refactor: batch paykit contact cleanup

[ben-kaufman]

Description

This PR builds on #1098 to batch and harden private Paykit contact cleanup:

• Reads one shared SDK linked-peer snapshot per multi-contact publication or cleanup pass, with one batch-level retry instead of per-contact fallback reads.
• Clears receiver-scoped payment lists, drains outbound private messages, and checks pending status once for the whole batch.
• Processes all pending deleted contacts through the collection cleanup path instead of repeating snapshots and drains per contact.
• Serializes publication and cleanup with the same mutex and preserves endpoint cache state that changes while remote cleanup is suspended.
• Keeps per-contact failure isolation for attributable clear failures: successful contacts are cleared while failed contacts remain cached for retry.
• Drops malformed persisted cleanup keys with a warning because they cannot address SDK state, allowing cleanup to converge.
• Batches deleted-contact marker updates into one DataStore write.

Error-path behavior is intentionally conservative:

• If linked-peer inspection fails twice, cleanup still clears locally tracked published paths, but retains all affected contacts for retry.
• If the shared post-drain inspection fails, the whole batch remains cached and pending because delivery status cannot be attributed safely per contact.

Preview

N/A — repository-only refactor.

QA Notes

Manual Tests

N/A

Automated Checks

• PrivatePaykitRepoTest.kt: all 43 focused tests passed, including malformed-state recovery, shared snapshot retry, batched deleted-contact retry, drain-inspection failure, partial cleanup failure, and a concurrent cache update during remote cleanup.
• ./gradlew compileDevDebugKotlin passed as part of the focused and full test runs.
• ./gradlew testDevDebugUnitTest passed.
• ./gradlew detekt passed with only pre-existing findings outside this change.
• git diff --check passed.

[greptile-apps]

Greptile Summary

This PR batches private Paykit contact publication and cleanup work. The main changes are:

• Reads linked peers once per multi-contact pass.
• Clears payment lists before one shared message drain.
• Preserves cached state for contacts with reported cleanup failures.
• Adds tests for batching and partial failure handling.

Confidence Score: 4/5

The batched cleanup path can skip remote cleanup or block unrelated contacts, so these cases should be fixed before merging.

• Invalid keys can be treated as successfully cleaned and lose their retry marker.
• One linked-peer lookup failure now prevents cleanup for the whole batch.
• A peer change during the pass can leave remote state without matching cached cleanup state.

app/src/main/java/to/bitkit/repositories/PrivatePaykitRepo.kt

Important Files Changed

Filename	Overview
app/src/main/java/to/bitkit/repositories/PrivatePaykitRepo.kt	Adds shared linked-peer snapshots, batched cleanup and draining, and per-contact cache reconciliation.
app/src/test/java/to/bitkit/repositories/PrivatePaykitRepoTest.kt	Adds valid tests for one linked-peer read, one shared drain, and partial cleanup failure.

_{Reviews (1): Last reviewed commit: "refactor: batch paykit contact cleanup" | Re-trigger Greptile}

[jvsena42]

Reviewed the batching refactor against the parallel iOS findings on synonymdev/bitkit-ios#638.

The structure is a clear improvement over the base, and the per-contact failure isolation contract does hold — I traced clear-failure, drain-still-pending, and link-inspection-failure, and each ends with the contact in failedPublicKeys, cache retained, retry marker intact. Test coverage is also meaningfully ahead of the iOS PR, which ships none. ./gradlew testDevDebugUnitTest --tests to.bitkit.repositories.PrivatePaykitRepoTest passes locally.

iOS parity summary

iOS #638 finding	Android status
greptile P1 — shared lookup failure drops cleanup + retries	✅ already fixed in 0b91ae2 (ahead of the iOS fix)
N+1 linkedPeers() on snapshot failure	⚠️ present
duplicated snapshot-or-fallback block	⚠️ partly — helper exists but isn't reused in cleanup
clears now attempted after link-inspection failure	⚠️ present, undocumented
drain-check failure blast radius = whole batch	⚠️ present
widened reentrancy window on the deferred state wipe	⚠️ present in analogous form
eager vs deferred failure markers	n/a — Android markers are caller-driven

Requesting changes on one item: the invalid-key handling added in 0b91ae2 turns a previously self-healing no-op into a permanently non-converging failure that blocks profile-delete, disable-sharing, and wipe cleanup for all contacts. Details inline. Everything else is either an iOS-parity item worth resolving consistently across platforms, or a cleanup.

Description nits: "no user-facing behavior changes are intended" isn't quite accurate — the error path now issues clearPrivatePaymentList calls it previously skipped, and transient-failure granularity is coarser. Both worth a bullet.

Changelog: correctly omitted for a refactor: with no user-facing change.

[ben-kaufman]

Restacked onto the current #1098 head after the base rewrite. The cleanup review-fix commit previously referenced as b6689414e is now signed commit cd4cd1ec7; the cleanup diff is unchanged.

The current build failure is the same inherited #1084 WatchOnlyAccountStore.serializedExtendedPubkey compile failure already present on #1084 and #1098.

[ovitrif]

Approved. The batched cleanup keeps per-contact clear isolation, drops malformed pending keys without failing the pass, and preserves cache state that changes while remote cleanup is suspended.