Guard MSONAtoms definition behind ASE package availability#3645
Guard MSONAtoms definition behind ASE package availability#3645janosh merged 5 commits intomaterialsproject:masterfrom
MSONAtoms definition behind ASE package availability#3645Conversation
that 👍 |
As opposed to a linter rule that only catches my fix ;) |
9d43e87 to
0424e72
Compare
|
Makes sense. Thanks! |
|
I'm actually a bit confused why |
If anyone is interested, click details below... I guess the reason is that Found 788 errors in 81 files (checked 279 source files) |
|
types and we're running i realized i never bothered to check if https://github.com/RobertCraigie/pyright-python re comment, pretty sure |
|
on a related note, the args: ["--ignore-missing-imports", "--scripts-are-modules"] |
|
Could we get this in then? I don't know how quickly you'd want to make another release but this does break several other packages at the moment (even just within the MP ecosystem -- atomate2, emmet, mp_api etc). Any fixing or new linting rules could be tried out in #3646 or otherwise handled by the maintainers as they see fit going forward |
2ed483c to
ef43267
Compare
|
i'll take a look at this tomorrow morning. it'll definitely be in the next release. we need one for several other issues as well |
…vert MSONAtoms implementation
…whole test_ase.py module
Scratch that, misread it! LGTM |
i hope not. the behavior is meant to be unchanged. given this class only raises on init, i don't see how you'd get an error on import? |
You just beat my ninja edit, in the time it took me to walk to my computer to test it out - never review PRs from your phone! Thanks for updating this @janosh |
Summary
Hopefully closes #3644. Not sure if this is the approach you want to take.
Todo
It would be nice to add a test or linter rule that would have caught the initial issue.