DOC: Partial fix SA04 errors in docstrings #28792 (feedback needed)

[dilex42]

• xref DOC: Fix SA04 errors in docstrings #28792

Would like feedback on a few issues before continuing to work on this issue :

• File pandas/core/accessor.py . How to handle multiline text? Couldn't find anywhere good recomendations.
• File pandas/core/ops/docstrings.py . Changed logic of templates. Is this okay? Also new lines with formatted text are too long, would like a reccomended sollution for this.
• File pandas/core/window/rolling.py and some others in regard to window functions have broken links that can be fixed by adding pandas. prefix to them(as in changed code). Is this right thing to do and should I make this changes whenever i see this?

Sorry for obvious mistakes, just starting to contribute. Thanks.

[pep8speaks]

Hello @dilex42! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

There are currently no PEP 8 issues detected in this Pull Request. Cheers! 🍻

Comment last updated at 2020-03-29 07:54:22 UTC

[datapythonista]

Thanks for the work on this @dilex42

I think with the two suggested changes we can get most of this merged. We can discuss any remaining question you have in the new PR.

[dilex42]

@datapythonista Thanks for the feedback. Will do the changes.
Before merging this changes there is still a problem with docs to window functions. The references doesn't work without pandas prefix but validation script gives errors if prefix present. I guess there is some inconsistency either in doc generator or validator.
Also a general rule how to deal with multiline text would be appreciated.

[datapythonista]

Can you open an issue for the prefix on the validation. Leave them the way they render correctly. If the validation is breaking the CI, remove the validation of that specific error in ci/code_checks.sh please.

For the multiline, I think the best is to use whatever makes code more readable. I think in most cases would be a multiline string, but a \n can be better for simple things. But it's more a mix of common sense and personal preference than a rule. Always keeping readability in mind.

[dilex42]

Thanks

[datapythonista]

Really nice, couple of suggestions, but looks great.

[dilex42]

Should be good now.

[datapythonista]

Looks perfect, thanks @dilex42

[WillAyd]

Thanks @dilex42