Feat: Add benchmark

[woile]

Warning

This PR introduced a breaking change, because it removes support for python 3.8

This PR adds benchmark capabilities to the repo by using pytest-benchmark

This will allow us to guarantee no regressions on performance on the benchmarked code.

As of now, I've added a check for benchmark-compare-fail=min:5%:

will make the suite fail if Min is 5% slower for any test

But it's worth exploring and adding more options.

How it works?

• on merge -> create benchmarks for the current release. It runs ./script/bench-current, it will create a json file inside .benchmarks/ per platform and python version (probably a lot of files because we test against many versions, which is good), the results will be commited and pushed back to the primary branch, it will not trigger a new release.
• on PR -> test no regressions are introduced. It runs ./script/bench-current, creating a json file, but this time, it won't be commited. Instead it will just check that it's not over a certain threshold.

./scripts/bench-compare is a convenience script which you can run locally.

[github-actions]

PR Preview Action v1.4.8
🚀 Deployed preview to https://kpn.github.io/kstreams/pr-preview/pr-242/
on branch gh-pages at 2024-11-21 14:38 UTC

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 95.73%. Comparing base (c222b81) to head (0b3d40d).
Report is 3 commits behind head on master.

Additional details and impacted files

@@           Coverage Diff           @@
##           master     #242   +/-   ##
=======================================
  Coverage   95.73%   95.73%           
=======================================
  Files          24       24           
  Lines         961      961           
=======================================
  Hits          920      920           
  Misses         41       41           

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

---

🚨 Try these New Features:

• Flaky Tests Detection - Detect and resolve failed and flaky tests