}
|
|
|
|
// TODO: This check is necessary because of the possibility of multiple
|
|
// calls to createStreamInfoFromMediaTag_ before either has resolved.
|
|
if (this.uriToStreamInfosMap_.has(verbatimMediaPlaylistUri)) {
|
|
return this.uriToStreamInfosMap_.get(verbatimMediaPlaylistUri);
|
|
}
|
|
this.uriToStreamInfosMap_.set(verbatimMediaPlaylistUri, streamInfo);
|
|
return streamInfo;
|
|
}
|
|
|
|
/**
|
|
* Parse EXT-X-MEDIA media tag into a Stream object.
|
|
*
|
|
* @param {shaka.hls.Tag} tag
|
|
* @return {!Promise.<!shaka.hls.HlsParser.StreamInfo>}
|
|
* @private
|
|
*/
|
|
async createStreamInfoFromImageTag_(tag) {
|
|
goog.asserts.assert(tag.name == 'EXT-X-IMAGE-STREAM-INF',
|
|
'Should only be called on image tags!');
|
|
/** @type {string} */
|
|
const type = shaka.util.ManifestParserUtils.ContentType.IMAGE;
|
|
|
|
const verbatimImagePlaylistUri = tag.getRequiredAttrValue('URI');
|
|
const codecs = tag.getAttributeValue('CODECS', 'jpeg') || '';
|
|
|
|
// Check if the stream has already been created as part of another Variant
|
|
// and return it if it has.
|
|
if (this.uriToStreamInfosMap_.has(verbatimImagePlaylistUri)) {
|
|
return this.uriToStreamInfosMap_.get(verbatimImagePlaylistUri);
|
|
}
|
|
|
|
const language = tag.getAttributeValue('LANGUAGE');
|
|
const name = tag.getAttributeValue('NAME');
|
|
|
|
const characteristics = tag.getAttributeValue('CHARACTERISTICS');
|
|
|
|
const streamInfo = this.createStreamInfo_(
|
|
verbatimImagePlaylistUri, codecs, type, language, /* primary= */ false,
|
|
name, /* channelsCount= */ null, /* closedCaptions= */ null,
|
|
characteristics, /* forced= */ false, /* sampleRate= */ null,
|
|
/* spatialAudio= */ false);
|
|
|
|
// TODO: This check is necessary because of the possibility of multiple
|
|
// calls to createStreamInfoFromImageTag_ before either has resolved.
|
|
if (this.uriToStreamInfosMap_.has(verbatimImagePlaylistUri)) {
|
|
return this.uriToStreamInfosMap_.get(verbatimImagePlaylistUri);
|
|
}
|
|
|
|
// Parse misc attributes.
|
|
const resolution = tag.getAttributeValue('RESOLUTION');
|
|
if (resolution) {
|
|
// The RESOLUTION tag represents the resolution of a single thumbnail, not
|
|
// of the entire sheet at once (like we expect in the output).
|
|
// So multiply by the layout size.
|
|
|
|
// Since we need to have generated the segment index for this, we can't
|
|
// lazy-load in this situation.
|
|
await streamInfo.stream.createSegmentIndex();
|
|
|
|
const reference = streamInfo.stream.segmentIndex.get(0);
|
|
const layout = reference.getTilesLayout();
|
|
if (layout) {
|
|
streamInfo.stream.width =
|
|
Number(resolution.split('x')[0]) * Number(layout.split('x')[0]);
|
|
streamInfo.stream.height =
|
|
Number(resolution.split('x')[1]) * Number(layout.split('x')[1]);
|
|
// TODO: What happens if there are multiple grids, with different
|
|
// layout sizes, inside this image stream?
|
|
}
|
|
}
|
|
const bandwidth = tag.getAttributeValue('BANDWIDTH');
|
|
if (bandwidth) {
|
|
streamInfo.stream.bandwidth = Number(bandwidth);
|
|
}
|
|
|
|
this.uriToStreamInfosMap_.set(verbatimImagePlaylistUri, streamInfo);
|
|
return streamInfo;
|
|
}
|
|
|
|
/**
|
|
* Parse an EXT-X-STREAM-INF media tag into a Stream object.
|
|
*
|
|
* @param {!shaka.hls.Tag} tag
|
|
* @param {!Array.<string>} allCodecs
|
|
* @param {string} type
|
|
* @return {!shaka.hls.HlsParser.StreamInfo}
|
|
* @private
|
|
*/
|
|
createStreamInfoFromVariantTag_(tag, allCodecs, type) {
|
|
goog.asserts.assert(tag.name == 'EXT-X-STREAM-INF',
|
|
'Should only be called on variant tags!');
|
|
const verbatimMediaPlaylistUri = tag.getRequiredAttrValue('URI');
|
|
|
|
if (this.uriToStreamInfosMap_.has(verbatimMediaPlaylistUri)) {
|
|
return this.uriToStreamInfosMap_.get(verbatimMediaPlaylistUri);
|
|
}
|
|
|
|
const closedCaptions = this.getClosedCaptions_(tag, type);
|
|
const codecs = shaka.util.ManifestParserUtils.guessCodecs(type, allCodecs);
|
|
const streamInfo = this.createStreamInfo_(verbatimMediaPlaylistUri,
|
|
codecs, type, /* language= */ null, /* primary= */ false,
|
|
/* name= */ null, /* channelcount= */ null, closedCaptions,
|
|
/* characteristics= */ null, /* forced= */ false,
|
|
/* sampleRate= */ null, /* spatialAudio= */ false);
|
|
// TODO: This check is necessary because of the possibility of multiple
|
|
// calls to createStreamInfoFromVariantTag_ before either has resolved.
|
|
if (this.uriToStreamInfosMap_.has(verbatimMediaPlaylistUri)) {
|
|
return this.uriToStreamInfosMap_.get(verbatimMediaPlaylistUri);
|
|
}
|
|
|
|
this.uriToStreamInfosMap_.set(verbatimMediaPlaylistUri, streamInfo);
|
|
return streamInfo;
|
|
}
|
|
|
|
|
|
/**
|
|
* @param {string} verbatimMediaPlaylistUri
|
|
* @param {string} codecs
|
|
* @param {string} type
|
|
* @param {?string} languageValue
|
|
* @param {boolean} primary
|
|
* @param {?string} name
|
|
* @param {?number} channelsCount
|
|
* @param {Map.<string, string>} closedCaptions
|
|
* @param {?string} characteristics
|
|
* @param {boolean} forced
|
|
* @param {?number} sampleRate
|
|
* @param {boolean} spatialAudio
|
|
* @return {!shaka.hls.HlsParser.StreamInfo}
|
|
* @private
|
|
*/
|
|
createStreamInfo_(verbatimMediaPlaylistUri, codecs, type, languageValue,
|
|
primary, name, channelsCount, closedCaptions, characteristics, forced,
|
|
sampleRate, spatialAudio) {
|
|
// TODO: Refactor, too many parameters
|
|
const initialMediaPlaylistUris = shaka.hls.Utils.constructUris(
|
|
[this.masterPlaylistUri_], verbatimMediaPlaylistUri,
|
|
this.globalVariables_);
|
|
|
|
// This stream is lazy-loaded inside the createSegmentIndex function.
|
|
// So we start out with a stream object that does not contain the actual
|
|
// segment index, then download when createSegmentIndex is called.
|
|
const stream = this.makeStreamObject_(codecs, type, languageValue, primary,
|
|
name, channelsCount, closedCaptions, characteristics, forced,
|
|
sampleRate, spatialAudio);
|
|
const streamInfo = {
|
|
stream,
|
|
type,
|
|
verbatimMediaPlaylistUri,
|
|
// These values are filled out or updated after lazy-loading:
|
|
absoluteMediaPlaylistUris: initialMediaPlaylistUris,
|
|
minTimestamp: 0,
|
|
maxTimestamp: 0,
|
|
mediaSequenceToStartTime: new Map(),
|
|
canSkipSegments: false,
|
|
canBlockReload: false,
|
|
hasEndList: false,
|
|
firstSequenceNumber: -1,
|
|
nextMediaSequence: -1,
|
|
nextPart: -1,
|
|
loadedOnce: false,
|
|
};
|
|
|
|
/** @param {!AbortSignal} abortSignal */
|
|
const downloadSegmentIndex = async (abortSignal) => {
|
|
// Download the actual manifest.
|
|
const response = await this.requestManifest_(
|
|
streamInfo.absoluteMediaPlaylistUris, /* isPlaylist= */ true);
|
|
if (abortSignal.aborted) {
|
|
return;
|
|
}
|
|
|
|
// Record the final URI after redirects.
|
|
const absoluteMediaPlaylistUri = response.uri;
|
|
|
|
// Record the redirected, final URI of this media playlist when we parse
|
|
// it.
|
|
/** @type {!shaka.hls.Playlist} */
|
|
const playlist = this.manifestTextParser_.parsePlaylist(
|
|
response.data, absoluteMediaPlaylistUri);
|
|
|
|
let mimeType = undefined;
|
|
|
|
// If no codec info was provided in the manifest and codec guessing is
|
|
// disabled we try to get necessary info from the media data.
|
|
if (!this.codecInfoInManifest_ && this.config_.hls.disableCodecGuessing) {
|
|
const basicInfo = await this.getMediaPlaylistBasicInfo_(playlist);
|
|
|
|
goog.asserts.assert(
|
|
type === basicInfo.type, 'Media types should match!');
|
|
|
|
mimeType = basicInfo.mimeType;
|
|
codecs = basicInfo.codecs;
|
|
}
|
|
|
|
const wasLive = this.isLive_();
|
|
const realStreamInfo = await this.convertParsedPlaylistIntoStreamInfo_(
|
|
playlist, verbatimMediaPlaylistUri, absoluteMediaPlaylistUri, codecs,
|
|
type, languageValue, primary, name, channelsCount, closedCaptions,
|
|
characteristics, forced, sampleRate, spatialAudio, mimeType);
|
|
if (abortSignal.aborted) {
|
|
return;
|
|
}
|
|
|
|
const realStream = realStreamInfo.stream;
|
|
|
|
if (this.isLive_() && !wasLive) {
|
|
// Now that we know that the presentation is live, convert the timeline
|
|
// to live.
|
|
this.changePresentationTimelineToLive_(playlist);
|
|
}
|
|
|
|
// Copy values from the real stream info to our initial one.
|
|
streamInfo.absoluteMediaPlaylistUri = absoluteMediaPlaylistUri;
|
|
streamInfo.minTimestamp = realStreamInfo.minTimestamp;
|
|
streamInfo.maxTimestamp = realStreamInfo.maxTimestamp;
|
|
streamInfo.canSkipSegments = realStreamInfo.canSkipSegments;
|
|
streamInfo.canBlockReload = realStreamInfo.canBlockReload;
|
|
streamInfo.hasEndList = realStreamInfo.hasEndList;
|
|
streamInfo.mediaSequenceToStartTime =
|
|
realStreamInfo.mediaSequenceToStartTime;
|
|
streamInfo.nextMediaSequence = realStreamInfo.nextMediaSequence;
|
|
streamInfo.nextPart = realStreamInfo.nextPart;
|
|
streamInfo.loadedOnce = true;
|
|
stream.segmentIndex = realStream.segmentIndex;
|
|
stream.encrypted = realStream.encrypted;
|
|
stream.drmInfos = realStream.drmInfos;
|
|
stream.keyIds = realStream.keyIds;
|
|
stream.mimeType = realStream.mimeType;
|
|
stream.bandwidth = realStream.bandwidth;
|
|
stream.codecs = realStream.codecs || stream.codecs;
|
|
|
|
// Since we lazy-loaded this content, the player may need to create new
|
|
// sessions for the DRM info in this stream.
|
|
if (stream.drmInfos.length) {
|
|
this.playerInterface_.newDrmInfo(stream);
|
|
}
|
|
|
|
const ContentType = shaka.util.ManifestParserUtils.ContentType;
|
|
if (type == ContentType.VIDEO || type == ContentType.AUDIO) {
|
|
for (const otherStreamInfo of this.uriToStreamInfosMap_.values()) {
|
|
if (!otherStreamInfo.loadedOnce && otherStreamInfo.type == type) {
|
|
// To aid manifest filtering, assume before loading that all video
|
|
// renditions have the same MIME type. (And likewise for audio.)
