Factory reset

[mikeymike9000]

I'm using Factory along with Firebase/Firestore and wondering how to best reset repositories when a user logs out of Firebase so that no data or listeners remain in my data repository. I'd been running into situations where some data persisted and also crashes due to dangling references, though I could not find any strong references to self and wondered if it might be internal to Firestore.

Object 0x600002e0b660 of class PayloadRepo deallocated with non-zero retain
count 3. This object's deinit, or something called from it, may have created a
strong reference to self which outlived deinit, resulting in a dangling reference.

I'd also been 'blindly' using .reset() (e.g. Container.shared.PayloadRepo.reset()) but I think that may have also been part of the issue.

I now have a kludgey 'Factory reset' that appears to remove all listeners and data, here are relevant code snippets:

extension Container {
  public var auth: Factory { Factory(self) { return Auth.auth() }.singleton } // Firebase Auth
  public var payloadRepo: Factory { self { PayloadRepo() }.singleton } // my data repo
}

---

public class PayloadRepo: ObservableObject {
...
  func unsubscribeAsync() async {
    listenerRegistration?.remove()
    listenerRegistration = nil
    await withCheckedContinuation { continuation in
      DispatchQueue.main.async { [weak self] in
        guard let self = self else { continuation.resume(); return }
        self.payloads = []
        self.user = nil
        continuation.resume()
      }
    }
  }

  deinit {
    unsubscribe(); cancellables.removeAll()
  }

  func unsubscribe() {
    Task { await unsubscribeAsync() }
  }
...
}

---

public class AuthService {
  @Injected(\.auth) private var auth
...
func signOut() async {
  do {
    await Container.shared.companyRepo.resolve().unsubscribeAsync()

    // wait 300ms for Firestore processes to 'settle'
    try await Task.sleep(nanoseconds: 300_000_000)
    await Container.shared.companyRepo.resolve().unsubscribeAsync()

    try await Task.sleep(nanoseconds: 300_000_000)
    try auth.signOut()

    try await Task.sleep(nanoseconds: 300_000_000)
    await Container.shared.companyRepo.resolve().unsubscribeAsync()
  }
}

I plan to replace this 'hack' (delays and repeated unsubscribes) when I can carve out some time to actually understand the inner workings.

Is there a best practice to 'simply' 'clear out' Factories?
Should I also be using .reset()?
Is this hack as horrible as it seems and does anyone have any advice that might expedite my journey?

thanks for any help!! Mike