We may currently throw if the load completes before we finished preparation.
Don't throw and instead check for this condition in maybeThrowPrepareError.
PiperOrigin-RevId: 237237000
These are thrown for non-IOException encountered during loading. Also, they
are thrown from an unexpected position and retrying is likely to fail anyway
because the Extractor is left in an unrecoverable state.
PiperOrigin-RevId: 237043948
SQLiteException is a runtime exception, which makes it easy to
forget to handle it. This change converts SQLiteExceptions into
a checked exception, which is then handled appropriately.
PiperOrigin-RevId: 237038793
Currently a flags-only DecoderInputBuffer may hold only the end-of-stream
flag. This change makes flags-only buffer read any kind of flag set in
the next sample. If no sample is available RESULT_NOTHING_READ is
returned.
PiperOrigin-RevId: 237027581
nalPrefix.readUnsignedIntToInt() will throw if the unsigned output doesn't fit
in an int. Before this change, in the error case we'd retry reading from after
the NAL prefix position. Throw a ParserException so that reading an obviously
invalid NAL prefix will be treated as a fatal exception by the default load
error handling policy.
PiperOrigin-RevId: 236839690
There are 6 different situations which are now more explicitly covered to
ensure we handle them correctly:
1. Radio button (non-adaptive, single group only)
Selected -> Do nothing.
Unselected -> Replace override with selected track.
2. Check box (non-adaptive, but multi-group allowed)
Selected -> Remove track (and override). If last one, set isDisabled flag.
Unselected -> Replace override with selected track.
3. Check box (adaptive allowed, single or multi group)
Selected -> Remove track. If last, remove override and maybe set isDisabled.
Unselected -> Add track to override.
PiperOrigin-RevId: 236668642
The release needs to post to the MediaSource thread and also needs to post
back to the DownloadHelper thread to prevent sending updates after release has
been called. The point where we release the downloadHelper also needs to
dismiss the dialog in case the creation has already been queued before the
release.
PiperOrigin-RevId: 236652309
This is more in line with "forgetting all previous state" as documented for
resetState=True. Also only apply the current position as initial seek position
in this case.
PiperOrigin-RevId: 236623630
We currently throw if a pending seek position was valid when the user issued
it on the app thread, but can't be resolved on the playback thread because the
timeline changed in the meantime. Throwing in this case seems wrong as the
user could not have known about the issue (and the seek position was actually
valid). Also, in other cases where the currently playing period is no longer
in the new timeline, we gracefully use a subsequent period or transition to
ENDED state instead of throwing. So it seems more consistent to transition to
ENDED state as well.
PiperOrigin-RevId: 236274862
Instead just set the currently clicked item as override. If the already
selected button is clicked, the override is reset with the same value.
PiperOrigin-RevId: 236118429
We currently do not show check boxes if the renderer does not support adaptive
switches between tracks in one group. However, we assume check boxes in the
onClick method.
Use same logic for both methods to be consistent.
PiperOrigin-RevId: 236118372
The flag in ExoPlayer.prepare is documented as keeping the current window
index and window position. We are currently using the current period UID and
period position instead. This causes problems when the media source is changed
but the position is not reset.
Using the initial seek position instead ensures we actually use the window
index and position.
Issue:#5520
PiperOrigin-RevId: 236101024
Adding new error types may cause issues when listeners assume a fixed set of
error types and don't handle arbitrary defaults.
Fixing error handling in one case and improving documentation to make people
aware of the issue.
PiperOrigin-RevId: 236093265
While disabled the renderer does not have non-null stream formats. This means
that setting the operating rate could cause a NullPointerException if there was
a codec for reuse.
Check for being enabled/started before trying to set the operating rate. After
the renderer is enabled it should receive a new input format which will update
the operating rate as needed.
PiperOrigin-RevId: 235494384
If the DefaultAudioSink was reconfigured in a way that was compatible with the
previous configuration just after having been reset, we would try to drain audio
processors despite not having an AudioTrack. This could result in a
NullPointerException if speed adjustment was active.
Fix this behavior by only trying to drain audio processors if we actually have
an AudioTrack.
PiperOrigin-RevId: 235355466
- Expose constructor (package private for now, for tests only)
- Add some tests for cache initialization
- Add some TODOs for handling initialization failure
PiperOrigin-RevId: 235188386
*** Reason for rollback ***
Breaks ExoPlayer's gradle build
*** Original change description ***
Migrate Mockito Matchers.{hamcrestMethods} to MockitoHamcrest.{hamcrestMethods}
All methods accepting a hamcrest Matcher in org.mockito.Matchers have been changed
to only accept an ArgumentMatcher. The corresponding methods are now available on
MockitoHamcrest.
