Changes included:
* 79f2ffe: Fixes required for export.
* e1ca63a: Implement empty schema shouldn't have a version number.
* 8b76be4: Refactor VisibilityStore from a no-op implementation into an interface.
* 47ba533: Remove the max repeatable length limit for GenericDocument
* 52ca287: Add GetSchemaResponse cts test.
* 56585f5: Change PutDocumentsRequestTest to be a cts test.
Bug: 183050495
Bug: 180058203
Bug: 191592792
Bug: 189161227
Bug: 183239766
Test: Presubmit
Change-Id: I30f51a18d697d3a5e43d2c63549ab19a36bbe99e
Callers that don't target S can schedule exact alarms so should get a
return value of true when they call canScheduleExactAlarm.
Test: atest FrameworksMockingServicesTests:AlarmManagerServiceTest
Bug: 191328951
Change-Id: I1cd1d0fb3d3d922360494552e653ed540bfe5227
Included changes:
* a8c984: Add more tests for logging
* 501a51: Remove some deprecated TODOs.
* 639f25: Add SchemaMigrationStats.
* 0b8885: Use consistent terminology for 3p and system access.
* 064c36: Add a CTS test for snippeting window sizes and max match counts.
Bug: 173532925
Bug: 187879464
Bug: 180058203
Test: Presubmit
Change-Id: I878fd6c7a42cccb237898392aac8c615830eb564
Adding this check in a testable way requires refactoring VisibilityStore
to pull it out of management by the AppSearchImpl object. It is now
initialized as a sibling of AppSearchImpl, managed by
AppSearchUserInstanceManager.
This has several benefits:
* Breaks the complicated initialization inter-dependency between
AppSearchImpl and VisibilityStore
* Reduces duplicative singleton managers
* Allows AppSearchImpl to be tested more easily by accepting a
"hasSystemSurfaceable" boolean that can be set in tests
* Reduces the number of times we have to call into VisStore; we can
determine whether the caller has system access in advance instead of
repeating the check for every schema type
Bug: 180058203
Bug: 183031844
Test: GlobalSearchSessionCtsTest#testReportSystemUsage_ForbiddenFromNonSystem
Test: GlobalSearchSessionPlatformCtsTest#testReportSystemUsage
Test: VisibilityStoreTest
Change-Id: I84c84819f287628ccf8af369f5481a8e90255f62
- Create a new XML tag for the new format. The new format should
persist the arrays of values without assuming they necessarily
have different bits.
- Upon reading, if the new tag is present, use that. If the new
tag is not present but the old tag is present, then limit the
contents to the capabilities/transports that existed in R.
- Catch all exceptions and ignore the requests if they can't be
re-read from disk. This is a measure to avoid any mistake where
the device couldn't boot when JS tries to restore state.
Bug: 183071974
Test: atest JobStoreTest
Original-Change: https://android-review.googlesource.com/1679762
Merged-In: I56cde50d2adab81134c8be4f6996d68018f4212a
Change-Id: I56cde50d2adab81134c8be4f6996d68018f4212a
Set explicit alarm window sizes so we can control the maximum alarm
delay for Doze alarms.
Bug: 187947479
Bug: 188468510
Test: atest DeviceIdleTest
Test: atest FrameworksMockingServicesTests:DeviceIdleControllerTest
Change-Id: I40c0e1a0fce18886ea817245d9a064a509eb6e08
Ensure SYSTEM_INTERACTION can bring an app out of the RESTRICTED bucket
(temporarily) if the app is only in the bucket due to timeout.
Bug: 191297958
Test: atest AppStandbyControllerTests
Change-Id: Ie76d224710228a8654d4d57ee0c73bb9441ef28f
Since we cannot access locked user's data and clear it when a package
got removed. When a user is unlocking, we should prune their data and
remove all data that doesn't belong to any of installed packages.
This will also help in handling the case where device gets
rebooted before the data is cleared for an uninstalled package.
Bug: 186151459
Test: atest --rebuild-module-info CtsAppSearchHostTestCases
Change-Id: Ic6d151621a88df7806901781d3edfec65700c35d
Callers that have power exemption should be allowed to set smaller
windows for their alarms.
