GH-50277: [CI][Python] Avoid using generators for test parametrization on newer Pytest

[raulcd]

Rationale for this change

On the hypothesis tests we treat Warnings as errors and the job has been failing with the latest Pytest version (9.1.0). This is related to:

• Deprecate iterables which are not collections (e.g. generators) for tests parametrization pytest-dev/pytest#13409

With the latest pytest using generators for parametrization has been deprecated.

What changes are included in this PR?

Change the generator that was causing the failures and return a list of supported_factories instead.

Are these changes tested?

Yes both locally and the archery test that was failing.

Are there any user-facing changes?

No

• GitHub Issue: [Python][CI] test-conda-python-3.11-hypothesis fails with newest pytest #50277

[github-actions]

⚠️ GitHub issue #50277 has been automatically assigned in GitHub to PR creator.

[raulcd]

@github-actions crossbow submit test-conda-python-3.11-hypothesis

[Copilot]

Pull request overview

This PR updates PyArrow’s Python memory-pool tests to avoid passing a generator into pytest.mark.parametrize, which triggers PytestRemovedIn10Warning (treated as errors in CI) on newer pytest versions.

Changes:

• Convert supported_factories() from a generator (yield) to a concrete list-returning function.
• Collect supported memory pool factory callables into a list before returning, ensuring parametrize receives a Collection.

[github-actions]

Revision: cdb6e0e

Submitted crossbow builds: ursacomputing/crossbow @ actions-b28209428c

Task	Status
test-conda-python-3.11-hypothesis

[raulcd]

@AlenkaF if you can take a quick look

[raulcd]

@github-actions crossbow submit test-conda-python-3.11-hypothesis

[github-actions]

Revision: b716114

Submitted crossbow builds: ursacomputing/crossbow @ actions-966d54caba

Task	Status
test-conda-python-3.11-hypothesis

[conbench-apache-arrow]

After merging your PR, Conbench analyzed the 4 benchmarking runs that have been run so far on merge-commit 3f4a04e.

There were no benchmark performance regressions. 🎉

The full Conbench report has more details. It also includes information about 2 possible false positives for unstable benchmarks that are known to sometimes produce them.