Closed auspicacious closed 4 weeks ago
isort done.
pycodestyle
doesn't seem to catch anything that black
doesn't, at least in our current codebase. It is also less flexible when it comes to line length of docstrings. Not using pycodestyle
.
pylint
detects a lot of helpful stuff, some of it surprisingly specific. We can't enable everything at once (it complains about any TODO
s left in code) but we can enable quite a lot.
Added pylint
with most checks enabled, except those related to documentation and code smells like having too many input arguments.
Added coverage
for information about code coverage, just for fun.
I think this task can be considered completed for now. We can probably consider this our baseline for new development going forward. There are a number of other tools, but, presumably, the return on investment will decrease with each one.
In https://github.com/moonshot-nagayama-pj/PnPQ/issues/31 we added black and ruff to PnPQ.
However, there's some things that black doesn't really do, like sorting imports or enforcing naming conventions.
Black has a page on how to use other tools alongside it: https://black.readthedocs.io/en/stable/guides/using_black_with_other_tools.html
We should integrate isort and pycodestyle at a minimum, and experiment with others.
This is part of epic https://github.com/moonshot-nagayama-pj/testbed-meta/issues/13