Test: atest FrameworksMockingServicesTests:AlarmManagerServiceTest
Bug: 185199076
Change-Id: I5f6515a917da7f87891f79fb637d48a3d7862a1b
We need to store other things in the appsearch data dir other than what
Icing stores, such as the storage info cache.
Bug: 177685938
Bug: 181787682
Test: Presubmit
Change-Id: I903c34a2c891d617217d44f5126d0d31c8c6bf39
Non-updatable APEXes are going to be only present on a dm-verity
protected read-only partition. For such APEXes there is no need in
embedding a hashtree.
This is a no-op change in preparation of switching default value of
generate_hashtree to true.
Bug: 190621617
Test: m
Change-Id: I0283fb03908a87490e5e29b803553977b8b6729f
The permission SCHEDULE_EXACT_ALARM state changes at the following
boundaries:
1. App-op: This gets toggled by the user via Settings.
2. Deny-list: Packages can be added to the deny list via DeviceConfig
APIs.
3. Package changes: A package's manifest may changes when it gets
updated.
In both cases 1 and 2, if alarm manager detects a permission change
from revoked to granted, it sends the
ACTION_SCHEDULE_EXACT_ALARM_PERMISSION_STATE_CHANGED broadcast to the
app.
If it detects a permission change from granted to revoked, it kills
all the processes within the hosting uid.
In all three cases, when the permssion changes from granted to revoked,
all the exact alarms scheduled by the relevant package are removed.
Package updates are treated differently as they require processes to
restart anyway. The broadcast is not needed in this case as there
are no alarms expected to have been lost by a previous revocation, and
apps can always use ACTION_MY_PACKAGE_REPLACED for post-update setup as
usual.
All this only applies to packages that have the change
REQUIRE_EXACT_ALARM_PERMISSION enabled.
Also changed canScheduleExactAlarms to return false if the change is not
enabled for the calling package.
Test: atest FrameworksMockingServicesTests:AlarmManagerServiceTest
atest CtsAlarmManagerTestCases
Bug: 179541791
Bug: 187206399
Change-Id: Icd68275701f2804be65b3a10a7dd81bbf6e2a0bb
This was removed in Iff38291b964de4116bcdefe4fd74136c4878c543 in favor
of using the bootclasspath fragment. However the bootclasspath fragment
for media does not contain the API tracking files, so it back for now
to make sure we can finalize the media API files.
This CL may be revertable in the future if we can get the
java_sdk_library into the bootclasspath_fragment somehow.
Bug: 190807367
Bug: 188427719
Test: m media-module-sdk and inspect the zip for api tracking files
Change-Id: Ia01c32d4c6c6bad734df053da0024a139a808a54
Flushing takes a few ms, it seems like a pretty good trade off compared
to keeping an icing instance in memory forever.
Bug: 179390369
Test: presubmit
Change-Id: Ic0493137dc8c33d44bd074bbd9ca4ae442049513
Bug: 189291729
Test: atest ./tests/cts/hostside/src/com/android/cts/net/HostsideRestrictBackgroundNetworkTests.java
BYPASS_INCLUSIVE_LANGUAGE_REASON=Existing public API
Change-Id: I7e26eaef4fe9d7d7b4b3a1740608e0fe821537d2
This subsumes the java_sdk_library in the sdk. Also add the service
java_sdk_library as it was missing.
Bug: 188427719
Test: build_mainline_modules.sh
Change-Id: Iff38291b964de4116bcdefe4fd74136c4878c543
The implementation of handleIncomingUser is copied from
ag/13200660 (1d1547b).
Bug: 181787682
Test: Presubmit
Change-Id: I946199a8844e7fd6ea5574b873da56d8b8630f62
There isn't a need to require that the app has permission on other
user as well. With this change, we are checking the apps existence
on other users to limit access such that apps won't be able to
access any more data than they would've accessed by having
INTERACT_ACROSS_USERS permission and talking to their instances
on other users.
Bug: 175844032
Test: atest --test-mapping apex/blobstore
Change-Id: I6f13e3e7a240d55c02aecf31e10f750b9e4d5702