For more information see http://go/mockito-2-lsc
Tested:
TAP --sample for global presubmit queue
http://test/OCL:235033293:BASE:234998256:1550793421835:c2ccbbda
***
PiperOrigin-RevId: 235186705
All methods accepting a hamcrest Matcher in org.mockito.Matchers have been changed
to only accept an ArgumentMatcher. The corresponding methods are now available on
MockitoHamcrest.
For more information see http://go/mockito-2-lsc
Tested:
TAP --sample for global presubmit queue
http://test/OCL:235033293:BASE:234998256:1550793421835:c2ccbbda
PiperOrigin-RevId: 235153912
*** Reason for rollback ***
Reverting as this may break playback on other Amlogic devices and/or playback of non-interlaced content.
*** Original change description ***
Add max video size workaround for Amlogic decoder.
The Amlogic awesome decoder reduces the video size of interlaced videos by half
if the internal configuration isn't force reset with new maximum input size
values. The product of these new values must exceed 1920x1088 to force the
reset.
Issue:#5003
***
PiperOrigin-RevId: 234967314
Currently IMA determines that the overlay is obstructing, even
if it's empty. Register it as friendly, which means we're
assuming that anything the apps puts in it is necessary for
playback.
PiperOrigin-RevId: 234963065
HlsSampleStream#read should return end of stream when
there is no mapping for the sample stream, instead of
nothing read. This allows the player to transition to
ended.
Issue:#5524
PiperOrigin-RevId: 234764027
- Use Cache UID for CacheContentIndex and CacheFileMetadataIndex. This
enables SD card swapping for a single device.
- I'm hopeful of finding a way to get the Cache UID to DefaultDownloadIndex
so we can do the same there.
PiperOrigin-RevId: 234662753
Horizontal orientation doesn't make sense. I thought about
overriding the setter to throw something, but it's probably
not worth the extra lines.
PiperOrigin-RevId: 234632530
This bug affects any playlist that uses initialization segments. In practice,
almost exclusively fragmented mp4 segmented playlists are affected.
The bottom line is that extractors are chosen for reuse after the
initialization segment connection is open. However, reused extractors
do not need re-parsing the init segment, so loading the initialization is
wasteful.
PiperOrigin-RevId: 234479467
TrackSelectionView requires an initialization with a call to init(...).
That's why we shouldn't let the view retain its view hierarchy automatically
as the views won't be backed by data if restored. Instead add a listener
which lets the containing activity/fragment save and restore the state if
needed.
PiperOrigin-RevId: 234152491
This is a temporary fix to prevent NullPointerException exceptions.
Though writing to cache will still fail if the cache folder isn't a real
folder.
There are a few thing we can try:
* The listing might be failing because the cache folder is just created.
We can wait and try again.
* If the cache folder is a regular file, we can try deleting and creating
a folder with the same name.
PiperOrigin-RevId: 234121925
Previously we would get a new AudioTrack and flush all audio processors if any
AudioProcessor needed to be flushed on reconfiguration. This was problematic for
the case of TrimmingAudioProcessor because it could become active or inactive
due to transitioning to a period with gapless metadata or without it (we don't
keep it active all the time because it is wasteful to populate its end buffer
for content that is not gapless).
This change handles the case where we don't need an AudioTrack but
AudioProcessors do need to be flushed. In this case we drain all the audio
processors when next handling data then switch to the new configuration.
This avoids truncation when period transitions change whether
TrimmingAudioProcessor is active but don't require a new AudioTrack, and is also
a step towards draining the AudioTrack when transitioning between periods if we
do need a new AudioTrack.
To do this, it needs to be possible to drain any pending output data from an
AudioProcessor after it's configured to a new format, so this change makes sure
AudioProcessors allow calling playToEndOfStream and getOutput after
reconfiguration and before flush.
PiperOrigin-RevId: 234033552
This cleanup is in preparation for draining audio processors on reconfiguration
to a format that doesn't require a new AudioTrack to be created.
PiperOrigin-RevId: 233990983
Stop encoding/decoding presentation time as part of the message.
What's actually in emsg boxes is a presentation time delta,
which is why it's only 32 bits, and hence why it doesn't handle
large absolute timestamps. We were using this field to hold
absolute timestamps only for the purpose of passing presentation
times from DashManifestParser.parseEvent back to the calling
method. After this change, we return Pair<Long, EventMessage>
instead.