|
|
otherStreamInfo.stream.mimeType = realStream.mimeType;
|
|
}
|
|
}
|
|
}
|
|
|
|
if (type == ContentType.TEXT) {
|
|
const firstSegment = realStream.segmentIndex.get(0);
|
|
if (firstSegment && firstSegment.initSegmentReference) {
|
|
stream.mimeType = 'application/mp4';
|
|
}
|
|
}
|
|
|
|
// Add finishing touches to the stream that can only be done once we have
|
|
// more full context on the media as a whole.
|
|
if (this.hasEnoughInfoToFinalizeStreams_()) {
|
|
if (!this.streamsFinalized_) {
|
|
// Mark this manifest as having been finalized, so we don't go through
|
|
// this whole process of finishing touches a second time.
|
|
this.streamsFinalized_ = true;
|
|
// Finalize all of the currently-loaded streams.
|
|
const streamInfos = Array.from(this.uriToStreamInfosMap_.values());
|
|
const activeStreamInfos =
|
|
streamInfos.filter((s) => s.stream.segmentIndex);
|
|
this.finalizeStreams_(activeStreamInfos);
|
|
// With the addition of this new stream, we now have enough info to
|
|
// figure out how long the streams should be. So process all streams
|
|
// we have downloaded up until this point.
|
|
this.determineDuration_();
|
|
// Finally, start the update timer, if this asset has been determined
|
|
// to be a livestream.
|
|
const delay = this.getUpdatePlaylistDelay_();
|
|
if (delay > 0) {
|
|
this.updatePlaylistTimer_.tickAfter(/* seconds= */ delay);
|
|
}
|
|
} else {
|
|
// We don't need to go through the full process; just finalize this
|
|
// single stream.
|
|
this.finalizeStreams_([streamInfo]);
|
|
}
|
|
}
|
|
};
|
|
|
|
/** @type {Promise} */
|
|
let creationPromise = null;
|
|
/** @type {!AbortController} */
|
|
let abortController = new AbortController();
|
|
const safeCreateSegmentIndex = () => {
|
|
// An operation is already in progress. The second and subsequent
|
|
// callers receive the same Promise as the first caller, and only one
|
|
// download operation will occur.
|
|
if (creationPromise) {
|
|
return creationPromise;
|
|
}
|
|
|
|
// Create a new AbortController to be able to cancel this specific
|
|
// download.
|
|
abortController = new AbortController();
|
|
|
|
// Create a Promise tied to the outcome of downloadSegmentIndex(). If
|
|
// downloadSegmentIndex is rejected, creationPromise will also be
|
|
// rejected.
|
|
creationPromise = new Promise((resolve) => {
|
|
resolve(downloadSegmentIndex(abortController.signal));
|
|
});
|
|
return creationPromise;
|
|
};
|
|
|
|
stream.createSegmentIndex = safeCreateSegmentIndex;
|
|
|
|
stream.closeSegmentIndex = () => {
|
|
// If we're mid-creation, cancel it.
|
|
if (creationPromise && !stream.segmentIndex) {
|
|
abortController.abort();
|
|
}
|
|
// If we have a segment index, release it.
|
|
if (stream.segmentIndex) {
|
|
stream.segmentIndex.release();
|
|
stream.segmentIndex = null;
|
|
}
|
|
// Clear the creation Promise so that a new operation can begin.
|
|
creationPromise = null;
|
|
};
|
|
|
|
return streamInfo;
|
|
}
|
|
|
|
/**
|
|
* @return {number}
|
|
* @private
|
|
*/
|
|
getMinDuration_() {
|
|
let minDuration = Infinity;
|
|
for (const streamInfo of this.uriToStreamInfosMap_.values()) {
|
|
if (streamInfo.stream.segmentIndex && streamInfo.stream.type != 'text') {
|
|
// Since everything is already offset to 0 (either by sync or by being
|
|
// VOD), only maxTimestamp is necessary to compute the duration.
|
|
minDuration = Math.min(minDuration, streamInfo.maxTimestamp);
|
|
}
|
|
}
|
|
return minDuration;
|
|
}
|
|
|
|
/**
|
|
* @return {number}
|
|
* @private
|
|
*/
|
|
getLiveDuration_() {
|
|
let maxTimestamp = Infinity;
|
|
let minTimestamp = Infinity;
|
|
for (const streamInfo of this.uriToStreamInfosMap_.values()) {
|
|
if (streamInfo.stream.segmentIndex && streamInfo.stream.type != 'text') {
|
|
maxTimestamp = Math.min(maxTimestamp, streamInfo.maxTimestamp);
|
|
minTimestamp = Math.min(minTimestamp, streamInfo.minTimestamp);
|
|
}
|
|
}
|
|
return maxTimestamp - minTimestamp;
|
|
}
|
|
|
|
/**
|
|
* @param {!Array.<!shaka.extern.Stream>} streams
|
|
* @private
|
|
*/
|
|
notifySegmentsForStreams_(streams) {
|
|
const references = [];
|
|
for (const stream of streams) {
|
|
if (!stream.segmentIndex) {
|
|
// The stream was closed since the list of streams was built.
|
|
continue;
|
|
}
|
|
stream.segmentIndex.forEachTopLevelReference((reference) => {
|
|
references.push(reference);
|
|
});
|
|
}
|
|
this.presentationTimeline_.notifySegments(references);
|
|
}
|
|
|
|
/**
|
|
* @param {!Array.<!shaka.hls.HlsParser.StreamInfo>} streamInfos
|
|
* @private
|
|
*/
|
|
finalizeStreams_(streamInfos) {
|
|
if (!this.isLive_()) {
|
|
const minDuration = this.getMinDuration_();
|
|
for (const streamInfo of streamInfos) {
|
|
streamInfo.stream.segmentIndex.fit(/* periodStart= */ 0, minDuration);
|
|
}
|
|
}
|
|
this.notifySegmentsForStreams_(streamInfos.map((s) => s.stream));
|
|
if (this.config_.hls.ignoreManifestProgramDateTime) {
|
|
this.syncStreamsWithSequenceNumber_(streamInfos);
|
|
} else {
|
|
this.syncStreamsWithProgramDateTime_(streamInfos);
|
|
}
|
|
}
|
|
|
|
/**
|
|
* There are some values on streams that can only be set once we know about
|
|
* both the video and audio content, if present.
|
|
* This checks if there is at least one video downloaded (if the media has
|
|
* video), and that there is at least one audio downloaded (if the media has
|
|
* audio).
|
|
* @return {boolean}
|
|
* @private
|
|
*/
|
|
hasEnoughInfoToFinalizeStreams_() {
|
|
if (!this.manifest_) {
|
|
return false;
|
|
}
|
|
const videos = [];
|
|
const audios = [];
|
|
for (const variant of this.manifest_.variants) {
|
|
if (variant.video) {
|
|
videos.push(variant.video);
|
|
}
|
|
if (variant.audio) {
|
|
audios.push(variant.audio);
|
|
}
|
|
}
|
|
if (videos.length > 0 && !videos.some((stream) => stream.segmentIndex)) {
|
|
return false;
|
|
}
|
|
if (audios.length > 0 && !audios.some((stream) => stream.segmentIndex)) {
|
|
return false;
|
|
}
|
|
return true;
|
|
}
|
|
|
|
/**
|
|
* @param {!shaka.hls.Playlist} playlist
|
|
* @param {string} verbatimMediaPlaylistUri
|
|
* @param {string} absoluteMediaPlaylistUri
|
|
* @param {string} codecs
|
|
* @param {string} type
|
|
* @param {?string} languageValue
|
|
* @param {boolean} primary
|
|
* @param {?string} name
|
|
* @param {?number} channelsCount
|
|
* @param {Map.<string, string>} closedCaptions
|
|
* @param {?string} characteristics
|
|
* @param {boolean} forced
|
|
* @param {?number} sampleRate
|
|
* @param {boolean} spatialAudio
|
|
* @param {(string|undefined)} mimeType
|
|
* @return {!Promise.<!shaka.hls.HlsParser.StreamInfo>}
|
|
* @private
|
|
*/
|
|
async convertParsedPlaylistIntoStreamInfo_(playlist, verbatimMediaPlaylistUri,
|
|
absoluteMediaPlaylistUri, codecs, type, languageValue, primary, name,
|
|
channelsCount, closedCaptions, characteristics, forced, sampleRate,
|
|
spatialAudio, mimeType = undefined) {
|
|
if (playlist.type != shaka.hls.PlaylistType.MEDIA) {
|
|
// EXT-X-MEDIA and EXT-X-IMAGE-STREAM-INF tags should point to media
|
|
// playlists.
|
|
throw new shaka.util.Error(
|
|
shaka.util.Error.Severity.CRITICAL,
|
|
shaka.util.Error.Category.MANIFEST,
|
|
shaka.util.Error.Code.HLS_INVALID_PLAYLIST_HIERARCHY);
|
|
}
|
|
|
|
/** @type {!Array.<!shaka.hls.Tag>} */
|
|
const variablesTags = shaka.hls.Utils.filterTagsByName(playlist.tags,
|
|
'EXT-X-DEFINE');
|
|
|
|
const mediaVariables = this.parseMediaVariables_(
|
|
variablesTags, absoluteMediaPlaylistUri);
|
|
|
|
goog.asserts.assert(playlist.segments != null,
|
|
'Media playlist should have segments!');
|
|
|
|
this.determinePresentationType_(playlist);
|
|
|
|
if (this.isLive_()) {
|
|
this.determineLastTargetDuration_(playlist);
|
|
}
|
|
|
|
if (!mimeType) {
|
|
mimeType = await this.guessMimeType_(type, codecs, playlist,
|
|
mediaVariables);
|
|
}
|
|
|
|
const {drmInfos, keyIds, encrypted, aesEncrypted} =
|
|
this.parseDrmInfo_(playlist, mimeType);
|
|
|
|
if (encrypted && !drmInfos.length && !aesEncrypted) {
|
|
throw new shaka.util.Error(
|
|
shaka.util.Error.Severity.CRITICAL,
|
|
shaka.util.Error.Category.MANIFEST,
|
|
shaka.util.Error.Code.HLS_KEYFORMATS_NOT_SUPPORTED);
|
|
}
|
|
|
|
const stream = this.makeStreamObject_(codecs, type, languageValue, primary,
|
|
name, channelsCount, closedCaptions, characteristics, forced,
|
|
sampleRate, spatialAudio);
|
|
stream.encrypted = encrypted;
|
|
stream.drmInfos = drmInfos;
|
|
stream.keyIds = keyIds;
|
|
stream.mimeType = mimeType;
|
|
|
|
const mediaSequenceToStartTime = this.isLive_() ?
|
|
this.mediaSequenceToStartTimeByType_.get(type) : new Map();
|
|
|
|
const {segments, bandwidth} = this.createSegments_(
|
|
playlist, stream, mediaSequenceToStartTime, mediaVariables);
|
|
stream.bandwidth = bandwidth;
|
|
|
|
// This new calculation is necessary for Low Latency streams.
|
|
if (this.isLive_()) {
|
|
this.determineLastTargetDuration_(playlist);
|
|
}
|
|
|
|
const firstStartTime = segments[0].startTime;
|
|
const lastSegment = segments[segments.length - 1];
|
|
const lastEndTime = lastSegment.endTime;
|
|
/** @type {!shaka.media.SegmentIndex} */
|
|
const segmentIndex = new shaka.media.SegmentIndex(segments);
|
|
stream.segmentIndex = segmentIndex;
|
|
|
|
const serverControlTag = shaka.hls.Utils.getFirstTagWithName(
|
|
playlist.tags, 'EXT-X-SERVER-CONTROL');
|
|
const canSkipSegments = serverControlTag ?
|
|
serverControlTag.getAttribute('CAN-SKIP-UNTIL') != null : false;
|
|
const canBlockReload = serverControlTag ?
