fix(android): mark tombstone and ANR exits as reported when the event is dropped

[markushi]

📜 Description

The last reported marker (last_tombstone_report / last_anr_report) was only written as a side
effect of caching the envelope on disk. An event dropped by beforeSend never gets there, so the
same ApplicationExitInfo was turned into an event again at every app start.

ApplicationExitInfoHistoryDispatcher now writes the marker as well when captureEvent returns
SentryId.EMPTY_ID, through a new ApplicationExitInfoPolicy.markReported(long). The successful
path is unchanged.

💡 Motivation and Context

A discarded crash must stay discarded. This is how signal handler events already behave, because
OutboxSender deletes the outbox file independent of the result of beforeSend.

• resolves: Tombstone events discarded from beforeSend are re-reported on every app launch #5973
• resolves: JAVA-697

💚 How did you test it?

New test in ApplicationExitIntegrationTestBase, so it runs for both TombstoneIntegrationTest
and AnrV2IntegrationTest. It failed before the change and passes now. Full
:sentry-android-core:testReleaseUnitTest is green.

📝 Checklist

• I added GH Issue ID & Linear ID
• I added tests to verify the changes.
• No new PII added or SDK only sends newly added PII if sendDefaultPII is enabled.
• I updated the docs if needed.
• I updated the wizard if needed.
• Review from the native team if needed.
• No breaking change or entry added to the changelog.
• No breaking change for hybrid SDKs or communicated to hybrid SDKs.
• Public API changes reviewed by another Mobile SDK team member or implemented according to the develop docs spec.

🔮 Next steps

TombstonePolicy deletes the matching native outbox file before the capture, so native data cannot
come back if the merged event is lost. That is a separate defect.

[linear-code]

JAVA-697

[sentry]

📲 Install Builds

Android

🔗 App Name	App ID	Version	Configuration
SDK Size	io.sentry.tests.size	8.53.0 (1)	release

⚙️ sentry-android Build Distribution Settings

[sentry]

Bug: The code marks an exit event as reported if captureEvent returns SentryId.EMPTY_ID. This ID is also returned for transient I/O errors, causing events to be permanently lost instead of retried.
_{Severity: HIGH}

Suggested Fix

The captureEvent method should provide a way to distinguish between intentional drops and transient failures. For example, it could throw a specific exception for I/O errors instead of returning SentryId.EMPTY_ID. The calling code in ApplicationExitInfoHistoryDispatcher should then catch this exception and avoid calling policy.markReported(), allowing the event to be retried later.

Prompt for AI Agent

Review the code at the location below. A potential bug has been identified by an AI
agent. Verify if this is a real issue. If it is, propose a fix; if not, explain why it's
not valid.

Location:
sentry-android-core/src/main/java/io/sentry/android/core/ApplicationExitInfoHistoryDispatcher.java#L188-L195

Potential issue: The `captureEvent` method returns `SentryId.EMPTY_ID` for both
intentional event drops (e.g., via `beforeSend` or sampling) and transient failures like
an `IOException` during envelope processing. The new logic in
`ApplicationExitInfoHistoryDispatcher` treats any `SentryId.EMPTY_ID` return as an
intentional drop and calls `policy.markReported()`. This means if an application exit
event fails to be captured due to a temporary network or I/O issue, it will be
permanently marked as reported and will not be retried on the next application start,
leading to the loss of critical ANR or crash data.

_{Did we get this right? 👍 / 👎 to inform future reviews.}

[0xadam-brown]

Nice to fix this 👍

Lgtm save for the ambiguity of the EMPTY_ID, as called out by SentryBot here.