CodSpeed Performance Report

Merging #1235 will not alter performance

_{Comparing kumaraditya303:master (e751492) with master (e0e61c2)}

Summary

✅ 245 untouched benchmarks

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 99.85%. Comparing base (e0e61c2) to head (e751492).
⚠️ Report is 27 commits behind head on master.

@@            Coverage Diff             @@
##           master    #1235      +/-   ##
==========================================
- Coverage   99.85%   99.85%   -0.01%     
==========================================
  Files          26       26              
  Lines        3510     3507       -3     
  Branches      252      252              
==========================================
- Hits         3505     3502       -3     
  Misses          3        3              
  Partials        2        2              

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

@kumaraditya303 I did not notice this earlier but it seems some tests are failing for some reason let me go in and see if I can suggest any changes. The others are on a tighter schedule than me apparently.

I did not notice this earlier but it seems some tests are failing for some reason let me go in and see if I can suggest any changes. The others are on a tighter schedule than me apparently.

The failing tests are leak tests, it is likely that it is because of gc changes in 3.14 or just higher memory usage on 3.14 interpreter though I haven't looked at it closely.

I did not notice this earlier but it seems some tests are failing for some reason let me go in and see if I can suggest any changes. The others are on a tighter schedule than me apparently.

The failing tests are leak tests, it is likely that it is because of gc changes in 3.14 or just higher memory usage on 3.14 interpreter though I haven't looked at it closely.

Makes sense. I mainly implemented the newer memory-leak test to prevent newer memory leak issues in the future since there have been several incidents in MultiDict's history of this happening repetitively.

I am surprised it's not related to that one and only istr. When tests/test_istr.py::test_leak[py] is ran in isolation it passes which I find very strange.

@kumaraditya303 Great job, I'll be sure to contact one of the contributors over on matrix If I don't see any activity here in the next couple days.

Great job, I'll be sure to contact one of the contributors over on matrix If I don't see any activity here in the next couple days.

Thanks I have updated the tests for CPython 3.14, specifically these two changes:

• tests/test_istr.py: Changed it to check for case_insensitive_str_class instances rather than all objects which isn't stable.
• tests/isolated/multidict_pop.py: Change it to check for relative increase in memory rather than absolute memery usage.