fix: Flasky-tests #20413

[aadityarshah]

Purpose / Description

While working on my PR #20412, I faced failure in JUnitTests (macos run 1) and on reviewing it, I observed that no changes I had made during any of my commits affected the files in which errors were faced and on further reviewing, I realized that these were triggered due to the specific development environment (Linux/Windows dual-boot and non-UTC TimeZone. Hence, I raised the issue #20413. This PR aims to resolve the issue.

Fixes

Fixes #20413

Approach

This change addresses the following three categories of instabilities (all of those reported in the issue #20413):

• ResourceID #0x0: Added a filter in the file PrefsSearchBarTest.kt in the directory AnkiDroid/src/test/java/com/ichi2/anki/preferences to ignore non-XML indexed items (IDs of 0), which were causing Resources$NotFoundException.
• TimeZone Determinism: Modified MockTime.timestamp in the MockTime.kt inside directory common/src/main/java/com/ichi2/anki/common/time to use a forced GMT calendar and updated AlarmManagerServiceTest.kt inside the directory AnkiDroid/src/test/java/com/ichi2/anki/services/AlarmManagerServiceTest.kt to set a UTC default during execution which prevents assertions from failing in a non-UTC TimeZone.
• Race Conditions: Replaced advanceRobolectricLooper() with ShadowLooper.runUiThreadTasksIncludingDelayedTasks() in ReviewerTest.kt wherever required and added advanceRobolectricLooper() where required in DeckPickerTest.kt, both of the files present in the directory AnkiDroid/src/test/java/com/ichi2/anki. This ensures the full asynchronous chain (Collection -> Scheduler -> WebView) completes before assertions run.

How Has This Been Tested?

• Used OpenJDK version 24 to run ./gradlew testPlayDebugUnitTest which ran successfully on implementing the following changes allowing me to conclude that the changes were made successfully.

Learning (optional, can help others)

• AI Disclosure: I used AI as a debugging consultant to help interpret the stack traces and understand Robolectric looper function. It helped me fix the issues in Race Conditions. I implemented and reviewed the changes manually ensuring there are no unnecessary changes in the code.
• Insights: I learned about how virtual time works with Robolectric, as well as the need for proper looper flushes, especially with WebView-based activities like the Reviewer. I also learned more about how JNI binaries are handled by the build system, depending on the host OS architecture.

Checklist

Please, go through these checks before submitting the PR.

• You have a descriptive commit message with a short title (first line, max 50 chars).
• You have commented your code, particularly in hard-to-understand areas
• You have performed a self-review of your own code
• UI changes: include screenshots of all affected screens (in particular showing any new or changed strings)
• UI Changes: You have tested your change using the Google Accessibility Scanner

[david-allison]

Won't this fail at the same time in UTC?

[david-allison]

This seems to be a no-op

[david-allison]

Why is there sometimes a resource of 0?

It's best to figure this out, rather than ignore something in the test which we don't understand.

Is this 0 a missing resource, or something which was incorrectly loaded?

I think the best option would be to add a log of the count + names of all selected resource IDs, then we know if it's extra or missing.

[david-allison]

Please revert all these spacing changes, they make it difficult to understand the change

[david-allison]

These appear to be equivalent

        fun advanceRobolectricLooper() {
            Shadows.shadowOf(Looper.getMainLooper()).runToEndOfTasks()
        }

  public static void runUiThreadTasksIncludingDelayedTasks() {
    getShadowMainLooper().runToEndOfTasks();
  }

[aadityarshah]

@david-allison Thank you for your review, I am moving this PR to draft while I work through your feedback and address the necessary changes. I will update you once I am ready and will make the PR open once updates are done.