New InitAggregationData abstraction/trait

[fraillt]

Changes

• Added new trait InitAggregationData, along with few helper types, which allow to cleanly define logic how aggregation is initialized. E.g.

impl<T> InitAggregationData for Sum<T>
where
    T: Number,
{
    type Aggr = data::Sum<T>;

    fn create_new(&self, time: AggregateTime) -> Self::Aggr {
        data::Sum {
            data_points: vec![],
            start_time: time.start,
            time: time.current,
            temporality: self.temporality,
            is_monotonic: self.monotonic,
        }
    }

    fn reset_existing(&self, existing: &mut Self::Aggr, time: AggregateTime) {
        existing.data_points.clear();
        existing.start_time = time.start;
        existing.time = time.current;
        existing.temporality = self.temporality;
        existing.is_monotonic = self.monotonic;
    }
}

• Using this implementation, we can significantly simplify code in collect functions (in 10 places). E.g.

-        let time = self.init_time.cumulative();
-        let s_data = dest.and_then(|d| d.as_mut().downcast_mut::<data::Sum<T>>());
-        let mut new_agg = if s_data.is_none() {
-            Some(data::Sum {
-                data_points: vec![],
-                start_time: time.start,
-                time: time.current,
-                temporality: Temporality::Cumulative,
-                is_monotonic: self.monotonic,
-            })
-        } else {
-            None
-        };
-        let s_data = s_data.unwrap_or_else(|| new_agg.as_mut().expect("present if s_data is none"));
-
-        s_data.start_time = time.start;
-        s_data.time = time.current;
-        s_data.temporality = Temporality::Cumulative;
-        s_data.is_monotonic = self.monotonic;
+        let mut s_data = AggregationData::init(self, dest, self.init_time.cumulative());

This could be expanded further to also construct "data points", which might make code simper once we introduce ability to change "value map" implementations (for different temporality in #2328).

Merge requirement checklist

• CONTRIBUTING guidelines followed
• Unit tests added/updated (if applicable)
• Appropriate CHANGELOG.md files updated for non-trivial, user-facing changes
• Changes in public API reviewed (if applicable)

[codecov]

Codecov Report

Attention: Patch coverage is 93.79845% with 8 lines in your changes missing coverage. Please review.

Project coverage is 77.8%. Comparing base (a1dda22) to head (e5d3549).

Files with missing lines	Patch %	Lines
...-sdk/src/metrics/internal/exponential_histogram.rs	60.0%	8 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##            main   #2500     +/-   ##
=======================================
- Coverage   77.9%   77.8%   -0.1%     
=======================================
  Files        123     123             
  Lines      22888   22818     -70     
=======================================
- Hits       17839   17774     -65     
+ Misses      5049    5044      -5     

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