configurable sessions expiration

[rpassas]

Summary

Original thread here. I didn't refactor SessionCache into an interface, but I'm happy to do so while we're looking at it.

This is meant to be a simple addition to allow the wrapper to create a ValidationService with a configurable SessionCache. That is, when resetExpirationAfterAccess is true, an active session's expiration time resets when it is accessed.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (161e92d) 11.63% compared to head (bdec0db) 11.65%.
Report is 52 commits behind head on master.

Additional details and impacted files

@@             Coverage Diff              @@
##             master    #1501      +/-   ##
============================================
+ Coverage     11.63%   11.65%   +0.01%     
- Complexity    30029    30092      +63     
============================================
  Files          2154     2157       +3     
  Lines        665568   665830     +262     
  Branches     196667   196744      +77     
============================================
+ Hits          77464    77571     +107     
- Misses       559891   560024     +133     
- Partials      28213    28235      +22     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[dotasek]

The implementation seems simple and reasonable, but I'd like to see a few things happen:

1. The new behaviour needs some unit tests. There are some already written that you can probably copy and alter to be appropriate: https://github.com/hapifhir/org.hl7.fhir.core/blob/master/org.hl7.fhir.validation/src/test/java/org/hl7/fhir/validation/cli/services/SessionCacheTest.java

2. I really would like SessionCache to be an interface, and for this to be a PassiveExpiringSessionCache implementation of that.

3. Could we make setExpirationAfterAccess fluent? So we can just call SessionCache sessionCache = new SessionCache(...).setExpirationAfterAccess(true)

[dotasek]

See comments for requested changes.

[dotasek]

Thank you for the refactor @rpassas. There is a failing test on Azure pipelines: https://dev.azure.com/fhir-pipelines/fhir-core-library/_build/results?buildId=16210&view=logs&j=7448d812-dbed-5970-799b-77d1cf827579&t=5ba38598-e9de-532c-87f1-57d0b28805df

I will investigate why that is, and review the test in the process.

I also see that the interface you created has methods that are directly bound to the expiring cache implementation. I'm going to create a branch that separates the two that can be merged into this.

[dotasek]

@rpassas Please check these two branches.

https://github.com/hapifhir/org.hl7.fhir.core/tree/do-20231211-sessionCache-reset-on-fetch
https://github.com/hapifhir/org.hl7.fhir.validator-wrapper/tree/do-20231211-sessionCache-reset-on-fetch

Your test was nearly correct; I repaired it. I made some of the methods have more explicit meaning, and moved some methods to the implementation. Let me know if these changes work for you; I'd like to see them merged into your PRs before merging them to master/main.

[rpassas]

@rpassas Please check these two branches.

https://github.com/hapifhir/org.hl7.fhir.core/tree/do-20231211-sessionCache-reset-on-fetch https://github.com/hapifhir/org.hl7.fhir.validator-wrapper/tree/do-20231211-sessionCache-reset-on-fetch

Your test was nearly correct; I repaired it. I made some of the methods have more explicit meaning, and moved some methods to the implementation. Let me know if these changes work for you; I'd like to see them merged into your PRs before merging them to master/main.

@dotasek I've merged your requested changes (in the wrapper as well). They all look good to me - thanks!