fix[cache]: replace double dict cache logic

[braddf]

Pull Request

Description

Simplify caching logic by using standard cache with TTL (time-to-live) functionality to handle expiry of cached route responses.

Future notes (for Quartz API, if applicable):

• This in-memory cache still doesn't serve our multi-instance deployment well in production, Redis or similar would be a better solution, but we might be able to scale back to 1 instance in future anyway in the short- to medium-term.
• I'm not convinced about the "default" behaviour of this middleware being the cached data, intuitively I think we should return early if the cached data is found, and then fall back to requesting the data if it's not. Just semantics though.
• Overall, I'm still of the mind that this should be using a well-structured caching library instead of rolling our own, unless we need to do special things that common OS libraries can't – convo for future!

How Has This Been Tested?

• Locally against dev DB, manual API requests and automated using Postman Runner with 20 virtual users making continuous concurrent requests over span of 5 minutes.

If your changes affect data processing, have you plotted any changes? i.e. have you done a quick sanity check?

• Yes

Checklist:

• My code follows OCF's coding style guidelines
• I have performed a self-review of my own code
• I have made corresponding changes to the documentation
• I have added tests that prove my fix is effective or that my feature works
• I have checked my code and corrected any misspellings

[braddf]

Something flakey in the tests due to the cache persisting I think, might want to put some explicit logic in to clear this between test runs, as the monotonic timer that TTLCache uses will be affected by the freezegun functions, which I think actually makes everything more consistent, but throws up things like this!
If we expose the cache and currently_running instances on the returned wrapper, we can clear them in a pytest.fixture to ensure this is clean between tests.

[peterdudfield]

Nice one! I tested it locally too and it does seem to keep the memory much lower.
(I dont really know how, but I think I'm ok with that)