Jacob Walls
Jacob Walls
https://github.com/PyCQA/astroid/pull/1189 proposes to introduce `Constraint`. See: > ⚠️ If there is at least one inferred value, but all inferred values are filtered out, a new Uninferable value is yielded, rather...
pylint-dev/astroid#1189 landed. We should welcome a PR that adds `isinstance` constraints to nodes. This has the ability to address both false positives and false negatives (see #9405 for the negative).
Reproduced @ 27e72725fa8c216cb18ac833c20c658ec05ab9f5. We try to exclude lambdas that are called immediately, but we're fooled when an assumption of direct parentage is violated here: https://github.com/pylint-dev/pylint/blob/51c28bc922dfb8da42e680ed7190eb942889df01/pylint/checkers/utils.py#L447-L449 The lambda's parent is a...
If there are some messages that we don't want to demote as low as an extension (don't have any in mind, just saying...) then we could leave them enabled but...
> The too-* messages might actually be okay as is? I don't see that much benefit from moving them to extensions when users can also easily change the threshold values...
A promising experiment with disk caching was described https://github.com/pylint-dev/pylint/issues/2912#issuecomment-981712889. Then, the ability to "pin" the ASTs of 3rd-party modules is described [earlier in the same issue](https://github.com/pylint-dev/pylint/issues/2912).
@Pierre-Sassoulas I think @adamtuft is pointing out that we actually have this exact case in the test suite, just asserting the wrong result with a sad comment around it. When...
[pylint primer result](https://github.com/jacobtylerwalls/pylint/pull/33#issuecomment-2094334243)
I hear you. Plus, what if we get negative user-feedback on this first push -- don't want to sink time into a big numpy yak shave that turns out to...