memory leak in ModelSyncedEventEmitter

[lawmicha]

Also I found a small leak in ModelSyncedEventEmitter, you can find a patch for this here:
memoryLeak.patch

This memory leak might be better tracked by another issue but as the code to easily reproduce it is here, I let you organise this as you want.

Originally posted by @gbitaudeau in #3259 (comment)

[github-actions]

This issue was opened by a maintainer of this repository; updates will be posted here. If you are also experiencing this issue, please comment here with any relevant information so that we're aware and can prioritize accordingly.

[lawmicha]

diff --git a/AmplifyPlugins/DataStore/Sources/AWSDataStorePlugin/Sync/InitialSync/ModelSyncedEventEmitter.swift b/AmplifyPlugins/DataStore/Sources/AWSDataStorePlugin/Sync/InitialSync/ModelSyncedEventEmitter.swift
index 9a44dbbc4..9bd806267 100644
--- a/AmplifyPlugins/DataStore/Sources/AWSDataStorePlugin/Sync/InitialSync/ModelSyncedEventEmitter.swift
+++ b/AmplifyPlugins/DataStore/Sources/AWSDataStorePlugin/Sync/InitialSync/ModelSyncedEventEmitter.swift
@@ -65,7 +65,7 @@ final class ModelSyncedEventEmitter {
         self.syncOrchestratorSink = initialSyncOrchestrator?
             .publisher
             .receive(on: queue)
-            .filter(filterSyncOperationEvent(_:))
+            .filter { [weak self] in self?.filterSyncOperationEvent($0) ?? false }
             .sink(receiveCompletion: { _ in },
                   receiveValue: { [weak self] value in
                     self?.onReceiveSyncOperationEvent(value: value)
@@ -74,7 +74,7 @@ final class ModelSyncedEventEmitter {
         self.reconciliationQueueSink = reconciliationQueue?
             .publisher
             .receive(on: queue)
-            .filter(filterReconciliationQueueEvent(_:))
+            .filter { [weak self] in self?.filterReconciliationQueueEvent($0) ?? false }
             .sink(receiveCompletion: { _ in },
                   receiveValue: { [weak self] value in
                     self?.onReceiveReconciliationEvent(value: value)

[5d]

Confirmed the memory leak does exist, and the proposed patch addresses and resolves the issue. Will created a PR for review and run tests.

[5d]

The patch has been released in version 2.33.4. Closing this ticket.

[github-actions]

This issue is now closed. Comments on closed issues are hard for our team to see.
If you need more assistance, please open a new issue that references this one.
If you wish to keep having a conversation with other community members under this issue feel free to do so.