Guard against Banner/Interstitial use of nullptr callbacks

[DellaBitta]

Description

Provide details of the change, and generalize the change in the PR title above.

Check to see if Interstitial and BannerView callback pointers have been nullified before invoking them in Java.

---

Testing

Describe how you've tested these changes. Link any manually triggered Integration tests or CPP binary SDK Packaging Github Action workflows, if applicable.

Ran iTests on Android phone.
iTest Run

---

Type of Change

Place an x the applicable box:

• Bug fix. Add the issue # below if applicable.
• New feature. A non-breaking change which adds functionality.
• Other, such as a build process or documentation change.

---

[DellaBitta]

This wouldn't have done anything, anyway, as the mAdView was already set to null. Also there's a synchronous bounding box changed event immediately prior to this.

[sunmou99]

Do we need to add some logs or throw errors when it's a nullptr?

[DellaBitta]

Do we need to add some logs or throw errors when it's a nullptr?

I don't think so. The most common scenario is starting the process of destroying the C++ object when suddenly a listener is invoked from Android and iOS about movement, clicks, etc. This isn't a misconfiguration just a race condition. Or, say, respond to AdClosed by destroying the object before the final bounding box listener event triggers a -1,-1,-1,-1 bounding box, since AdClosed and the BoundingBox change are different events.

[sunmou99]

Check no pointer erros. LGTM.

[github-actions]

❌  Integration test FAILED

Requested by @DellaBitta on commit 27203e2
Last updated: Thu Nov 18 12:48 PST 2021
View integration test log & download artifacts

Failures	Configs
admob	[TEST] [ERROR] [Android] [All os] [android_target]
auth	[TEST] [ERROR] [Android] [windows] [android_target]
messaging	[TEST] [ERROR] [Android] [macos] [android_target]
remote_config	[TEST] [ERROR] [Android] [windows] [android_target]

Add flaky tests to go/fpl-cpp-flake-tracker