chore(ci,tooling): speed up json-loader job

[danceratopz]

🗒️ Description

The json-loader job takes ~23 min. This PR applies two small quick wins to reduce this:

• Run the job on Python 3.14, matching fill, so coverage uses sys.monitoring instead of the slower C tracer. Coverage was enabled for json-loader in fix(ci,tooling): include json-loader in coverage reports #2975, but the job still runs on the default 3.13, reintroducing the C-tracer overhead that chore(ci): add setup-uv action, bump default ci python to 3.14 #2310 removed elsewhere (see the regression tracked in py3 CI job timing regression (~18min → ~38min) after PR #2252 #2314).
• Run the fixture-consuming pytest phase with the same worker count as the fill phase (-n {{ xdist_workers }}, which is auto in CI). It was pinned to --maxprocesses 6, capping it at 6 workers instead of scaling with the runner. The cap carried over from the tox configuration in feat(ci,tooling): replace tox with just #2555.

🔗 Related Issues or PRs

• fix(ci,tooling): include json-loader in coverage reports #2975
• chore(ci): add setup-uv action, bump default ci python to 3.14 #2310
• py3 CI job timing regression (~18min → ~38min) after PR #2252 #2314

✅ Checklist

• All: Ran fast static checks to avoid unnecessary CI fails, see also Code Standards and Enabling Pre-commit Checks:

  just static

• All: PR title adheres to the repo standard - it will be used as the squash commit message and should start type(scope):.
• All: Considered updating the online docs in the ./docs/ directory.
• All: Set appropriate labels for the changes (only maintainers can apply labels).

Cute Animal Picture

🐱

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 92.82%. Comparing base (5f8c109) to head (854a6e9).
⚠️ Report is 1 commits behind head on forks/amsterdam.

Additional details and impacted files

@@                 Coverage Diff                 @@
##           forks/amsterdam    #3010      +/-   ##
===================================================
- Coverage            93.20%   92.82%   -0.39%     
===================================================
  Files                  620      620              
  Lines                38777    36714    -2063     
  Branches              3342     3341       -1     
===================================================
- Hits                 36144    34081    -2063     
  Misses                1773     1773              
  Partials               860      860              

Flag	Coverage Δ
unittests	92.82% <ø> (-0.39%)	⬇️

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

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[danceratopz]

Hey @SamWilsn, thanks for the review here. I reverted the attempts to speed-up the json-loader job by splitting up the test_tools_new_fork.py::test_end_to_end test across different workers. This didn't seem to be working as intended. I think we can leave this as for a follow-up and get the other quick wins in for speed-up.

The new_fork tests take ~225s each and it looks like they get placed on one xdist worker, leading to poorly-balanced workers/starvation. As these tests are unit tests for the new fork tool and have nothing todo with JSON, I'd consider moving these out of the json-loader tests; there might be other candidates, will have to check. This would make it easier to split them and is cleaner.

[danceratopz]

@SamWilsn added an issue here:

• Move new-fork CLI tests out of the json-loader job #3022