- Apply the initial scale to overview while it is in the backround app
state to ensure the side pages are positioned accordingly. Since this
affects the computation of the target rects for the overview
animation, also ensure that we calculate those with the final target
overview scale applied.
Bug: 117439562
Change-Id: I5e00cf1683e31b22cf937b966d52f8cbe745aaba
> getBitmapInfo() is called on background thead which blocks
on the main thread to get InvariantDeviceProfile
> getDefaultIcon() can be called on the main thread, which
would get blocked on a previous call of getBitmapInfo()
Bug: 117588621
Change-Id: I6549398c7e0a49284fe59950965495074ab8cdd0
- Skip handling loading visible task data after resetting overview, we
currently keep the tasks bound to their task views (some calls are not
synchronous with the lifecycle of recents, can be changed later), so
the next call to loadVisibleTasks before the tasks have been loaded would
load snapshots for old tasks
- Skip loading tasks if the previously requested load plan is still valid,
this can happen because setCurrentTask() will be called multiple times
during swipe up
Bug: 117603579
Change-Id: Ie58ded14dedbb3934d08163ca7939d89310c3ecf
- Add libs/plugin_core.jar
- Include plugin_core in Launcher3 build (it is already present other
builds as part of the updated shared lib)
- Add PluginEnablerImpl that uses SharedPrefs to enable/disable plugin
components
- Add src_plugins, where plugin interfaces will live. It has a build
rule to create a jar that plugin projects will depend on.
- Copy PluginPreferencesFragment from sysui but using our implementation
for PluginEnabler
Bug: 115877296
Change-Id: I3db54677eaceb10f92018c0f9d18920ad9ffac39
1. Make waitXXX methods fail if the condition diesn’t turn true.
2. Waiting for loading to complete in tearDown instead of reloading the
model
3. Avoiding waiting for load-complete where loading didn’t start
4. Disabling last test in AddConfigWidgetTest
5. Waiting for loading to complete inside setupAndVerifyContents(), not
outside
6. Unifying how we wait for loader to complete
7. Adding more logging
Bug: 117332845
Test: running all Nexus tests
Change-Id: I3070e1ac2b9161179cc3e0800b0cd8162807389a
> Items ids were already being typecasted to int when being bound on the UI
> Using a consistent type allow better use of platform data-structures
> Adding IntArray and IntSet as a replacement for various Collection classes
Change-Id: Id3c650ed2420c2bfca3bd7671d2b705b56112371
This is the first step in adding a flag toggler UI to launcher.
The change migrates a single flag (QSB_ON_FIRST_SCREEN) from a
boolean constant to a boolean method. In future, that will allow
us to return different values at runtime.
Bug: 117223984
Change-Id: I1e62c91dd941b8145166021bc0aa157733e62ea0
It started to work locally in condition, I believe, identical to those
on ATP.
Bug: 112899690
Test: adb shell am instrument -e size large -w -r -e debug false com.google.android.apps.nexuslauncher.tests/androidx.test.runner.AndroidJUnitRunner
Change-Id: I2decb2142d15e025402c0c32b7b161dd28d102bb
This is to prevent tests from affecting tests running after them.
Bug: 117332845
Test: Ran all tests; will see on TAP
Change-Id: Ie93ef30477eb069c6b4df48c6fceaceb968edaf4
This allows running in-proc tests without enabling test harness on
Android and restarting Launcher.
Bug: 110103162
Test: TaplTests
Change-Id: Ife44eff4760eb1349a10847efe2c1f27643bf146
Bug: 117157432
Bug: 116744159
Context: In vary rare case, Pointer id can be higher than
MotionEvent.getPointerCount. When I was fabricating ACTION_POINTER_DOWN
event, I was using pointer id instead of index to create MotionEvent.
This resulted in stack trace as seen in the bug.
Change-Id: I802c2a45a274f0dda5984f1d266334fc8b6b0bea
Hyunyoung Song