[pigeon] Fix race in event channel handler removal integration test - #13528

This PR fixes the tree. Linux_android custom_package_tests master is failing on main (all 3 attempts of #4084) in the Pigeon integration test event channels handle multiple instances:

The following MissingPluginException was thrown while de-activating platform stream on channel
dev.flutter.pigeon.pigeon_integration_tests.EventChannelMethods.streamConsistentNumbers.1:
MissingPluginException(No implementation found for method cancel on channel
dev.flutter.pigeon.pigeon_integration_tests.EventChannelMethods.streamConsistentNumbers.1)

Cause

The test plugin unregisters the stream handler for instance "1" (to exercise event channel unregistration, added in #13438) immediately after calling endOfStream(). On the Dart side, the end of the stream closes the EventChannel's broadcast stream, which then sends a cancel method call back on the same channel. If the handler has already been removed by the time that cancel call arrives, it fails with a MissingPluginException, which fails the test. Whether it fails depends on timing, so it shows up as a flake.

Fix

Unregister the handler from the stream handler's onCancel instead, i.e. only after the cancel call has been received, in both the Kotlin (Android) and Swift (iOS/macOS) test plugins. The generated Pigeon stream handler wrappers already forward the platform onCancel to the user handler, so no generator or generated-code changes are needed. The test still exercises unregistration; it just no longer races with the channel teardown.

This only changes test plugin code under platform_tests/, so there is no change to the published package.

Testing

On an Android emulator (Pixel 9a), running only event channels handle multiple instances:

The full android_kotlin_integration_tests suite (680 tests) also passes with this fix, as do dart_unittests and flutter_unittests. The Swift change mirrors the Kotlin one and is covered by the iOS/macOS integration tests in CI.

Related

flutter/flutter#179154 tracks a different flake in the same test on macOS Swift (Bad state: Cannot add new events after calling close, i.e. an event sent after endOfStream). This PR does not address that one, and the macOS Swift skip is left in place.

Pre-Review Checklist

• I read the Contributor Guide and followed the process outlined there for submitting PRs.
• I read the AI contribution guidelines and understand my responsibilities, or I am not using AI tools.
• I read the Tree Hygiene page, which explains my responsibilities.
• I read and followed the relevant style guides and ran the auto-formatter.
• I signed the CLA.
• The title of the PR starts with the name of the package surrounded by square brackets, e.g. [shared_preferences]
• I linked to at least one issue that this PR fixes in the description above.
• I followed the version and CHANGELOG instructions, using semantic versioning and the repository CHANGELOG style, or I have commented below to indicate which documented exception this PR falls under¹.
• I updated/added any relevant documentation (doc comments with ///).
• I added new tests to check the change I am making, or I have commented below to indicate which test exemption this PR falls under¹.
• All existing and new tests are passing.

If you need help, consider asking for advice on the #hackers-new channel on Discord.