Implement fallback cache keys

What does this MR do?

Implements the runner logic necessary to support multiple fallback keys for cache retrieval. I assume that merging and releasing this MR needs to be coordinated with the linked MR in the gitlab repo

Why was this MR needed?

Adds the backend implementation to support fallback_keys field in the cache configuration in CI.

What's the best way to test this MR?

2 new unit tests have been added:

• TestAbstractShell_extractCacheWithMultipleFallbackKeys_AllKeys confirms that all fallback keys are attempted in order
• TestAbstractShell_extractCacheWithMultipleFallbackKeys_FirstFallbackKeyRemovedDueToProtection confirms that the first fallback key is skipped due to having the protected suffix

What are the relevant issue numbers?

Relates to gitlab!110467 (merged)

Relates to gitlab#22213 (closed)