|
|
serverControlTag.getAttribute('CAN-BLOCK-RELOAD') != null : false;
|
|
|
|
const mediaSequenceNumber = shaka.hls.Utils.getFirstTagWithNameAsNumber(
|
|
playlist.tags, 'EXT-X-MEDIA-SEQUENCE', 0);
|
|
|
|
const {nextMediaSequence, nextPart} =
|
|
this.getNextMediaSequenceAndPart_(mediaSequenceNumber, segments);
|
|
|
|
return {
|
|
stream,
|
|
type,
|
|
verbatimMediaPlaylistUri,
|
|
absoluteMediaPlaylistUris: [absoluteMediaPlaylistUri],
|
|
minTimestamp: firstStartTime,
|
|
maxTimestamp: lastEndTime,
|
|
canSkipSegments,
|
|
canBlockReload,
|
|
hasEndList: false,
|
|
firstSequenceNumber: -1,
|
|
nextMediaSequence,
|
|
nextPart,
|
|
mediaSequenceToStartTime,
|
|
loadedOnce: false,
|
|
};
|
|
}
|
|
|
|
/**
|
|
* Get the next msn and part
|
|
*
|
|
* @param {number} mediaSequenceNumber
|
|
* @param {!Array.<!shaka.media.SegmentReference>} segments
|
|
* @return {{nextMediaSequence: number, nextPart:number}}}
|
|
* @private
|
|
*/
|
|
getNextMediaSequenceAndPart_(mediaSequenceNumber, segments) {
|
|
const currentMediaSequence = mediaSequenceNumber + segments.length - 1;
|
|
let nextMediaSequence = currentMediaSequence;
|
|
let nextPart = -1;
|
|
if (!segments.length) {
|
|
nextMediaSequence++;
|
|
return {
|
|
nextMediaSequence,
|
|
nextPart,
|
|
};
|
|
}
|
|
const lastSegment = segments[segments.length - 1];
|
|
const partialReferences = lastSegment.partialReferences;
|
|
if (!lastSegment.partialReferences.length) {
|
|
nextMediaSequence++;
|
|
if (lastSegment.hasByterangeOptimization()) {
|
|
nextPart = 0;
|
|
}
|
|
return {
|
|
nextMediaSequence,
|
|
nextPart,
|
|
};
|
|
}
|
|
nextPart = partialReferences.length - 1;
|
|
const lastPartialReference =
|
|
partialReferences[partialReferences.length - 1];
|
|
if (!lastPartialReference.isPreload()) {
|
|
nextMediaSequence++;
|
|
nextPart = 0;
|
|
}
|
|
return {
|
|
nextMediaSequence,
|
|
nextPart,
|
|
};
|
|
}
|
|
|
|
|
|
/**
|
|
* Creates a stream object with the given parameters.
|
|
* The parameters that are passed into here are only the things that can be
|
|
* known without downloading the media playlist; other values must be set
|
|
* manually on the object after creation.
|
|
* @param {string} codecs
|
|
* @param {string} type
|
|
* @param {?string} languageValue
|
|
* @param {boolean} primary
|
|
* @param {?string} name
|
|
* @param {?number} channelsCount
|
|
* @param {Map.<string, string>} closedCaptions
|
|
* @param {?string} characteristics
|
|
* @param {boolean} forced
|
|
* @param {?number} sampleRate
|
|
* @param {boolean} spatialAudio
|
|
* @return {!shaka.extern.Stream}
|
|
* @private
|
|
*/
|
|
makeStreamObject_(codecs, type, languageValue, primary, name, channelsCount,
|
|
closedCaptions, characteristics, forced, sampleRate, spatialAudio) {
|
|
// Fill out a "best-guess" mimeType, for now. It will be replaced once the
|
|
// stream is lazy-loaded.
|
|
const mimeType = this.guessMimeTypeBeforeLoading_(type, codecs) ||
|
|
this.guessMimeTypeFallback_(type);
|
|
const roles = [];
|
|
if (characteristics) {
|
|
for (const characteristic of characteristics.split(',')) {
|
|
roles.push(characteristic);
|
|
}
|
|
}
|
|
|
|
let kind = undefined;
|
|
if (type == shaka.util.ManifestParserUtils.ContentType.TEXT) {
|
|
if (roles.includes('public.accessibility.transcribes-spoken-dialog') &&
|
|
roles.includes('public.accessibility.describes-music-and-sound')) {
|
|
kind = shaka.util.ManifestParserUtils.TextStreamKind.CLOSED_CAPTION;
|
|
} else {
|
|
kind = shaka.util.ManifestParserUtils.TextStreamKind.SUBTITLE;
|
|
}
|
|
}
|
|
|
|
// If there are no roles, and we have defaulted to the subtitle "kind" for
|
|
// this track, add the implied subtitle role.
|
|
if (!roles.length &&
|
|
kind === shaka.util.ManifestParserUtils.TextStreamKind.SUBTITLE) {
|
|
roles.push(shaka.util.ManifestParserUtils.TextStreamKind.SUBTITLE);
|
|
}
|
|
|
|
return {
|
|
id: this.globalId_++,
|
|
originalId: name,
|
|
groupId: null,
|
|
createSegmentIndex: () => Promise.resolve(),
|
|
segmentIndex: null,
|
|
mimeType,
|
|
codecs,
|
|
kind: (type == shaka.util.ManifestParserUtils.ContentType.TEXT) ?
|
|
shaka.util.ManifestParserUtils.TextStreamKind.SUBTITLE : undefined,
|
|
encrypted: false,
|
|
drmInfos: [],
|
|
keyIds: new Set(),
|
|
language: this.getLanguage_(languageValue),
|
|
originalLanguage: languageValue,
|
|
label: name, // For historical reasons, since before "originalId".
|
|
type,
|
|
primary,
|
|
// TODO: trick mode
|
|
trickModeVideo: null,
|
|
emsgSchemeIdUris: null,
|
|
frameRate: undefined,
|
|
pixelAspectRatio: undefined,
|
|
width: undefined,
|
|
height: undefined,
|
|
bandwidth: undefined,
|
|
roles,
|
|
forced,
|
|
channelsCount,
|
|
audioSamplingRate: sampleRate,
|
|
spatialAudio,
|
|
closedCaptions,
|
|
hdr: undefined,
|
|
videoLayout: undefined,
|
|
tilesLayout: undefined,
|
|
accessibilityPurpose: null,
|
|
external: false,
|
|
};
|
|
}
|
|
|
|
/**
|
|
* @param {!shaka.hls.Playlist} playlist
|
|
* @param {string} mimeType
|
|
* @return {{
|
|
* drmInfos: !Array.<shaka.extern.DrmInfo>,
|
|
* keyIds: !Set.<string>,
|
|
* encrypted: boolean,
|
|
* aesEncrypted: boolean
|
|
* }}
|
|
* @private
|
|
*/
|
|
parseDrmInfo_(playlist, mimeType) {
|
|
/** @type {!Array.<!shaka.hls.Tag>} */
|
|
const drmTags = [];
|
|
if (playlist.segments) {
|
|
for (const segment of playlist.segments) {
|
|
const segmentKeyTags = shaka.hls.Utils.filterTagsByName(segment.tags,
|
|
'EXT-X-KEY');
|
|
drmTags.push(...segmentKeyTags);
|
|
}
|
|
}
|
|
|
|
let encrypted = false;
|
|
let aesEncrypted = false;
|
|
|
|
/** @type {!Array.<shaka.extern.DrmInfo>}*/
|
|
const drmInfos = [];
|
|
const keyIds = new Set();
|
|
|
|
for (const drmTag of drmTags) {
|
|
const method = drmTag.getRequiredAttrValue('METHOD');
|
|
if (method != 'NONE') {
|
|
encrypted = true;
|
|
|
|
if (method == 'AES-128') {
|
|
// These keys are handled separately.
|
|
aesEncrypted = true;
|
|
} else {
|
|
// According to the HLS spec, KEYFORMAT is optional and implicitly
|
|
// defaults to "identity".
|
|
// https://datatracker.ietf.org/doc/html/draft-pantos-hls-rfc8216bis-11#section-4.4.4.4
|
|
const keyFormat =
|
|
drmTag.getAttributeValue('KEYFORMAT') || 'identity';
|
|
const drmParser =
|
|
shaka.hls.HlsParser.KEYFORMATS_TO_DRM_PARSERS_[keyFormat];
|
|
|
|
const drmInfo = drmParser ? drmParser(drmTag, mimeType) : null;
|
|
if (drmInfo) {
|
|
if (drmInfo.keyIds) {
|
|
for (const keyId of drmInfo.keyIds) {
|
|
keyIds.add(keyId);
|
|
}
|
|
}
|
|
drmInfos.push(drmInfo);
|
|
} else {
|
|
shaka.log.warning('Unsupported HLS KEYFORMAT', keyFormat);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
return {drmInfos, keyIds, encrypted, aesEncrypted};
|
|
}
|
|
|
|
/**
|
|
* @param {!shaka.hls.Tag} drmTag
|
|
* @param {!shaka.hls.Playlist} playlist
|
|
* @param {?Map.<string, string>=} variables
|
|
* @return {!shaka.extern.aes128Key}
|
|
* @private
|
|
*/
|
|
parseAES128DrmTag_(drmTag, playlist, variables) {
|
|
// Check if the Web Crypto API is available.
|
|
if (!window.crypto || !window.crypto.subtle) {
|
|
shaka.log.alwaysWarn('Web Crypto API is not available to decrypt ' +
|
|
'AES-128. (Web Crypto only exists in secure origins like https)');
|
|
throw new shaka.util.Error(
|
|
shaka.util.Error.Severity.CRITICAL,
|
|
shaka.util.Error.Category.MANIFEST,
|
|
shaka.util.Error.Code.NO_WEB_CRYPTO_API);
|
|
}
|
|
|
|
// HLS RFC 8216 Section 5.2:
|
|
// An EXT-X-KEY tag with a KEYFORMAT of "identity" that does not have an IV
|
|
// attribute indicates that the Media Sequence Number is to be used as the
|
|
// IV when decrypting a Media Segment, by putting its big-endian binary
|
|
// representation into a 16-octet (128-bit) buffer and padding (on the left)
|
|
// with zeros.
|
|
let firstMediaSequenceNumber = 0;
|
|
let iv;
|
|
const ivHex = drmTag.getAttributeValue('IV', '');
|
|
if (!ivHex) {
|
|
// Media Sequence Number will be used as IV.
|
|
firstMediaSequenceNumber = shaka.hls.Utils.getFirstTagWithNameAsNumber(
|
|
playlist.tags, 'EXT-X-MEDIA-SEQUENCE', 0);
|
|
} else {
|
|
// Exclude 0x at the start of string.
|
|
iv = shaka.util.Uint8ArrayUtils.fromHex(ivHex.substr(2));
|
|
if (iv.byteLength != 16) {
|
|
throw new shaka.util.Error(
|
|
shaka.util.Error.Severity.CRITICAL,
|
|
shaka.util.Error.Category.MANIFEST,
|
|
shaka.util.Error.Code.AES_128_INVALID_IV_LENGTH);
|
|
}
|
|
}
|
|
|
|
const keyUri = shaka.hls.Utils.constructUris(
|
|
[playlist.absoluteUri], drmTag.getRequiredAttrValue('URI'), variables);
|
|
|
|
const requestType = shaka.net.NetworkingEngine.RequestType.KEY;
|
|
const request = shaka.net.NetworkingEngine.makeRequest(
|
|
[keyUri], this.config_.retryParameters);
|
|
|
|
const keyInfo = {method: 'AES-128', iv, firstMediaSequenceNumber};
|
|
|
|
// Don't download the key object until the segment is parsed, to avoid a
|
|
// startup delay for long manifests with lots of keys.
|
|
keyInfo.fetchKey = async () => {
|
|
const keyResponse = await this.makeNetworkRequest_(request, requestType);
|
|
|
|
// keyResponse.status is undefined when URI is "data:text/plain;base64,"
|
|
if (!keyResponse.data || keyResponse.data.byteLength != 16) {
|
|
throw new shaka.util.Error(
|
|
shaka.util.Error.Severity.CRITICAL,
|
|
shaka.util.Error.Category.MANIFEST,
|
|
shaka.util.Error.Code.AES_128_INVALID_KEY_LENGTH);
|
|
}
|
|
|
|
const algorithm = {
|
|
name: 'AES-CBC',
|
|
};
|
|
keyInfo.cryptoKey = await window.crypto.subtle.importKey(
|
|
'raw', keyResponse.data, algorithm, true, ['decrypt']);
|
|
keyInfo.fetchKey = undefined; // No longer needed.
|
|
};
|
|
|
|
return keyInfo;
|
|
}
|
|
|
|
|
|
/**
|
|
* @param {!shaka.hls.Playlist} playlist
|
|
* @private
|
|
*/
|
|
determinePresentationType_(playlist) {
|
|
const PresentationType = shaka.hls.HlsParser.PresentationType_;
|
|
const presentationTypeTag =
|
|
shaka.hls.Utils.getFirstTagWithName(playlist.tags,
|
|
'EXT-X-PLAYLIST-TYPE');
|
|
const endListTag =
|
|
shaka.hls.Utils.getFirstTagWithName(playlist.tags, 'EXT-X-ENDLIST');
|
|
|
|
const isVod = (presentationTypeTag && presentationTypeTag.value == 'VOD') ||
|
|
endListTag;
|
|
const isEvent = presentationTypeTag &&
|
|
presentationTypeTag.value == 'EVENT' && !isVod;
|
|
const isLive = !isVod && !isEvent;
|
|
|
|
if (isVod) {
|
|
this.setPresentationType_(PresentationType.VOD);
|
|
} else {
|
|
// If it's not VOD, it must be presentation type LIVE or an ongoing EVENT.
