Mechanical fix/workaround for tests/tools/pyroma/test_pyroma_tool.py

[valeriupredoi]

Description

This PR provides a mechanical fix/workaround (ie not a fix at the root cause, please see below for reasoning) for the failed test

Testing steps

• tested locally (on my machine) following the test steps from the Github Action (GA) workflow: result: 100% pass rate
• tested on GA via my personal fork: I replicated the test failure one can see on the main master

Clues: the files.files are identical on both machines; the messages indeed differ by two elements:

• Files and messages on local machine:

[PosixPath('/home/valeriu/prospector-fork/tests/tools/pyroma/testdata/pkg1/__init__.py'), PosixPath('/home/valeriu/prospector-fork/tests/tools/pyroma/testdata/pkg1/this_one_is_fine/setup.py'), PosixPath('/home/valeriu/prospector-fork/tests/tools/pyroma/testdata/pkg1/this_one_is_fine/__init__.py')]
[pyroma-PYR05, pyroma-PYR06, pyroma-PYRUNKNOWN, pyroma-PYR09, pyroma-PYR12, pyroma-PYR05, pyroma-PYR06, pyroma-PYRUNKNOWN, pyroma-PYR09, pyroma-PYR12]

• Files and messages on GA machine:

[PosixPath('/home/runner/work/prospector/prospector/tests/tools/pyroma/testdata/pkg1/__init__.py'), PosixPath('/home/runner/work/prospector/prospector/tests/tools/pyroma/testdata/pkg1/this_one_is_fine/__init__.py'), PosixPath('/home/runner/work/prospector/prospector/tests/tools/pyroma/testdata/pkg1/this_one_is_fine/setup.py')]
[pyroma-PYR05, pyroma-PYR06, pyroma-PYRUNKNOWN, pyroma-PYR09, pyroma-PYR12, pyroma-PYRUNKNOWN, pyroma-PYR05, pyroma-PYR06, pyroma-PYRUNKNOWN, pyroma-PYR09, pyroma-PYR12, pyroma-PYRUNKNOWN]

As you can see, the GA results include two additional pyroma-PYRUNKNOWN members in the messages list. My theory is that the setup.py file gets a little bit altered (trailing space character, newline character etc - could be anything non-major - something minor) which gives differing results at the point of pyroma run. File system peculiarities, I reckon.

PS: I've kept the change to the GA workflow file for possible further testing, I'll remove that if this will get merged. Cheers, fellas 🍺

Related Issue

This failing test needs be fixed so we can get #645 in and consequently release a version that supports Python 3.12

Motivation and Context

Fixes failing test tests/tools/pyroma/test_pyroma_tool.py

How Has This Been Tested?

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to change)

Checklist:

• My code follows the code style of this project.
• My change requires a change to the documentation.
• I have updated the documentation accordingly.
• My change requires a change to the dependencies
• I have updated the dependencies accordingly
• I have read the CONTRIBUTING document.
• I have added tests to cover my changes.
• All new and existing tests passed.

[Pierre-Sassoulas]

Thank you for your work @valeriupredoi. I don't have enough knowledge of prospector to review that, sorry.

[valeriupredoi]

@Pierre-Sassoulas not a problem at all, mate. Can any other of the prospector core devs review maybe? 🍺

[Pierre-Sassoulas]

@carlio would you mind reviewing this to unblock python 3.12 support down the line, please ? :)

[valeriupredoi]

am just a bit worried that @carlio 's been off GH scope for many months, do hope he's on a nice sabbatical/holiday 👍

[robertalexa]

@valeriupredoi @Pierre-Sassoulas Forgive me if I am about to say something stupid, i am sailing uncharted water here, but:

Shouldn't this simply be a case of getting the unique values for messages, in this case with a len of 5 (pyroma-PYR05, pyroma-PYR06, pyroma-PYRUNKNOWN, pyroma-PYR09, pyroma-PYR12), and thus satisfying execution in any context?

If order is irrelevant (which i feel it is), messages could be converted into a set and back to a list, assert the len and continue with the iteration.

Thoughts?

[Pierre-Sassoulas]

Thoughts?

I'm starting to think that this test possibly being invalid is not as impactful than prospector still not being available for python 3.12 / pylint 3.0. (And the numerous CVE it contains from some outdated dependencies)