Fix SimpleExemplarReservoir for cumulative

[cijothomas]

Fixes SimplerExemplarReservoir for cumulative temporality to reset internal state during collect. Before this change, newer measurements had no fair chance to make it to the reservoir. This PR resets the "number_of_measurements_seen" with every collect, so new measurements has fair chance to make it to reservoir, subject to the reservoir sampling algorithm.

Spec ref: https://github.com/open-telemetry/opentelemetry-specification/blob/main/specification/metrics/sdk.md#simplefixedsizeexemplarreservoir

Merge requirement checklist

• CONTRIBUTING guidelines followed (license requirements, nullable enabled, static analysis, etc.)
• Unit tests added/updated
• Appropriate CHANGELOG.md files updated for non-trivial changes
• Changes in public API reviewed (if applicable)

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (6250307) 83.38% compared to head (345325e) 83.12%.
Report is 15 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #5230      +/-   ##
==========================================
- Coverage   83.38%   83.12%   -0.27%     
==========================================
  Files         297      271      -26     
  Lines       12531    11970     -561     
==========================================
- Hits        10449     9950     -499     
+ Misses       2082     2020      -62     

Flag	Coverage Δ
unittests	?
unittests-Solution-Experimental	83.09% <100.00%> (?)
unittests-Solution-Stable	83.08% <100.00%> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.

Files	Coverage Δ
...tation.AspNetCore/Implementation/HttpInListener.cs	89.72% <100.00%> (+0.14%)	⬆️
...emetry/Metrics/Exemplar/SimpleExemplarReservoir.cs	80.43% <100.00%> (ø)

... and 33 files with indirect coverage changes

[alanwest]

unrelated but kind of related nit: maybe this variable should be named isDelta for readability?

[cijothomas]

yes... The Reservoir itself is not necessarily aware of temporality concept. All it needs to know is if it should reset its internal state or not.. The caller knows temporality and passes reset to this...

[CodeBlanch]

Not a blocker for this PR, but just looking at this API shape, is the correct thing to do to expose bool reset?

SDKs are free to decide whether "collect" should also reset internal storage for delta temporal aggregation collection, or use a more optimal implementation.

Curious what other languages have done to solve this.

[cijothomas]

The current APIs need to be revised before exposing them to public (its marked internal), and there are many TODOs to be solved.