|
|
if (isLive) {
|
|
this.setPresentationType_(PresentationType.LIVE);
|
|
} else {
|
|
this.setPresentationType_(PresentationType.EVENT);
|
|
}
|
|
}
|
|
}
|
|
|
|
|
|
/**
|
|
* @param {!shaka.hls.Playlist} playlist
|
|
* @private
|
|
*/
|
|
determineLastTargetDuration_(playlist) {
|
|
let lastTargetDuration = Infinity;
|
|
const segments = playlist.segments;
|
|
if (segments.length) {
|
|
let segmentIndex = segments.length - 1;
|
|
while (segmentIndex >= 0) {
|
|
const segment = segments[segmentIndex];
|
|
const extinfTag =
|
|
shaka.hls.Utils.getFirstTagWithName(segment.tags, 'EXTINF');
|
|
if (extinfTag) {
|
|
// The EXTINF tag format is '#EXTINF:<duration>,[<title>]'.
|
|
// We're interested in the duration part.
|
|
const extinfValues = extinfTag.value.split(',');
|
|
lastTargetDuration = Number(extinfValues[0]);
|
|
break;
|
|
}
|
|
segmentIndex--;
|
|
}
|
|
}
|
|
|
|
const targetDurationTag = this.getRequiredTag_(playlist.tags,
|
|
'EXT-X-TARGETDURATION');
|
|
const targetDuration = Number(targetDurationTag.value);
|
|
const partialTargetDurationTag =
|
|
shaka.hls.Utils.getFirstTagWithName(playlist.tags, 'EXT-X-PART-INF');
|
|
if (partialTargetDurationTag) {
|
|
this.partialTargetDuration_ = Number(
|
|
partialTargetDurationTag.getRequiredAttrValue('PART-TARGET'));
|
|
}
|
|
// Get the server-recommended min distance from the live edge.
|
|
const serverControlTag = shaka.hls.Utils.getFirstTagWithName(
|
|
playlist.tags, 'EXT-X-SERVER-CONTROL');
|
|
// According to the HLS spec, updates should not happen more often than
|
|
// once in targetDuration. It also requires us to only update the active
|
|
// variant. We might implement that later, but for now every variant
|
|
// will be updated. To get the update period, choose the smallest
|
|
// targetDuration value across all playlists.
|
|
// 1. Update the shortest one to use as update period and segment
|
|
// availability time (for LIVE).
|
|
if (this.lowLatencyMode_ && this.partialTargetDuration_) {
|
|
// For low latency streaming, use the partial segment target duration.
|
|
if (this.lowLatencyByterangeOptimization_) {
|
|
// We always have at least 1 partial segment part, and most servers
|
|
// allow you to make a request with _HLS_msn=X&_HLS_part=0 with a
|
|
// distance of 4 partial segments. With this we ensure that we
|
|
// obtain the minimum latency in this type of case.
|
|
if (this.partialTargetDuration_ * 5 <= lastTargetDuration) {
|
|
this.lastTargetDuration_ = Math.min(
|
|
this.partialTargetDuration_, this.lastTargetDuration_);
|
|
} else {
|
|
this.lastTargetDuration_ = Math.min(
|
|
lastTargetDuration, this.lastTargetDuration_);
|
|
}
|
|
} else {
|
|
this.lastTargetDuration_ = Math.min(
|
|
this.partialTargetDuration_, this.lastTargetDuration_);
|
|
}
|
|
// Use 'PART-HOLD-BACK' as the presentation delay for low latency mode.
|
|
this.lowLatencyPresentationDelay_ = serverControlTag ? Number(
|
|
serverControlTag.getRequiredAttrValue('PART-HOLD-BACK')) : 0;
|
|
} else {
|
|
this.lastTargetDuration_ = Math.min(
|
|
lastTargetDuration, this.lastTargetDuration_);
|
|
// Use 'HOLD-BACK' as the presentation delay for default if defined.
|
|
const holdBack = serverControlTag ?
|
|
serverControlTag.getAttribute('HOLD-BACK') : null;
|
|
this.presentationDelay_ = holdBack ? Number(holdBack.value) : 0;
|
|
}
|
|
// 2. Update the longest target duration if need be to use as a
|
|
// presentation delay later.
|
|
this.maxTargetDuration_ = Math.max(
|
|
targetDuration, this.maxTargetDuration_);
|
|
}
|
|
|
|
|
|
/**
|
|
* @param {!shaka.hls.Playlist} playlist
|
|
* @private
|
|
*/
|
|
changePresentationTimelineToLive_(playlist) {
|
|
// The live edge will be calculated from segments, so we don't need to
|
|
// set a presentation start time. We will assert later that this is
|
|
// working as expected.
|
|
|
|
// The HLS spec (RFC 8216) states in 6.3.3:
|
|
//
|
|
// "The client SHALL choose which Media Segment to play first ... the
|
|
// client SHOULD NOT choose a segment that starts less than three target
|
|
// durations from the end of the Playlist file. Doing so can trigger
|
|
// playback stalls."
|
|
//
|
|
// We accomplish this in our DASH-y model by setting a presentation
|
|
// delay of configured value, or 3 segments duration if not configured.
|
|
// This will be the "live edge" of the presentation.
|
|
let presentationDelay;
|
|
if (this.config_.defaultPresentationDelay) {
|
|
presentationDelay = this.config_.defaultPresentationDelay;
|
|
} else if (this.lowLatencyPresentationDelay_) {
|
|
presentationDelay = this.lowLatencyPresentationDelay_;
|
|
} else if (this.presentationDelay_) {
|
|
presentationDelay = this.presentationDelay_;
|
|
} else {
|
|
const playlistSegments = playlist.segments.length;
|
|
let delaySegments = this.config_.hls.liveSegmentsDelay;
|
|
if (delaySegments > (playlistSegments - 2)) {
|
|
delaySegments = Math.max(1, playlistSegments - 2);
|
|
}
|
|
presentationDelay = this.maxTargetDuration_ * delaySegments;
|
|
}
|
|
|
|
this.presentationTimeline_.setPresentationStartTime(0);
|
|
this.presentationTimeline_.setDelay(presentationDelay);
|
|
this.presentationTimeline_.setStatic(false);
|
|
}
|
|
|
|
/**
|
|
* Get the InitSegmentReference for a segment if it has a EXT-X-MAP tag.
|
|
* @param {!shaka.hls.Playlist} playlist
|
|
* @param {!Array.<!shaka.hls.Tag>} tags Segment tags
|
|
* @param {?Map.<string, string>=} variables
|
|
* @return {shaka.media.InitSegmentReference}
|
|
* @private
|
|
*/
|
|
getInitSegmentReference_(playlist, tags, variables) {
|
|
/** @type {?shaka.hls.Tag} */
|
|
const mapTag = shaka.hls.Utils.getFirstTagWithName(tags, 'EXT-X-MAP');
|
|
|
|
if (!mapTag) {
|
|
return null;
|
|
}
|
|
// Map tag example: #EXT-X-MAP:URI="main.mp4",BYTERANGE="720@0"
|
|
const verbatimInitSegmentUri = mapTag.getRequiredAttrValue('URI');
|
|
const absoluteInitSegmentUris = shaka.hls.Utils.constructUris(
|
|
[playlist.absoluteUri], verbatimInitSegmentUri, variables);
|
|
|
|
const mapTagKey = [
|
|
absoluteInitSegmentUris.toString(),
|
|
mapTag.getAttributeValue('BYTERANGE', ''),
|
|
].join('-');
|
|
if (!this.mapTagToInitSegmentRefMap_.has(mapTagKey)) {
|
|
/** @type {shaka.extern.aes128Key|undefined} */
|
|
let aes128Key = undefined;
|
|
let byteRangeTag = null;
|
|
for (const tag of tags) {
|
|
if (tag.name == 'EXT-X-KEY') {
|
|
if (tag.getRequiredAttrValue('METHOD') == 'AES-128' &&
|
|
tag.id < mapTag.id) {
|
|
aes128Key = this.parseAES128DrmTag_(tag, playlist, variables);
|
|
}
|
|
} else if (tag.name == 'EXT-X-BYTERANGE' && tag.id < mapTag.id) {
|
|
byteRangeTag = tag;
|
|
}
|
|
}
|
|
const initSegmentRef = this.createInitSegmentReference_(
|
|
absoluteInitSegmentUris, mapTag, byteRangeTag, aes128Key);
|
|
this.mapTagToInitSegmentRefMap_.set(mapTagKey, initSegmentRef);
|
|
}
|
|
return this.mapTagToInitSegmentRefMap_.get(mapTagKey);
|
|
}
|
|
|
|
/**
|
|
* Create an InitSegmentReference object for the EXT-X-MAP tag in the media
|
|
* playlist.
|
|
* @param {!Array.<string>} absoluteInitSegmentUris
|
|
* @param {!shaka.hls.Tag} mapTag EXT-X-MAP
|
|
* @param {shaka.hls.Tag=} byteRangeTag EXT-X-BYTERANGE
|
|
* @param {shaka.extern.aes128Key=} aes128Key
|
|
* @return {!shaka.media.InitSegmentReference}
|
|
* @private
|
|
*/
|
|
createInitSegmentReference_(absoluteInitSegmentUris, mapTag, byteRangeTag,
|
|
aes128Key) {
|
|
let startByte = 0;
|
|
let endByte = null;
|
|
let byterange = mapTag.getAttributeValue('BYTERANGE');
|
|
if (!byterange && byteRangeTag) {
|
|
byterange = byteRangeTag.value;
|
|
}
|
|
// If a BYTERANGE attribute is not specified, the segment consists
|
|
// of the entire resource.
|
|
if (byterange) {
|
|
const blocks = byterange.split('@');
|
|
const byteLength = Number(blocks[0]);
|
|
startByte = Number(blocks[1]);
|
|
endByte = startByte + byteLength - 1;
|
|
|
|
if (aes128Key) {
|
|
// MAP segment encrypted with method 'AES-128', when served with
|
|
// HTTP Range, has the unencrypted size specified in the range.
|
|
// See: https://tools.ietf.org/html/draft-pantos-hls-rfc8216bis-08#section-6.3.6
|
|
const length = (endByte + 1) - startByte;
|
|
if (length % 16) {
|
|
endByte += (16 - (length % 16));
|
|
}
|
|
}
|
|
}
|
|
|
|
const initSegmentRef = new shaka.media.InitSegmentReference(
|
|
() => absoluteInitSegmentUris,
|
|
startByte,
|
|
endByte,
|
|
/* mediaQuality= */ null,
|
|
/* timescale= */ null,
|
|
/* segmentData= */ null,
|
|
aes128Key);
|
|
return initSegmentRef;
|
|
}
|
|
|
|
/**
|
|
* Parses one shaka.hls.Segment object into a shaka.media.SegmentReference.
