Fix #11686. Trigger a cache update immediately after reading from cache.

[rebornix]

Fixes #11686.

• Pull request represents a single change (i.e. not fixing disparate/unrelated things in a single PR).
• Title summarizes what is changing.
• Has a news entry file (remember to thank yourself!).
• Appropriate comments and documentation strings in the code.
• Has sufficient logging.
• Has telemetry for feature-requests.
• Unit tests & system/integration tests are added/updated.
• Test plan is updated as appropriate.
• package-lock.json has been regenerated by running npm install (if dependencies have changed).

[codecov-commenter]

Codecov Report

Merging #11687 (8d7ed47) into main (ecad7ae) will increase coverage by 0%.
The diff coverage is 100%.

@@           Coverage Diff           @@
##            main   #11687    +/-   ##
=======================================
  Coverage     61%      62%            
=======================================
  Files        483      483            
  Lines      34362    34368     +6     
  Branches    5571     5573     +2     
=======================================
+ Hits       21245    21562   +317     
+ Misses     11018    10740   -278     
+ Partials    2099     2066    -33     

Impacted Files	Coverage Δ
src/kernels/jupyter/finder/remoteKernelFinder.ts	88% <100%> (+7%)	⬆️
src/platform/common/cancellation.ts	62% <0%> (-6%)	⬇️
.../standalone/survey/dataScienceSurveyBanner.node.ts	66% <0%> (-6%)	⬇️
src/kernels/debuggerStartupCodeProvider.ts	100% <0%> (ø)
...c/kernels/execution/cellExecutionMessageHandler.ts	77% <0%> (+<1%)	⬆️
src/kernels/kernel.ts	82% <0%> (+<1%)	⬆️
src/notebooks/controllers/kernelRanking/helpers.ts	57% <0%> (+<1%)	⬆️
src/platform/common/utils/localize.ts	76% <0%> (+<1%)	⬆️
.../kernels/raw/launcher/kernelEnvVarsService.node.ts	78% <0%> (+<1%)	⬆️
src/kernels/kernelDependencyService.node.ts	85% <0%> (+<1%)	⬆️
... and 40 more

[DonJayamanne]

Please could you check the changes here #11692
I think that might be a better way, I didn't want to push changes to your PR
please use my PR as a suggestion.

I think its better to trigger the change event closer to where the values are updated,
the problem today is the getFromCache updates the cached property when i think that's incorrect, as it should just return as the method implies, else the method should be named initializeFromCache

The reason, I'm rejecting this PR is it feels a little hackish to call updateCache after calling getFromCache & the fact that we're relying on the side effect for the event.

[DonJayamanne]

Appoving the PR as my nit with the current approach doesn't warrant a rejection,
it's merely a code smell

[rebornix]

@DonJayamanne thanks for reviewing the changes. I guess the naming of the functions is a bit confusing but the reason we need one more updateCache is not that we rely on its side effect, it's ensuring that if we initially loadCache successfully, making sure that we trigger a real kernel find. It's possible that we found something in the cache, and there is no relevant emitted to trigger a research, we never do the real kernel find before users run a cell. Changes #11692 here are good but they don't fix the particular bug.