Open stdedos opened 1 year ago
It seems it's a problem inside pylint-pytest, but right now it's unclear who's maintaining it so let's keep this issue opened until it's clearer.
Should pytest
pylint
capture these at the top level?
I like the "maybe-pylint-bug" template - but I think on that case pytest
pylint
may have been able to say "plugin xyz exploded. pls report to plugin-maintainer. I'll continue without it." or something along those lines?
There's three libs at work here:
It seems this bug is happening outside of pylint, if you want something to change in pytest you're going to need to open an issue in pytest, if you want something to change in pylint-pytest there's no clear maintainer ship right now, but we plan on maintaining it one way or the other in the midterm see https://github.com/PyCQA/meta/issues/56.
Sorry, sorry. I meant pylint
:facepalm:
Ha, right. No it's voluntary to not catch the error, we need the stacktrace and a clear fail when something goes wrong (so we get an issue opened and the users have information available).
I was thinking that, "if you did all that", but then just unloaded the plugin and continue, it would give pylint
s issue tracker less downstream issues.
But sure, I agree that explicit > implicit, so let's let it be.
I'm planning to touch pylint-pytest
soon with https://github.com/reverbc/pylint-pytest/pull/22 - so might as well give this a better look, on top of context.
Bug description
Using to "facilitate"
pylint
+pytest
, I receive:Configuration
No response
Command used
... and then files of the repo.
(i.e., called via
pre-commit
)(PS: add the following information on the
~/.cache/pylint/pylint-crash-2023-02-15-14-17-14.txt
file :pray:)Pylint output
Expected behavior
Not-a-stack-trace
Pylint version
OS / Environment
Ubuntu 20.04
Additional dependencies
No response