|
|
*
|
|
* @param {shaka.media.InitSegmentReference} initSegmentReference
|
|
* @param {shaka.media.SegmentReference} previousReference
|
|
* @param {!shaka.hls.Segment} hlsSegment
|
|
* @param {number} startTime
|
|
* @param {!Map.<string, string>} variables
|
|
* @param {!shaka.hls.Playlist} playlist
|
|
* @param {shaka.extern.Stream} stream
|
|
* @param {shaka.extern.aes128Key=} aes128Key
|
|
* @return {shaka.media.SegmentReference}
|
|
* @private
|
|
*/
|
|
createSegmentReference_(
|
|
initSegmentReference, previousReference, hlsSegment, startTime,
|
|
variables, playlist, stream, aes128Key) {
|
|
const tags = hlsSegment.tags;
|
|
const extinfTag =
|
|
shaka.hls.Utils.getFirstTagWithName(tags, 'EXTINF');
|
|
|
|
let endTime = 0;
|
|
let startByte = 0;
|
|
let endByte = null;
|
|
|
|
if (hlsSegment.partialSegments.length && !this.lowLatencyMode_) {
|
|
shaka.log.alwaysWarn('Low-latency HLS live stream detected, but ' +
|
|
'low-latency streaming mode is not enabled in Shaka ' +
|
|
'Player. Set streaming.lowLatencyMode configuration to ' +
|
|
'true, and see https://bit.ly/3clctcj for details.');
|
|
}
|
|
|
|
let syncTime = null;
|
|
if (!this.config_.hls.ignoreManifestProgramDateTime) {
|
|
const dateTimeTag =
|
|
shaka.hls.Utils.getFirstTagWithName(tags, 'EXT-X-PROGRAM-DATE-TIME');
|
|
if (dateTimeTag && dateTimeTag.value) {
|
|
syncTime = shaka.util.XmlUtils.parseDate(dateTimeTag.value);
|
|
goog.asserts.assert(syncTime != null,
|
|
'EXT-X-PROGRAM-DATE-TIME format not valid');
|
|
}
|
|
}
|
|
|
|
let status = shaka.media.SegmentReference.Status.AVAILABLE;
|
|
if (shaka.hls.Utils.getFirstTagWithName(tags, 'EXT-X-GAP')) {
|
|
status = shaka.media.SegmentReference.Status.MISSING;
|
|
}
|
|
|
|
if (!extinfTag) {
|
|
if (hlsSegment.partialSegments.length == 0) {
|
|
// EXTINF tag must be available if the segment has no partial segments.
|
|
throw new shaka.util.Error(
|
|
shaka.util.Error.Severity.CRITICAL,
|
|
shaka.util.Error.Category.MANIFEST,
|
|
shaka.util.Error.Code.HLS_REQUIRED_TAG_MISSING, 'EXTINF');
|
|
} else if (!this.lowLatencyMode_) {
|
|
// Without EXTINF and without low-latency mode, partial segments get
|
|
// ignored.
|
|
return null;
|
|
}
|
|
}
|
|
|
|
// Create SegmentReferences for the partial segments.
|
|
let partialSegmentRefs = [];
|
|
|
|
// Optimization for LL-HLS with byterange
|
|
// More info in https://tinyurl.com/hls-open-byte-range
|
|
let segmentWithByteRangeOptimization = false;
|
|
let getUrisOptimization = null;
|
|
let somePartialSegmentWithGap = false;
|
|
let isPreloadSegment = false;
|
|
|
|
if (this.lowLatencyMode_ && hlsSegment.partialSegments.length) {
|
|
const byterangeOptimizationSupport = (stream.mimeType == 'video/mp4' ||
|
|
stream.mimeType == 'audio/mp4') && window.ReadableStream;
|
|
|
|
let partialSyncTime = syncTime;
|
|
for (let i = 0; i < hlsSegment.partialSegments.length; i++) {
|
|
const item = hlsSegment.partialSegments[i];
|
|
const pPreviousReference = i == 0 ?
|
|
previousReference : partialSegmentRefs[partialSegmentRefs.length - 1];
|
|
const pStartTime = (i == 0) ? startTime : pPreviousReference.endTime;
|
|
|
|
// If DURATION is missing from this partial segment, use the target
|
|
// partial duration from the top of the playlist, which is a required
|
|
// attribute for content with partial segments.
|
|
const pDuration = Number(item.getAttributeValue('DURATION')) ||
|
|
this.partialTargetDuration_;
|
|
|
|
// If for some reason we have neither an explicit duration, nor a target
|
|
// partial duration, we should SKIP this partial segment to avoid
|
|
// duplicating content in the presentation timeline.
|
|
if (!pDuration) {
|
|
continue;
|
|
}
|
|
|
|
const pEndTime = pStartTime + pDuration;
|
|
|
|
let pStartByte = 0;
|
|
let pEndByte = null;
|
|
if (item.name == 'EXT-X-PRELOAD-HINT') {
|
|
// A preload hinted partial segment may have byterange start info.
|
|
const pByterangeStart = item.getAttributeValue('BYTERANGE-START');
|
|
pStartByte = pByterangeStart ? Number(pByterangeStart) : 0;
|
|
// A preload hinted partial segment may have byterange length info.
|
|
const pByterangeLength = item.getAttributeValue('BYTERANGE-LENGTH');
|
|
if (pByterangeLength) {
|
|
pEndByte = pStartByte + Number(pByterangeLength) - 1;
|
|
} else if (pStartByte) {
|
|
// If we have a non-zero start byte, but no end byte, follow the
|
|
// recommendation of https://tinyurl.com/hls-open-byte-range and
|
|
// set the end byte explicitly to a large integer.
|
|
pEndByte = Number.MAX_SAFE_INTEGER;
|
|
}
|
|
} else {
|
|
const pByterange = item.getAttributeValue('BYTERANGE');
|
|
[pStartByte, pEndByte] =
|
|
this.parseByteRange_(pPreviousReference, pByterange);
|
|
}
|
|
const pUri = item.getAttributeValue('URI');
|
|
if (!pUri) {
|
|
continue;
|
|
}
|
|
|
|
let partialStatus = shaka.media.SegmentReference.Status.AVAILABLE;
|
|
if (item.getAttributeValue('GAP') == 'YES') {
|
|
partialStatus = shaka.media.SegmentReference.Status.MISSING;
|
|
somePartialSegmentWithGap = true;
|
|
}
|
|
|
|
let uris = null;
|
|
const getPartialUris = () => {
|
|
if (uris == null) {
|
|
goog.asserts.assert(pUri, 'Partial uri should be defined!');
|
|
uris = shaka.hls.Utils.constructUris(
|
|
[playlist.absoluteUri], pUri, variables);
|
|
}
|
|
return uris;
|
|
};
|
|
|
|
if (byterangeOptimizationSupport &&
|
|
pStartByte >= 0 && pEndByte != null) {
|
|
getUrisOptimization = getPartialUris;
|
|
segmentWithByteRangeOptimization = true;
|
|
}
|
|
|
|
const partial = new shaka.media.SegmentReference(
|
|
pStartTime,
|
|
pEndTime,
|
|
getPartialUris,
|
|
pStartByte,
|
|
pEndByte,
|
|
initSegmentReference,
|
|
/* timestampOffset= */ 0,
|
|
/* appendWindowStart= */ 0,
|
|
/* appendWindowEnd= */ Infinity,
|
|
/* partialReferences= */ [],
|
|
/* tilesLayout= */ '',
|
|
/* tileDuration= */ null,
|
|
partialSyncTime,
|
|
partialStatus,
|
|
aes128Key);
|
|
if (item.name == 'EXT-X-PRELOAD-HINT') {
|
|
partial.markAsPreload();
|
|
isPreloadSegment = true;
|
|
}
|
|
// The spec doesn't say that we can assume INDEPENDENT=YES for the
|
|
// first partial segment. It does call the flag "optional", though, and
|
|
// that cases where there are no such flags on any partial segments, it
|
|
// is sensible to assume the first one is independent.
|
|
if (item.getAttributeValue('INDEPENDENT') != 'YES' && i > 0) {
|
|
partial.markAsNonIndependent();
|
|
}
|
|
partialSegmentRefs.push(partial);
|
|
|
|
if (partialSyncTime) {
|
|
partialSyncTime += pDuration;
|
|
}
|
|
} // for-loop of hlsSegment.partialSegments
|
|
}
|
|
|
|
// If the segment has EXTINF tag, set the segment's end time, start byte
|
|
// and end byte based on the duration and byterange information.
|
|
// Otherwise, calculate the end time, start / end byte based on its partial
|
|
// segments.
|
|
// Note that the sum of partial segments durations may be slightly different
|
|
// from the parent segment's duration. In this case, use the duration from
|
|
// the parent segment tag.
|
|
if (extinfTag) {
|
|
// The EXTINF tag format is '#EXTINF:<duration>,[<title>]'.
|
|
// We're interested in the duration part.
|
|
const extinfValues = extinfTag.value.split(',');
|
|
const duration = Number(extinfValues[0]);
|
|
// Skip segments without duration
|
|
if (duration == 0) {
|
|
return null;
|
|
}
|
|
endTime = startTime + duration;
|
|
} else if (partialSegmentRefs.length) {
|
|
endTime = partialSegmentRefs[partialSegmentRefs.length - 1].endTime;
|
|
} else {
|
|
// Skip segments without duration and without partialsegments
|
|
return null;
|
|
}
|
|
|
|
if (segmentWithByteRangeOptimization) {
|
|
// We cannot optimize segments with gaps, or with a start byte that is
|
|
// not 0.
|
|
if (somePartialSegmentWithGap || partialSegmentRefs[0].startByte != 0) {
|
|
segmentWithByteRangeOptimization = false;
|
|
getUrisOptimization = null;
|
|
} else {
|
|
partialSegmentRefs = [];
|
|
}
|
|
}
|
|
|
|
// If the segment has EXT-X-BYTERANGE tag, set the start byte and end byte
|
|
// base on the byterange information. If segment has no EXT-X-BYTERANGE tag
|
|
// and has partial segments, set the start byte and end byte base on the
|
|
// partial segments.
|
|
const byterangeTag =
|
|
shaka.hls.Utils.getFirstTagWithName(tags, 'EXT-X-BYTERANGE');
|
|
if (byterangeTag) {
|
|
[startByte, endByte] =
|
|
this.parseByteRange_(previousReference, byterangeTag.value);
|
|
} else if (partialSegmentRefs.length) {
|
|
startByte = partialSegmentRefs[0].startByte;
|
|
endByte = partialSegmentRefs[partialSegmentRefs.length - 1].endByte;
|
|
}
|
|
|
|
let tilesLayout = '';
|
|
let tileDuration = null;
|
|
if (stream.type == shaka.util.ManifestParserUtils.ContentType.IMAGE) {
|
|
// By default in HLS the tilesLayout is 1x1
|
|
tilesLayout = '1x1';
|
|
const tilesTag =
|
|
shaka.hls.Utils.getFirstTagWithName(tags, 'EXT-X-TILES');
|
|
if (tilesTag) {
|
|
tilesLayout = tilesTag.getRequiredAttrValue('LAYOUT');
|
|
const duration = tilesTag.getAttributeValue('DURATION');
|
|
if (duration) {
|
|
tileDuration = Number(duration);
|
|
}
|
|
}
|
|
}
|
|
|
|
let uris = null;
|
|
const getUris = () => {
|
|
if (getUrisOptimization) {
|
|
return getUrisOptimization();
|
|
}
|
|
if (uris == null) {
|
|
uris = shaka.hls.Utils.constructUris([playlist.absoluteUri],
|
|
hlsSegment.verbatimSegmentUri, variables);
|
|
}
|
|
return uris || [];
|
|
};
|
|
|
|
const allPartialSegments = partialSegmentRefs.length > 0 &&
|
|
!!hlsSegment.verbatimSegmentUri;
|
|
|
|
const reference = new shaka.media.SegmentReference(
|
|
startTime,
|
|
endTime,
|
|
getUris,
|
|
startByte,
|
|
endByte,
|
|
initSegmentReference,
|
|
/* timestampOffset= */ 0,
|
|
/* appendWindowStart= */ 0,
|
|
/* appendWindowEnd= */ Infinity,
|
|
partialSegmentRefs,
|
|
tilesLayout,
|
|
tileDuration,
|
|
syncTime,
|
|
status,
|
|
aes128Key,
|
|
allPartialSegments,
|
|
);
|
|
|
|
if (segmentWithByteRangeOptimization) {
|
|
this.lowLatencyByterangeOptimization_ = true;
|
|
reference.markAsByterangeOptimization();
|
|
|
|
if (isPreloadSegment) {
|
|
reference.markAsPreload();
|
|
}
|
|
}
|
|
|
|
return reference;
|
|
}
|
|
|
|
|
|
/**
|
|
* Parse the startByte and endByte.
|
|
* @param {shaka.media.SegmentReference} previousReference
|
|
* @param {?string} byterange
|
|
* @return {!Array.<number>} An array with the start byte and end byte.
|
|
* @private
|
|
*/
|
|
parseByteRange_(previousReference, byterange) {
|
|
let startByte = 0;
|
|
let endByte = null;
|
|
// If BYTERANGE is not specified, the segment consists of the entire
|
|
// resource.
