Customizable Run conda tests workflow

[grusev]

Reference Issues/PRs

What does this implement or fix?

Idea is same as with #2379

The actual PR is also for "test_diff_long_stream_descriptor_mismatch" test on MacOS which used to be flaky. No more such evidences. But to prove that we need also MacOS builds that allow tests for flakyness

The new workflow gives option to run or not cpp tests as well as to pass aditional arguments to pytest or completely rewrite pytest command to be able to achieve repetitions to test for flakyness. Example:

 pytest -n auto -v --count=100 -x python/tests/integration/arcticdb/test_storage_lock.py

---

Test on MacOS
Test: python/tests/integration/arcticdb/version_store/test_basic_version_store.py::test_diff_long_stream_descriptor_mismatch
Command: pytest -n 10 -v --count=1000 -x python/tests/integration/arcticdb/version_store/test_basic_version_store.py::test_diff_long_stream_descriptor_mismatch
Log: https://github.com/man-group/ArcticDB/actions/runs/15462578452
Analysis: the test executed ok on mac_os without any errors

More advanced test:
Command: pytest -n auto -v --count=2000 -x python/tests/integration/arcticdb/version_store/test_basic_version_store.py::test_diff_long_stream_descriptor_mismatch
Log: https://github.com/man-group/ArcticDB/actions/runs/15530261173/job/43717520944

Any other comments?

Checklist

Checklist for code changes...

• Have you updated the relevant docstrings, documentation and copyright notice?
• Is this contribution tested against all ArcticDB's features?
• Do all exceptions introduced raise appropriate error messages?
• Are API changes highlighted in the PR description?
• Is the PR labelled as enhancement or bug so it appears in autogenerated release notes?