* DISPLAY_GAMMA_CALIBRATION
* LONG_TERM_ORBITS
* SERIAL_NUMBER
* While we're at it, cleanup tests for removal of other features too
Change-Id: I1a572cf3bcc5298d208a3dc234f9d82da1105bf0
* The preferences constraints system uses reflection to look up
fields via strings, we need to ensure they are not optimized
out by Proguard. Alternatively we can manually enumerate them,
but it's yet another list of constants to maintain. We should
have used a proper enumeration to begin with, but so it goes.
* The @VisibleForTesting annotation is honored by the platform for
ensuring that Proguard doesn't stomp fields. The support library
has an @Keep annotation, but it's only supported by Gradle. If
it becomes supported by the platform in the future, switch to
it instead.
* Also add tests for all of this.
Change-Id: I9f324e003e2a9de5604085b978d5f95bdd098d32
* Added a proper API for this thing
* Pass objects around with necessary metadata, pre-filtered and sorted
* Added event log and service dump
* Simplified locking and adjusted for performance / correctness
* Clarified/beautified code where possible
* TODO:
- Write profile descriptions
- Finish the UI
- Clean up javadocs
- Do something with per-app profiles, maybe
Change-Id: Ie2ef84031e81a18e95dcd62392c7e6939860d933
In the case that the bitmap being passed in was not a bitmap drawable,
we were not retaining any of the attributes from the original drawable.
This patch ensures we ask that drawable to draw on the canvas/bitmap so we
can use that information. Also add tests around it.
Change-Id: I3eefba6e6624fe0bed4965ddf9029320c40f7420
This patch completes the ThemeManagerTest and thus brings the themes
package to 100% test coverage! \o/
Change-Id: I3bdb41781f127c9554c83f56503d591371031a23
TICKET: CYNGNOS-3029
This fixes the assmption that a device will have a default theme
declared in CMSettings, which is not always the case. If no default
theme package is defined in CMSettings we use ThemeConfig.SYSTEM_DEFAULT
instead.
Change-Id: If7952ea984eea0accf3d0990871fb11e3ccc373a
Sometimes the runnable posted to runOnServiceThread() might run a bit
faster than the calls made to the mocked weather provider, which causes
Mockito to throw an exception. Add a timeout to instruct Mockito to
wait for a bit before it verifies if the method was actually called.
Also removed an unused variable in CMWeatherManagerTest
TICKET: CYNGNOS-3043
Change-Id: I53b6541aeb571a6c83d1c26b6723d74c3661e8cc
Java's assert doesn't really do much on android and just returns true, which
makes you feel good about your tests, but its just lying to you. On the long run
its best that you know the real result and fix it, hence switch to JUnit's assert.
CYNGNOS-3042
Change-Id: If7e5982b10abc1f842a52110973566818b1e3bf4
Tests for unique device id are enforced to be implemented correctly
on user builds.
Change-Id: I1c6b24bbf68fe1dce645744f8323c869fdeb9ada
TICKET: CYNGNOS-3026
Since powersave is ignored while the device is plugged in,
and switching pluggedmodes via runtime executing dumpsys commands
is blocked by sepolicy. Ignore the powersave portion of the test.
Change-Id: Ia2b206649a7fc1960cc25d16ffd22379e7c616e3
TICKET: CYNGNOS-2603
Since the API for PerformanceManager returns the number
of profiles supported we can assume what the profiles are
since the HAL provides them in an ordered manner. Thus,
iterate through the size of the number of profiles and verify
each one that's possible.
Change-Id: I87f6d1a847c849bd9e544c1e89a666726c61fe83
TICKET: CYNGNOS-2603
* This is a rework of the session callback API which previously
lived in the framework due to JNI usage. This has been split out
and cleaned up for CMSDK.
* The JNI library lives on the server side, and the app-level
callback has been changed to a protected broadcast. This allows
us to wake up registered services when these events occur.
* Additionally, we support listing all active audio sessions.
* Also brings some JNI love/hate over to CMSDK.
Change-Id: I31c293943474419e3db088bb7ffab75f7440ac0f