|
|
if (byterange) {
|
|
const blocks = byterange.split('@');
|
|
const byteLength = Number(blocks[0]);
|
|
if (blocks[1]) {
|
|
startByte = Number(blocks[1]);
|
|
} else {
|
|
goog.asserts.assert(previousReference,
|
|
'Cannot refer back to previous HLS segment!');
|
|
startByte = previousReference.endByte + 1;
|
|
}
|
|
endByte = startByte + byteLength - 1;
|
|
}
|
|
return [startByte, endByte];
|
|
}
|
|
|
|
/**
|
|
* Parses shaka.hls.Segment objects into shaka.media.SegmentReferences and
|
|
* get the bandwidth necessary for this segments If it's defined in the
|
|
* playlist.
|
|
*
|
|
* @param {!shaka.hls.Playlist} playlist
|
|
* @param {shaka.extern.Stream} stream
|
|
* @param {!Map.<number, number>} mediaSequenceToStartTime
|
|
* @param {!Map.<string, string>} variables
|
|
* @return {{segments: !Array.<!shaka.media.SegmentReference>,
|
|
* bandwidth: (number|undefined)}}
|
|
* @private
|
|
*/
|
|
createSegments_(playlist, stream, mediaSequenceToStartTime, variables) {
|
|
/** @type {Array.<!shaka.hls.Segment>} */
|
|
const hlsSegments = playlist.segments;
|
|
goog.asserts.assert(hlsSegments.length, 'Playlist should have segments!');
|
|
|
|
/** @type {shaka.media.InitSegmentReference} */
|
|
let initSegmentRef;
|
|
|
|
/** @type {shaka.extern.aes128Key|undefined} */
|
|
let aes128Key = undefined;
|
|
|
|
let discontinuitySequence = shaka.hls.Utils.getFirstTagWithNameAsNumber(
|
|
playlist.tags, 'EXT-X-DISCONTINUITY-SEQUENCE', 0);
|
|
const mediaSequenceNumber = shaka.hls.Utils.getFirstTagWithNameAsNumber(
|
|
playlist.tags, 'EXT-X-MEDIA-SEQUENCE', 0);
|
|
const skipTag = shaka.hls.Utils.getFirstTagWithName(
|
|
playlist.tags, 'EXT-X-SKIP');
|
|
const skippedSegments =
|
|
skipTag ? Number(skipTag.getAttributeValue('SKIPPED-SEGMENTS')) : 0;
|
|
let position = mediaSequenceNumber + skippedSegments;
|
|
let firstStartTime = 0;
|
|
|
|
// For live stream, use the cached value in the mediaSequenceToStartTime
|
|
// map if available.
|
|
if (this.isLive_() && mediaSequenceToStartTime.has(position)) {
|
|
firstStartTime = mediaSequenceToStartTime.get(position);
|
|
}
|
|
|
|
/** @type {!Array.<!shaka.media.SegmentReference>} */
|
|
const references = [];
|
|
|
|
let previousReference = null;
|
|
|
|
/** @type {!Array.<{bitrate: number, duration: number}>} */
|
|
const bitrates = [];
|
|
|
|
for (let i = 0; i < hlsSegments.length; i++) {
|
|
const item = hlsSegments[i];
|
|
const startTime =
|
|
(i == 0) ? firstStartTime : previousReference.endTime;
|
|
position = mediaSequenceNumber + skippedSegments + i;
|
|
|
|
const discontinuityTag = shaka.hls.Utils.getFirstTagWithName(
|
|
item.tags, 'EXT-X-DISCONTINUITY');
|
|
if (discontinuityTag) {
|
|
discontinuitySequence++;
|
|
}
|
|
|
|
// Apply new AES-128 tags as you see them, keeping a running total.
|
|
for (const drmTag of item.tags) {
|
|
if (drmTag.name == 'EXT-X-KEY') {
|
|
if (drmTag.getRequiredAttrValue('METHOD') == 'AES-128') {
|
|
aes128Key = this.parseAES128DrmTag_(drmTag, playlist, variables);
|
|
} else {
|
|
aes128Key = undefined;
|
|
}
|
|
}
|
|
}
|
|
|
|
mediaSequenceToStartTime.set(position, startTime);
|
|
|
|
initSegmentRef = this.getInitSegmentReference_(playlist,
|
|
item.tags, variables);
|
|
|
|
// If the stream is low latency and the user has not configured the
|
|
// lowLatencyMode, but if it has been configured to activate the
|
|
// lowLatencyMode if a stream of this type is detected, we automatically
|
|
// activate the lowLatencyMode.
|
|
if (!this.lowLatencyMode_) {
|
|
const autoLowLatencyMode = this.playerInterface_.isAutoLowLatencyMode();
|
|
if (autoLowLatencyMode) {
|
|
this.playerInterface_.enableLowLatencyMode();
|
|
this.lowLatencyMode_ = this.playerInterface_.isLowLatencyMode();
|
|
}
|
|
}
|
|
|
|
const reference = this.createSegmentReference_(
|
|
initSegmentRef,
|
|
previousReference,
|
|
item,
|
|
startTime,
|
|
variables,
|
|
playlist,
|
|
stream,
|
|
aes128Key);
|
|
|
|
if (reference) {
|
|
const bitrate = shaka.hls.Utils.getFirstTagWithNameAsNumber(
|
|
item.tags, 'EXT-X-BITRATE');
|
|
if (bitrate) {
|
|
bitrates.push({
|
|
bitrate,
|
|
duration: reference.endTime - reference.startTime,
|
|
});
|
|
} else if (bitrates.length) {
|
|
// It applies to every segment between it and the next EXT-X-BITRATE,
|
|
// so we use the latest bitrate value
|
|
const prevBitrate = bitrates.pop();
|
|
prevBitrate.duration += reference.endTime - reference.startTime;
|
|
bitrates.push(prevBitrate);
|
|
}
|
|
|
|
previousReference = reference;
|
|
reference.discontinuitySequence = discontinuitySequence;
|
|
|
|
if (this.config_.hls.ignoreManifestProgramDateTime &&
|
|
this.minSequenceNumber_ != null &&
|
|
position < this.minSequenceNumber_) {
|
|
// This segment is ignored as part of our fallback synchronization
|
|
// method.
|
|
} else {
|
|
references.push(reference);
|
|
}
|
|
}
|
|
}
|
|
|
|
let bandwidth = undefined;
|
|
if (bitrates.length) {
|
|
const duration = bitrates.reduce((sum, value) => {
|
|
return sum + value.duration;
|
|
}, 0);
|
|
bandwidth = Math.round(bitrates.reduce((sum, value) => {
|
|
return sum + value.bitrate * value.duration;
|
|
}, 0) / duration * 1000);
|
|
}
|
|
|
|
// If some segments have sync times, but not all, extrapolate the sync
|
|
// times of the ones with none.
|
|
const someSyncTime = references.some((ref) => ref.syncTime != null);
|
|
if (someSyncTime) {
|
|
for (let i = 0; i < references.length; i++) {
|
|
const reference = references[i];
|
|
if (reference.syncTime != null) {
|
|
// No need to extrapolate.
|
|
continue;
|
|
}
|
|
// Find the nearest segment with syncTime, in either direction.
|
|
// This looks forward and backward simultaneously, keeping track of what
|
|
// to offset the syncTime it finds by as it goes.
|
|
let forwardAdd = 0;
|
|
let forwardI = i;
|
|
/**
|
|
* Look forwards one reference at a time, summing all durations as we
|
|
* go, until we find a reference with a syncTime to use as a basis.
|
|
* This DOES count the original reference, but DOESN'T count the first
|
|
* reference with a syncTime (as we approach it from behind).
|
|
* @return {?number}
|
|
*/
|
|
const lookForward = () => {
|
|
const other = references[forwardI];
|
|
if (other) {
|
|
if (other.syncTime != null) {
|
|
return other.syncTime + forwardAdd;
|
|
}
|
|
forwardAdd -= other.endTime - other.startTime;
|
|
forwardI += 1;
|
|
}
|
|
return null;
|
|
};
|
|
let backwardAdd = 0;
|
|
let backwardI = i;
|
|
/**
|
|
* Look backwards one reference at a time, summing all durations as we
|
|
* go, until we find a reference with a syncTime to use as a basis.
|
|
* This DOESN'T count the original reference, but DOES count the first
|
|
* reference with a syncTime (as we approach it from ahead).
|
|
* @return {?number}
|
|
*/
|
|
const lookBackward = () => {
|
|
const other = references[backwardI];
|
|
if (other) {
|
|
if (other != reference) {
|
|
backwardAdd += other.endTime - other.startTime;
|
|
}
|
|
if (other.syncTime != null) {
|
|
return other.syncTime + backwardAdd;
|
|
}
|
|
backwardI -= 1;
|
|
}
|
|
return null;
|
|
};
|
|
while (reference.syncTime == null) {
|
|
reference.syncTime = lookBackward();
|
|
if (reference.syncTime == null) {
|
|
reference.syncTime = lookForward();
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
// Split the sync times properly among partial segments.
|
|
if (someSyncTime) {
|
|
for (const reference of references) {
|
|
let syncTime = reference.syncTime;
|
|
for (const partial of reference.partialReferences) {
|
|
partial.syncTime = syncTime;
|
|
syncTime += partial.endTime - partial.startTime;
|
|
}
|
|
}
|
|
}
|
|
|
|
// lowestSyncTime is a value from a previous playlist update. Use it to
|
|
// set reference start times. If this is the first playlist parse, we will
|
|
// skip this step, and wait until we have sync time across stream types.
|
|
const lowestSyncTime = this.lowestSyncTime_;
|
|
if (someSyncTime && lowestSyncTime != Infinity) {
|
|
for (const reference of references) {
|
|
reference.syncAgainst(lowestSyncTime);
|
|
}
|
|
}
|
|
|
|
return {
|
|
segments: references,
|
|
bandwidth,
|
|
};
|
|
}
|
|
|
|
/**
|
|
* Attempts to guess stream's mime type based on content type and URI.
|
|
*
|
|
* @param {string} contentType
|
|
* @param {string} codecs
|
|
* @return {?string}
|
|
* @private
|
|
*/
|
|
guessMimeTypeBeforeLoading_(contentType, codecs) {
|
|
if (contentType == shaka.util.ManifestParserUtils.ContentType.TEXT) {
|
|
if (codecs == 'vtt' || codecs == 'wvtt') {
|
|
// If codecs is 'vtt', it's WebVTT.
|
|
return 'text/vtt';
|
|
} else if (codecs && codecs !== '') {
|
|
// Otherwise, assume MP4-embedded text, since text-based formats tend
|
|
// not to have a codecs string at all.
|
|
return 'application/mp4';
|
|
}
|
|
}
|
|
|
|
if (contentType == shaka.util.ManifestParserUtils.ContentType.IMAGE) {
|
|
if (!codecs || codecs == 'jpeg') {
|
|
return 'image/jpeg';
|
|
}
|
|
}
|
|
|
|
if (contentType == shaka.util.ManifestParserUtils.ContentType.AUDIO) {
|
|
// See: https://bugs.chromium.org/p/chromium/issues/detail?id=489520
|
|
if (codecs == 'mp4a.40.34') {
|
|
return 'audio/mpeg';
|
|
}
|
|
}
|
|
|
|
// Not enough information to guess from the content type and codecs.
|
|
return null;
|
|
}
|
|
|
|
/**
|
|
* Get a fallback mime type for the content. Used if all the better methods
|
|
* for determining the mime type have failed.
|
|
*
|
|
* @param {string} contentType
|
|
* @return {string}
|
|
* @private
|
|
*/
|
|
guessMimeTypeFallback_(contentType) {
|
|
if (contentType == shaka.util.ManifestParserUtils.ContentType.TEXT) {
|
|
// If there was no codecs string and no content-type, assume HLS text
|
|
// streams are WebVTT.
|
|
return 'text/vtt';
|
|
}
|
|
// If the HLS content is lacking in both MIME type metadata and
|
|
// segment file extensions, we fall back to assuming it's MP4.
|
|
const map = shaka.hls.HlsParser.EXTENSION_MAP_BY_CONTENT_TYPE_[contentType];
|
|
return map['mp4'];
|
|
}
|
|
|
|
/**
|
|
* Attempts to guess stream's mime type based on content type, URI, and
|
|
* contents of the playlist.
|
|
*
|
|
* @param {string} contentType
|
|
* @param {string} codecs
|
|
* @param {!shaka.hls.Playlist} playlist
|
|
* @param {!Map.<string, string>} variables
|
|
* @return {!Promise.<string>}
|
|
* @private
|
|
*/
|
|
async guessMimeType_(contentType, codecs, playlist, variables) {
|
|
const HlsParser = shaka.hls.HlsParser;
|
|
const requestType = shaka.net.NetworkingEngine.RequestType.SEGMENT;
|
|
|
|
// If you wait long enough, requesting the first segment can fail
|
|
// because it has fallen off the left edge of DVR, so to be safer,
|
|
// let's request the middle segment.
