[x] Add unit tests and integration tests where applicable.
[x] Add a line to the release notes (docs/release_notes.rst) under "Current Development Version".
Make sure to include the PR number after you open and get one.
I'm not really happy about this heuristic, but it seems to do the work. It would be cleaner change the splitting of docstring to sections, but I'm afraid its not possible in general case due to numpy style having no indent. Anyway, this should fix the problem with google style convention and additional lines being considered part of the section.
The solution may not work in mixed tabs/spaces docstring where each line has different whitespace prefix, but I find this to be pretty obscure case (and it's obviously not covered by pydocstyle tests anyway).
Closes #618.
Make sure to include the PR number after you open and get one.
I'm not really happy about this heuristic, but it seems to do the work. It would be cleaner change the splitting of docstring to sections, but I'm afraid its not possible in general case due to
numpystyle having no indent. Anyway, this should fix the problem withgooglestyle convention and additional lines being considered part of the section.The solution may not work in mixed tabs/spaces docstring where each line has different whitespace prefix, but I find this to be pretty obscure case (and it's obviously not covered by
pydocstyletests anyway).