Issue: #5490
PiperOrigin-RevId: 233561731
builder.mActions is marked as @RestrictTo(LIBRARY_GROUP). We'll
just have to put up with notification flicker on KitKat if the
actions change.
PiperOrigin-RevId: 233420076
Switch from passing an ad UI ViewGroup to passing an object that can also
provide information about controls overlays.
Also switch to using a dedicated overlay for ads instead of the overlay frame
layout, which apps have easy access to.
PiperOrigin-RevId: 233393500
Supporting multiple overrides allows to select tracks from multiple groups, if
enabled. As more options are added, the creation of the dialog is moved to a
separate builder class.
PiperOrigin-RevId: 233366282
Imported from GitHub PR https://github.com/google/ExoPlayer/pull/5438
Plus the following changes:
- Only support profile 5 (handling other profiles requires
backward-compatibility changes in the renderer which are left for a later
change.)
- Only add KEY_PROFILE to the codec configuration MediaFormat for Dolby Vision.
- In MediaCodecUtil support all DV profiles that Android has constants for. This
includes ones that are "not supported for new applications". Since we don't
extract these profiles, this is currently only for the benefit of custom
extractors.
- Misc code style fixes and reordering for consistency.
Merge 37878b975c2bc082b0568e21cbe62bfcef97c10d into 67be9e7783
PiperOrigin-RevId: 233066799
Given the change to require setPlayer on AdsLoaders, it seems like a good
opportunity to clean up deprecated ads-related symbols.
PiperOrigin-RevId: 233020171
When SimpleCache uses a CacheFileMetadataIndex, it will be able to avoid
querying file.length() and renaming files, both of which are expensive
operations on some file systems.
PiperOrigin-RevId: 232664255
Now we're reusing the builder, it's necessary to set maxProgress
back to 0 to avoid the progress bar appearing for end state
notifications.
PiperOrigin-RevId: 232520007
DefaultDownloadIndex is currently just forwarding all calls to its
inner class. getDownloadTable() handles initialization, but this
doesn't really seem different to having an initialization method. It
doesn't guarantee initialization happens, since a bad method
implementation could try and access downloadTable directly, just as
a bad method implementation could forget to call the initialization
method. Hence this change removes the indirection.
PiperOrigin-RevId: 232482228
This combines all the different listeners in ExoVideosPlayerV2 into one and
moves the PlaybackLogger class towards ExoPlayer's AnalyticsCollector with
the same purpose.
In the future this allows two things:
1. Gradually move LogginClient implementations to AnalyticsListener to
prevent custom event forwarding.
2. Using ExoPlayer's QoE extension (which is also an AnalyticsListener).
PiperOrigin-RevId: 232321182
When calling releaseSource(), all pending messages will be removed. That means
that all action-on-completion callbacks which are somewhere in flight are
just dropped without being called. This change adds code to keep track of the
current state of each callback to allow all of them being called when the
source is released.
Issue:#5464
PiperOrigin-RevId: 232312528
This change enables transitioning to/from different Storage
implementations, to allow experimentally enabling (and if
necessary, disabling) SQLiteStorage. All that's left to do
is the final wiring to turn it on
PiperOrigin-RevId: 232304458
If there is data after the last samples in the container, we may request
continue loading after the last samples have been read but before the load has
completed. In this situation the buffered position is returned as
Long.MAX_VALUE, which prevents continuing loading, yet the media period is not
treated as fully buffered because its buffered position is not
C.TIME_END_OF_SOURCE.
PiperOrigin-RevId: 231406964
We can keep the reading media period and continue playing if we haven't read
beyond the new duration. Otherwise, we can keep the period, but need to reset
the renderers as we already read too far.
PiperOrigin-RevId: 231406252
It better describes what the class does. More importantly, we've had
inconsistent class names since we added offline support, for which we
added ProgressiveDownloader ("ExtractorDownloader" doesn't make any
sense). We could really do with aligning the names for clarity.
(Sorry)
PiperOrigin-RevId: 231387268
This requires to prepare the media source and the periods in a small helper similar
to the metadata retriever. It also gets rid of the need to have abstract protected
methods to load the manifest, to extract the track groups and to convert to stream keys,
as this can now be handled by the media period.
PiperOrigin-RevId: 231385590
Currently, we remove all variants if none of the stream keys contains any
variants. This causes HlsMediaPeriod to throw exceptions as it expects at least
one variant.
Change it to support master playlists without variants.