|
|
goog.asserts.assert(playlist.segments.length,
|
|
'Playlist should have segments!');
|
|
const middleSegmentIdx = Math.trunc((playlist.segments.length - 1) / 2);
|
|
|
|
const middleSegmentUris = shaka.hls.Utils.constructUris(
|
|
[playlist.absoluteUri],
|
|
playlist.segments[middleSegmentIdx].verbatimSegmentUri,
|
|
variables);
|
|
|
|
const parsedUri = new goog.Uri(middleSegmentUris[0]);
|
|
const extension = parsedUri.getPath().split('.').pop();
|
|
const map = HlsParser.EXTENSION_MAP_BY_CONTENT_TYPE_[contentType];
|
|
|
|
let mimeType = map[extension];
|
|
if (mimeType) {
|
|
return mimeType;
|
|
}
|
|
|
|
mimeType = HlsParser.RAW_FORMATS_TO_MIME_TYPES_[extension];
|
|
if (mimeType) {
|
|
return mimeType;
|
|
}
|
|
|
|
// The extension map didn't work, so guess based on codecs.
|
|
mimeType = this.guessMimeTypeBeforeLoading_(contentType, codecs);
|
|
if (mimeType) {
|
|
return mimeType;
|
|
}
|
|
|
|
// If unable to guess mime type, request a segment and try getting it
|
|
// from the response.
|
|
const headRequest = shaka.net.NetworkingEngine.makeRequest(
|
|
middleSegmentUris, this.config_.retryParameters);
|
|
headRequest.method = 'HEAD';
|
|
const type = shaka.net.NetworkingEngine.AdvancedRequestType.MEDIA_SEGMENT;
|
|
const response = await this.makeNetworkRequest_(
|
|
headRequest, requestType, {type});
|
|
|
|
const contentMimeType = response.headers['content-type'];
|
|
if (contentMimeType) {
|
|
// Split the MIME type in case the server sent additional parameters.
|
|
return contentMimeType.split(';')[0];
|
|
}
|
|
|
|
return this.guessMimeTypeFallback_(contentType);
|
|
}
|
|
|
|
/**
|
|
* Returns a tag with a given name.
|
|
* Throws an error if tag was not found.
|
|
*
|
|
* @param {!Array.<shaka.hls.Tag>} tags
|
|
* @param {string} tagName
|
|
* @return {!shaka.hls.Tag}
|
|
* @private
|
|
*/
|
|
getRequiredTag_(tags, tagName) {
|
|
const tag = shaka.hls.Utils.getFirstTagWithName(tags, tagName);
|
|
if (!tag) {
|
|
throw new shaka.util.Error(
|
|
shaka.util.Error.Severity.CRITICAL,
|
|
shaka.util.Error.Category.MANIFEST,
|
|
shaka.util.Error.Code.HLS_REQUIRED_TAG_MISSING, tagName);
|
|
}
|
|
|
|
return tag;
|
|
}
|
|
|
|
/**
|
|
* @param {shaka.extern.Stream} stream
|
|
* @param {?string} width
|
|
* @param {?string} height
|
|
* @param {?string} frameRate
|
|
* @param {?string} videoRange
|
|
* @param {?string} videoLayout
|
|
* @private
|
|
*/
|
|
addVideoAttributes_(stream, width, height, frameRate, videoRange,
|
|
videoLayout) {
|
|
if (stream) {
|
|
stream.width = Number(width) || undefined;
|
|
stream.height = Number(height) || undefined;
|
|
stream.frameRate = Number(frameRate) || undefined;
|
|
stream.hdr = videoRange || undefined;
|
|
stream.videoLayout = videoLayout || undefined;
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Makes a network request for the manifest and returns a Promise
|
|
* with the resulting data.
|
|
*
|
|
* @param {!Array.<string>} uris
|
|
* @param {boolean=} isPlaylist
|
|
* @return {!Promise.<!shaka.extern.Response>}
|
|
* @private
|
|
*/
|
|
requestManifest_(uris, isPlaylist) {
|
|
const requestType = shaka.net.NetworkingEngine.RequestType.MANIFEST;
|
|
|
|
const request = shaka.net.NetworkingEngine.makeRequest(
|
|
uris, this.config_.retryParameters);
|
|
|
|
const type = isPlaylist ?
|
|
shaka.net.NetworkingEngine.AdvancedRequestType.MEDIA_PLAYLIST :
|
|
shaka.net.NetworkingEngine.AdvancedRequestType.MASTER_PLAYLIST;
|
|
return this.makeNetworkRequest_(request, requestType, {type});
|
|
}
|
|
|
|
/**
|
|
* Called when the update timer ticks. Because parsing a manifest is async,
|
|
* this method is async. To work with this, this method will schedule the next
|
|
* update when it finished instead of using a repeating-start.
|
|
*
|
|
* @return {!Promise}
|
|
* @private
|
|
*/
|
|
async onUpdate_() {
|
|
shaka.log.info('Updating manifest...');
|
|
|
|
goog.asserts.assert(
|
|
this.getUpdatePlaylistDelay_() > 0,
|
|
'We should only call |onUpdate_| when we are suppose to be updating.');
|
|
|
|
// Detect a call to stop()
|
|
if (!this.playerInterface_) {
|
|
return;
|
|
}
|
|
|
|
try {
|
|
const startTime = Date.now();
|
|
await this.update();
|
|
|
|
// Keep track of how long the longest manifest update took.
|
|
const endTime = Date.now();
|
|
|
|
// This may have converted to VOD, in which case we stop updating.
|
|
if (this.isLive_()) {
|
|
const updateDuration = (endTime - startTime) / 1000.0;
|
|
this.averageUpdateDuration_.sample(1, updateDuration);
|
|
const delay = this.getUpdatePlaylistDelay_();
|
|
const finalDelay = Math.max(0,
|
|
delay - this.averageUpdateDuration_.getEstimate());
|
|
this.updatePlaylistTimer_.tickAfter(/* seconds= */ finalDelay);
|
|
}
|
|
} catch (error) {
|
|
// Detect a call to stop() during this.update()
|
|
if (!this.playerInterface_) {
|
|
return;
|
|
}
|
|
|
|
goog.asserts.assert(error instanceof shaka.util.Error,
|
|
'Should only receive a Shaka error');
|
|
|
|
if (this.config_.raiseFatalErrorOnManifestUpdateRequestFailure) {
|
|
this.playerInterface_.onError(error);
|
|
return;
|
|
}
|
|
|
|
// We will retry updating, so override the severity of the error.
|
|
error.severity = shaka.util.Error.Severity.RECOVERABLE;
|
|
this.playerInterface_.onError(error);
|
|
|
|
// Try again very soon.
|
|
this.updatePlaylistTimer_.tickAfter(/* seconds= */ 0.1);
|
|
}
|
|
|
|
// Detect a call to stop()
|
|
if (!this.playerInterface_) {
|
|
return;
|
|
}
|
|
|
|
this.playerInterface_.onManifestUpdated();
|
|
}
|
|
|
|
|
|
/**
|
|
* @return {boolean}
|
|
* @private
|
|
*/
|
|
isLive_() {
|
|
const PresentationType = shaka.hls.HlsParser.PresentationType_;
|
|
return this.presentationType_ != PresentationType.VOD;
|
|
}
|
|
|
|
|
|
/**
|
|
* @return {number}
|
|
* @private
|
|
*/
|
|
getUpdatePlaylistDelay_() {
|
|
// The HLS spec (RFC 8216) states in 6.3.4:
|
|
// "the client MUST wait for at least the target duration before
|
|
// attempting to reload the Playlist file again".
|
|
// For LL-HLS, the server must add a new partial segment to the Playlist
|
|
// every part target duration.
|
|
return this.lastTargetDuration_;
|
|
}
|
|
|
|
|
|
/**
|
|
* @param {shaka.hls.HlsParser.PresentationType_} type
|
|
* @private
|
|
*/
|
|
setPresentationType_(type) {
|
|
this.presentationType_ = type;
|
|
|
|
if (this.presentationTimeline_) {
|
|
this.presentationTimeline_.setStatic(!this.isLive_());
|
|
}
|
|
|
|
// If this manifest is not for live content, then we have no reason to
|
|
// update it.
|
|
if (!this.isLive_()) {
|
|
this.updatePlaylistTimer_.stop();
|
|
}
|
|
}
|
|
|
|
|
|
/**
|
|
* Create a networking request. This will manage the request using the
|
|
* parser's operation manager. If the parser has already been stopped, the
|
|
* request will not be made.
|
|
*
|
|
* @param {shaka.extern.Request} request
|
|
* @param {shaka.net.NetworkingEngine.RequestType} type
|
|
* @param {shaka.extern.RequestContext=} context
|
|
* @return {!Promise.<shaka.extern.Response>}
|
|
* @private
|
|
*/
|
|
makeNetworkRequest_(request, type, context) {
|
|
if (!this.operationManager_) {
|
|
throw new shaka.util.Error(
|
|
shaka.util.Error.Severity.CRITICAL,
|
|
shaka.util.Error.Category.PLAYER,
|
|
shaka.util.Error.Code.OPERATION_ABORTED);
|
|
}
|
|
|
|
const op = this.playerInterface_.networkingEngine.request(
|
|
type, request, context);
|
|
this.operationManager_.manage(op);
|
|
|
|
return op.promise;
|
|
}
|
|
|
|
/**
|
|
* @param {!shaka.hls.Tag} drmTag
|
|
* @param {string} mimeType
|
|
* @return {?shaka.extern.DrmInfo}
|
|
* @private
|
|
*/
|
|
static fairplayDrmParser_(drmTag, mimeType) {
|
|
if (mimeType == 'video/mp2t') {
|
|
throw new shaka.util.Error(
|
|
shaka.util.Error.Severity.CRITICAL,
|
|
shaka.util.Error.Category.MANIFEST,
|
|
shaka.util.Error.Code.HLS_MSE_ENCRYPTED_MP2T_NOT_SUPPORTED);
|
|
}
|
|
|
|
if (shaka.util.Platform.isMediaKeysPolyfilled()) {
|
|
throw new shaka.util.Error(
|
|
shaka.util.Error.Severity.CRITICAL,
|
|
shaka.util.Error.Category.MANIFEST,
|
|
shaka.util.Error.Code
|
|
.HLS_MSE_ENCRYPTED_LEGACY_APPLE_MEDIA_KEYS_NOT_SUPPORTED);
|
|
}
|
|
|
|
/*
|
|
* Even if we're not able to construct initData through the HLS tag, adding
|
|
* a DRMInfo will allow DRM Engine to request a media key system access
|
|
* with the correct keySystem and initDataType
|
|
*/
|
|
const drmInfo = shaka.util.ManifestParserUtils.createDrmInfo(
|
|
'com.apple.fps', [
|
|
{initDataType: 'sinf', initData: new Uint8Array(0), keyId: null},
|
|
]);
|
|
|
|
return drmInfo;
|
|
}
|
|
|
|
/**
|
|
* @param {!shaka.hls.Tag} drmTag
|
|
* @return {?shaka.extern.DrmInfo}
|
|
* @private
|
|
*/
|
|
static widevineDrmParser_(drmTag) {
|
|
const method = drmTag.getRequiredAttrValue('METHOD');
|
|
const VALID_METHODS = ['SAMPLE-AES', 'SAMPLE-AES-CTR'];
|
|
if (!VALID_METHODS.includes(method)) {
|
|
shaka.log.error('Widevine in HLS is only supported with [',
|
|
VALID_METHODS.join(', '), '], not', method);
|
|
return null;
|
|
}
|
|
|
|
const uri = drmTag.getRequiredAttrValue('URI');
|
|
const parsedData = shaka.net.DataUriPlugin.parseRaw(uri);
|
|
|
|
// The data encoded in the URI is a PSSH box to be used as init data.
