Turn EncodingWarning into errors and cleanup pytest.ini

[Avasam]

Summary of changes

Attempt at supeerseding #4234 (comment) w/o any potential breaking change
Essentially a band-aid for #3810 to make reading failed tests results more bearable.

Pull Request Checklist

• Changes have tests (these are the test changes)
• News fragment added in newsfragments/.
  (See documentation for details)

[Avasam]

6 EncodingWarnings left with this, which I think is good enough for this PR https://github.com/pypa/setuptools/actions/runs/8148548361/job/22271588677?pr=4255#step:9:1897

[Avasam]

@abravalheri Looks like there's still at least one EncodingWarning left to be able to fully re-enable them as errors.
Although the pytest context seems to hide the true location, which I assume is somewhere in the build_py command (likely in the run method).
Maybe the call to distutils.util.byte_compile ?

[abravalheri]

Yeah, a bit of a shame...

It might be the case there is a open()... read() ... exec() sequence somewhere or a ConfigParser.read() somewhere in distutils that fails only if SETUPTOOLS_USE_DISTUTILS=stdlib...

Worst case scenario, I was wondering if there is a way of introducing a pytest.mark.filterwarnings("ignore::DeprecationWarning") decorator conditionally to SETUPTOOLS_USE_DISTUTILS in the complicated function...

[Avasam]

Up to you, looks like the last ignore line I added works. Although it could hide other issues in that module.
We could "default" that one instead of "ignore".

I also didn't originally add a news fragment for this PR, but now that we're preventing more EncodingWarning from reaching user code, that's a beneficial user-facing change that's worth mentioning.

[abravalheri]

I think I understood the root of the problem: pytest.warns seems to reset the filter.
One last attempt of being more specific...

[abravalheri]

Now, that problem with stdlib's distutils is gone, but there seems to be another weird problem with pytest-mypy and pypy 😩: https://github.com/pypa/setuptools/actions/runs/8791484502/job/24126944440

Screenshots

This part is really weird:

Why the FileNotFoundError is not being caught by the try...except block?

[abravalheri]

@Avasam, this is the change for the windows errors:

--- setuptools/tests/test_windows_wrappers.py | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/setuptools/tests/test_windows_wrappers.py b/setuptools/tests/test_windows_wrappers.py index 3f321386f..b27268935 100644 --- a/setuptools/tests/test_windows_wrappers.py +++ b/setuptools/tests/test_windows_wrappers.py @@ -110,7 +110,11 @@ class TestCLI(WrapperTester): 'arg5 a\\\\b', ] proc = subprocess.Popen( - cmd, stdout=subprocess.PIPE, stdin=subprocess.PIPE, text=True + cmd, + stdout=subprocess.PIPE, + stdin=subprocess.PIPE, + text=True, + encoding="utf-8", ) stdout, stderr = proc.communicate('hello\nworld\n') actual = stdout.replace('\r\n', '\n') @@ -143,7 +147,11 @@ class TestCLI(WrapperTester): 'arg5 a\\\\b', ] proc = subprocess.Popen( - cmd, stdout=subprocess.PIPE, stdin=subprocess.PIPE, text=True + cmd, + stdout=subprocess.PIPE, + stdin=subprocess.PIPE, + text=True, + encoding="utf-8", ) stdout, stderr = proc.communicate('hello\nworld\n') actual = stdout.replace('\r\n', '\n') @@ -191,6 +199,7 @@ class TestCLI(WrapperTester): stdin=subprocess.PIPE, stderr=subprocess.STDOUT, text=True, + encoding="utf-8", ) stdout, stderr = proc.communicate() actual = stdout.replace('\r\n', '\n') @@ -240,6 +249,7 @@ class TestGUI(WrapperTester): stdin=subprocess.PIPE, stderr=subprocess.STDOUT, text=True, + encoding="utf-8", ) stdout, stderr = proc.communicate() assert not stdout --  2.43.2

[abravalheri]

I am fixing the macos CI error in the upstram/main (#4327) then rebasing this then merging.

Thank you vrey much @Avasam.