PiperOrigin-RevId: 231385547
Converting a single track selection to stream keys is only possible if the output
is independent from other track selections being made.
This is not the case for DASH and HLS embedded track groups which should select the
already selected primary track if possible (and thus needs to know whether a primary
track group is selected).
Also, update the test method to take a period index.
PiperOrigin-RevId: 231385490
The way it is currently, it's very unclear that an operation on the
version table will correctly belong to a transaction in code such as
this, taken from DefaultDownloadIndex:
writableDatabase.beginTransaction();
try {
writableDatabase.execSQL(...);
versionTable.setVersion(...);
writableDatabase.setTransactionSuccessful();
} finally {
writableDatabase.endTransaction();
}
This change explicitly passes the database, to make it obvious that
the operation will really go into the same transaction:
writableDatabase.beginTransaction();
try {
writableDatabase.execSQL(....);
VersionTable.setVersion(writableDatabase, ...);
writableDatabase.setTransactionSuccessful();
} finally {
writableDatabase.endTransaction();
}
PiperOrigin-RevId: 231374933
Float output is only possible from API 21, but the high-res int
to float conversion path was checking for 32-bit PCM not float
output capability.
PiperOrigin-RevId: 231172495
This will allow CachedContentIndex to start using database
storage without us having to change the SimpleCache and
CachedContentIndex constructors to require a Context or a
DatabaseProvider.
PiperOrigin-RevId: 230884501
If are allowing changing durations of periods, we shouldn't use the end position
of clipped content as part of the id as it may change. This change moves the end
position back to MediaPeriodInfo and adds the next ad group index to the id
instead to ensure we still have unique ids for all content parts.
PiperOrigin-RevId: 230878389
The combination of pre-16 API levels accounting for ~0.5% of the device
population, and that the most important components in ExoPlayer (e.g.
the MediaCodec renderers) have always required API level 16, mean it's
very unlikely this will negatively impact on anyone.
PiperOrigin-RevId: 230701808
We need to support serialization to/from an SQLite table. The
model of passing something around for each class to write into
doesn't work well for SQL, and it would be messy to have two
different structural designs for serialization. This change
centralizes the logic in CachedContentIndex, where a centralized
SQL based version can more easily sit alongside it.
PiperOrigin-RevId: 230692291
This method is a generalization of the existing hasReadStreamToEnd. It is
useful to determine whether a renderer already read beyond a new duration of
a period.
PiperOrigin-RevId: 230689165
1. The method kept track of the current period index to check if the next
period is still in the correct period. This is unneccessary since we no longer
use the period index but the actual uid in MediaPeriodId and mismatches are
already detected by canKeepMediaPeriodHolder.
2. We updated the MediaPeriodIndfo twice: once in getFollowingMediaPeriodInfo
and once in getUpdatedMediaPeriodInfo. That's confusing and difficult to
follow. The only difference is that getUpdatedMediaPeriodInfo keeps the
content position while getFollowingMediaPeriodInfo resets it. This is made more
explicit for readability.
3. The durations compatibility check for all following periods was broken as
it compared the same durations (partly due to the confusion caused by 2.)
PiperOrigin-RevId: 230519295
* Add AC-4 MIME type definition
* Add AC-4 format support in Mp4Extractor and TsExtractor
* Add AC-4 Extractor
* Add AC-4 playback support in MPEG-4, MPEG-DASH, TS and HLS
ExoPlaybackExceptions of type SOURCE are always associated with the loading
period and thus we can use the event time for the loading period in
onPlayerError. Renderer and unexpected exceptions are still associated with the
currently playing period.
Issue:#5407
PiperOrigin-RevId: 230216253
The Amlogic awesome decoder reduces the video size of interlaced videos by half
if the internal configuration isn't force reset with new maximum input size
values. The product of these new values must exceed 1920x1088 to force the
reset.
Issue:#5003
PiperOrigin-RevId: 230206675
That's the same position set in MediaPeriod.prepare (where it may be removed
in the future).
Having the position at an earlier point is necessary to fix an
issue with lazy preparation in ConcatenatingMediaSource where the prepare
position was assumed to be known but MediaPeriod.prepare hasn't been called
yet.
Issue:#5350
PiperOrigin-RevId: 229756637
[Problem]
CC1 is shown even when CC2-3-4 is selected
[Solution]
* use incoming channel and frame information
* Misc Codes need to handle use Frame information
* Add checks everywhere
[Test]
* Live channels and already existing test should not show any difference
in behavior when CC1 is selected.