|
|
const pssh = shaka.util.BufferUtils.toUint8(parsedData.data);
|
|
const drmInfo = shaka.util.ManifestParserUtils.createDrmInfo(
|
|
'com.widevine.alpha', [
|
|
{initDataType: 'cenc', initData: pssh},
|
|
]);
|
|
|
|
const keyId = drmTag.getAttributeValue('KEYID');
|
|
if (keyId) {
|
|
const keyIdLowerCase = keyId.toLowerCase();
|
|
// This value should begin with '0x':
|
|
goog.asserts.assert(
|
|
keyIdLowerCase.startsWith('0x'), 'Incorrect KEYID format!');
|
|
// But the output should not contain the '0x':
|
|
drmInfo.keyIds = new Set([keyIdLowerCase.substr(2)]);
|
|
}
|
|
return drmInfo;
|
|
}
|
|
|
|
/**
|
|
* See: https://docs.microsoft.com/en-us/playready/packaging/mp4-based-formats-supported-by-playready-clients?tabs=case4
|
|
*
|
|
* @param {!shaka.hls.Tag} drmTag
|
|
* @return {?shaka.extern.DrmInfo}
|
|
* @private
|
|
*/
|
|
static playreadyDrmParser_(drmTag) {
|
|
const method = drmTag.getRequiredAttrValue('METHOD');
|
|
const VALID_METHODS = ['SAMPLE-AES', 'SAMPLE-AES-CTR'];
|
|
if (!VALID_METHODS.includes(method)) {
|
|
shaka.log.error('PlayReady in HLS is only supported with [',
|
|
VALID_METHODS.join(', '), '], not', method);
|
|
return null;
|
|
}
|
|
|
|
const uri = drmTag.getRequiredAttrValue('URI');
|
|
const parsedData = shaka.net.DataUriPlugin.parseRaw(uri);
|
|
|
|
// The data encoded in the URI is a PlayReady Pro Object, so we need
|
|
// convert it to pssh.
|
|
const data = shaka.util.BufferUtils.toUint8(parsedData.data);
|
|
const systemId = new Uint8Array([
|
|
0x9a, 0x04, 0xf0, 0x79, 0x98, 0x40, 0x42, 0x86,
|
|
0xab, 0x92, 0xe6, 0x5b, 0xe0, 0x88, 0x5f, 0x95,
|
|
]);
|
|
const keyIds = new Set();
|
|
const psshVersion = 0;
|
|
const pssh =
|
|
shaka.util.Pssh.createPssh(data, systemId, keyIds, psshVersion);
|
|
const drmInfo = shaka.util.ManifestParserUtils.createDrmInfo(
|
|
'com.microsoft.playready', [
|
|
{initDataType: 'cenc', initData: pssh},
|
|
]);
|
|
|
|
return drmInfo;
|
|
}
|
|
|
|
/**
|
|
* See: https://datatracker.ietf.org/doc/html/draft-pantos-hls-rfc8216bis-11#section-5.1
|
|
*
|
|
* @param {!shaka.hls.Tag} drmTag
|
|
* @return {?shaka.extern.DrmInfo}
|
|
* @private
|
|
*/
|
|
static identityDrmParser_(drmTag) {
|
|
const method = drmTag.getRequiredAttrValue('METHOD');
|
|
const VALID_METHODS = ['SAMPLE-AES', 'SAMPLE-AES-CTR'];
|
|
if (!VALID_METHODS.includes(method)) {
|
|
shaka.log.error('Identity (ClearKey) in HLS is only supported with [',
|
|
VALID_METHODS.join(', '), '], not', method);
|
|
return null;
|
|
}
|
|
|
|
// NOTE: The ClearKey CDM requires a key-id to key mapping. HLS doesn't
|
|
// provide a key ID anywhere. So although we could use the 'URI' attribute
|
|
// to fetch the actual 16-byte key, without a key ID, we can't provide this
|
|
// automatically to the ClearKey CDM. Instead, the application will have
|
|
// to use player.configure('drm.clearKeys', { ... }) to provide the key IDs
|
|
// and keys or player.configure('drm.servers.org\.w3\.clearkey', ...) to
|
|
// provide a ClearKey license server URI.
|
|
return shaka.util.ManifestParserUtils.createDrmInfo(
|
|
'org.w3.clearkey', /* initDatas= */ null);
|
|
}
|
|
};
|
|
|
|
|
|
/**
|
|
* @typedef {{
|
|
* stream: !shaka.extern.Stream,
|
|
* type: string,
|
|
* verbatimMediaPlaylistUri: string,
|
|
* absoluteMediaPlaylistUris: !Array.<string>,
|
|
* minTimestamp: number,
|
|
* maxTimestamp: number,
|
|
* mediaSequenceToStartTime: !Map.<number, number>,
|
|
* canSkipSegments: boolean,
|
|
* canBlockReload: boolean,
|
|
* hasEndList: boolean,
|
|
* firstSequenceNumber: number,
|
|
* nextMediaSequence: number,
|
|
* nextPart: number,
|
|
* loadedOnce: boolean
|
|
* }}
|
|
*
|
|
* @description
|
|
* Contains a stream and information about it.
|
|
*
|
|
* @property {!shaka.extern.Stream} stream
|
|
* The Stream itself.
|
|
* @property {string} type
|
|
* The type value. Could be 'video', 'audio', 'text', or 'image'.
|
|
* @property {string} verbatimMediaPlaylistUri
|
|
* The verbatim media playlist URI, as it appeared in the master playlist.
|
|
* This has not been canonicalized into an absolute URI. This gives us a
|
|
* consistent key for this playlist, even if redirects cause us to update
|
|
* from different origins each time.
|
|
* @property {!Array.<string>} absoluteMediaPlaylistUris
|
|
* The absolute media playlist URIs, resolved relative to the master playlist
|
|
* and updated to reflect any redirects.
|
|
* @property {number} minTimestamp
|
|
* The minimum timestamp found in the stream.
|
|
* @property {number} maxTimestamp
|
|
* The maximum timestamp found in the stream.
|
|
* @property {!Map.<number, number>} mediaSequenceToStartTime
|
|
* A map of media sequence numbers to media start times.
|
|
* Only used for VOD content.
|
|
* @property {boolean} canSkipSegments
|
|
* True if the server supports delta playlist updates, and we can send a
|
|
* request for a playlist that can skip older media segments.
|
|
* @property {boolean} canBlockReload
|
|
* True if the server supports blocking playlist reload, and we can send a
|
|
* request for a playlist that can block reload until some segments are
|
|
* present.
|
|
* @property {boolean} hasEndList
|
|
* True if the stream has an EXT-X-ENDLIST tag.
|
|
* @property {number} firstSequenceNumber
|
|
* The sequence number of the first reference. Only calculated if needed.
|
|
* @property {number} nextMediaSequence
|
|
* The next media sequence.
|
|
* @property {number} nextPart
|
|
* The next part.
|
|
* @property {boolean} loadedOnce
|
|
* True if the stream has been loaded at least once.
|
|
*/
|
|
shaka.hls.HlsParser.StreamInfo;
|
|
|
|
|
|
/**
|
|
* @typedef {{
|
|
* audio: !Array.<shaka.hls.HlsParser.StreamInfo>,
|
|
* video: !Array.<shaka.hls.HlsParser.StreamInfo>
|
|
* }}
|
|
*
|
|
* @description Audio and video stream infos.
|
|
* @property {!Array.<shaka.hls.HlsParser.StreamInfo>} audio
|
|
* @property {!Array.<shaka.hls.HlsParser.StreamInfo>} video
|
|
*/
|
|
shaka.hls.HlsParser.StreamInfos;
|
|
|
|
|
|
/**
|
|
* @const {!Object.<string, string>}
|
|
* @private
|
|
*/
|
|
shaka.hls.HlsParser.RAW_FORMATS_TO_MIME_TYPES_ = {
|
|
'aac': 'audio/aac',
|
|
'ac3': 'audio/ac3',
|
|
'ec3': 'audio/ec3',
|
|
'mp3': 'audio/mpeg',
|
|
};
|
|
|
|
|
|
/**
|
|
* @const {!Object.<string, string>}
|
|
* @private
|
|
*/
|
|
shaka.hls.HlsParser.AUDIO_EXTENSIONS_TO_MIME_TYPES_ = {
|
|
'mp4': 'audio/mp4',
|
|
'mp4a': 'audio/mp4',
|
|
'm4s': 'audio/mp4',
|
|
'm4i': 'audio/mp4',
|
|
'm4a': 'audio/mp4',
|
|
'm4f': 'audio/mp4',
|
|
'cmfa': 'audio/mp4',
|
|
// MPEG2-TS also uses video/ for audio: https://bit.ly/TsMse
|
|
'ts': 'video/mp2t',
|
|
'tsa': 'video/mp2t',
|
|
};
|
|
|
|
|
|
/**
|
|
* @const {!Object.<string, string>}
|
|
* @private
|
|
*/
|
|
shaka.hls.HlsParser.VIDEO_EXTENSIONS_TO_MIME_TYPES_ = {
|
|
'mp4': 'video/mp4',
|
|
'mp4v': 'video/mp4',
|
|
'm4s': 'video/mp4',
|
|
'm4i': 'video/mp4',
|
|
'm4v': 'video/mp4',
|
|
'm4f': 'video/mp4',
|
|
'cmfv': 'video/mp4',
|
|
'ts': 'video/mp2t',
|
|
'tsv': 'video/mp2t',
|
|
};
|
|
|
|
|
|
/**
|
|
* @const {!Object.<string, string>}
|
|
* @private
|
|
*/
|
|
shaka.hls.HlsParser.TEXT_EXTENSIONS_TO_MIME_TYPES_ = {
|
|
'mp4': 'application/mp4',
|
|
'm4s': 'application/mp4',
|
|
'm4i': 'application/mp4',
|
|
'm4f': 'application/mp4',
|
|
'cmft': 'application/mp4',
|
|
'vtt': 'text/vtt',
|
|
'webvtt': 'text/vtt',
|
|
'ttml': 'application/ttml+xml',
|
|
};
|
|
|
|
|
|
/**
|
|
* @const {!Object.<string, string>}
|
|
* @private
|
|
*/
|
|
shaka.hls.HlsParser.IMAGE_EXTENSIONS_TO_MIME_TYPES_ = {
|
|
'jpg': 'image/jpeg',
|
|
'png': 'image/png',
|
|
'svg': 'image/svg+xml',
|
|
'webp': 'image/webp',
|
|
'avif': 'image/avif',
|
|
};
|
|
|
|
|
|
/**
|
|
* @const {!Object.<string, !Object.<string, string>>}
|
|
* @private
|
|
*/
|
|
shaka.hls.HlsParser.EXTENSION_MAP_BY_CONTENT_TYPE_ = {
|
|
'audio': shaka.hls.HlsParser.AUDIO_EXTENSIONS_TO_MIME_TYPES_,
|
|
'video': shaka.hls.HlsParser.VIDEO_EXTENSIONS_TO_MIME_TYPES_,
|
|
'text': shaka.hls.HlsParser.TEXT_EXTENSIONS_TO_MIME_TYPES_,
|
|
'image': shaka.hls.HlsParser.IMAGE_EXTENSIONS_TO_MIME_TYPES_,
|
|
};
|
|
|
|
|
|
/**
|
|
* @typedef {function(!shaka.hls.Tag, string):?shaka.extern.DrmInfo}
|
|
* @private
|
|
*/
|
|
shaka.hls.HlsParser.DrmParser_;
|
|
|
|
|
|
/**
|
|
* @const {!Object.<string, shaka.hls.HlsParser.DrmParser_>}
|
|
* @private
|
|
*/
|
|
shaka.hls.HlsParser.KEYFORMATS_TO_DRM_PARSERS_ = {
|
|
'com.apple.streamingkeydelivery':
|
|
shaka.hls.HlsParser.fairplayDrmParser_,
|
|
'urn:uuid:edef8ba9-79d6-4ace-a3c8-27dcd51d21ed':
|
|
shaka.hls.HlsParser.widevineDrmParser_,
|
|
'com.microsoft.playready':
|
|
shaka.hls.HlsParser.playreadyDrmParser_,
|
|
'identity':
|
|
shaka.hls.HlsParser.identityDrmParser_,
|
|
};
|
|
|
|
|
|
/**
|
|
* @enum {string}
|
|
* @private
|
|
*/
|
|
shaka.hls.HlsParser.PresentationType_ = {
|
|
VOD: 'VOD',
|
|
EVENT: 'EVENT',
|
|
LIVE: 'LIVE',
|
|
};
|
|
|
|
shaka.media.ManifestParser.registerParserByMime(
|
|
'application/x-mpegurl', () => new shaka.hls.HlsParser());
|
|
shaka.media.ManifestParser.registerParserByMime(
|
|
'application/vnd.apple.mpegurl', () => new shaka.hls.HlsParser());
|