* Live channels might have CC2-CC3-CC4
* Sarnoff test content has specific test cases for channel support
When the extra adaptation set of a switch group isn't defined in the manifest, we
currently assume it's the first adaptation group. This either leads to wrong grouping
or duplicate track groups.
Such a case may easily happen if the manifest is filtered such that only one of the
switch adaptation sets will be present in the manifest.
PiperOrigin-RevId: 229365379
The buffered position is currently based on the mimimum queued timestamp of
all AV tracks. If the tracks have unequal lengths, one track continues loading
without bounds as the "buffered position" will always stay at the shorter
track's duration.
This change adds an optional buffer flag to mark the last sample of the
stream. This is set in the Mp4Extractor only so far. ExtractorMediaSource
uses this flag to ignore AV streams in the buffered duration calculation if
they already finished loading.
Issue:#3670
PiperOrigin-RevId: 229359899
That was previously handled by the player. But since we switched to Handler
messages instead of player messages, we should do that manually.
PiperOrigin-RevId: 229341747
This is not necessary as the track selection needs to be updated with
updateSelectedTrack anyway. It's also error-prone as the selection code
calls into a protected method of a not fully initialized class.
PiperOrigin-RevId: 229331669
DataSpec.FLAG_ALLOW_CACHE_FRAGMENTATION is added to indicate to the
cache when fragmentation is allowed. This flag is set for progressive
requests only.
To avoid breaking changes, CacheDataSink defaults to ignoring the flag
(and enabling fragmentation) for now. Respecting the flag can be
enabled manually. DownloaderConstructorHelper enables respecting of
the flag.
Issue: #4253
PiperOrigin-RevId: 229176835
We currently forget whether a source is seekable at re-preparation. This was
implemented intentionally this way under the assumption that we really can't seek
until we have loaded the seek map again. However, seek operations are only
allowed after a media period is prepared. So there is no harm in remembering
whether a source is seekable.
This problem currently prevents reusing ClippingMediaSources with
ExtractorMediaSource and a non-zero start clip position.
Issue: #5351
PiperOrigin-RevId: 229169441
ExoMediaCrypto.requiresSecureDecoderComponent() is removed, and
FrameworkMediaCrypto.forceAllowInsecureDecoderComponents is made
public to allow determining whether a secure decoder is required
to be implemented in MediaCodecRenderer.
PiperOrigin-RevId: 228909771
There were some edge cases in which we'd forget to release DRM
sessions. For example if we read a format and acquired a
pendingDrmSession (in onInputFormatChanged), then immediately
read another format and overwrote pendingDrmSession, we'd
forget to release the one that's been overwritten.
This change hopefully makes release much clearer. We keep a list
of all drm sessions we're currently holding. Whenever we update
either drmSession or pendingDrmSession, we release any other
sessions that are in the list.
PiperOrigin-RevId: 228905465
The window object is used without being filled with data. This used to work
well for most cases as the same live stream is sending regular updates and the
first update is almost never used if it's not the first item in a playlist.
It causes problems when the first timeline update of a live stream is actually
used for playback (e.g. when the live stream is lazily prepared in a playlist
and played first).
PiperOrigin-RevId: 228530232
Reported in https://github.com/google/ExoPlayer/issues/3860
For failing examples see the github link above.
[Problem]
We drop matching control codes even if they are not received on
consecutive frames.
The specification says
"(4) If the first transmission of a control code pair passes parity,
it is acted upon within one video frame. If the NEXT frame contains
a perfect repeat of the same pair, the redundant code is ignored."
Keyword is the NEXT. The frames must arrive immediately after
each other.
See https://www.law.cornell.edu/cfr/text/47/79.101
[Solution]
Set an additional flag when any data is processed. Control code
duplication checks should be limited only for the first control
byte pairs processed after any control code.
[Test]
Sarnoff tests have equivalent CEA708 and CEA608 Streams.
This is the initialization part of mitigating issue #4253. The
remaining work is on the writing side, and is simply a case of
having startFile return File instances that are sharded into
sub-directories. We still need to decide what scheme we want
to use for doing that.
Issue: #4253
PiperOrigin-RevId: 228306327
Calls to File.length() can be O(N) where N is the number of files
in the containing folder. This is believed to be true for at least
FAT32. Repeated calls for the same file tend to be faster,
presumably due to caching in the file system, however are still
surprisingly expensive. Hence minimizing the number of calls is
preferable.
Issue: #4253
PiperOrigin-RevId: 228179921
