[MapProxy-dev] RFC: Introduce linting

[Tom Kralidis]

Big +1 here.  We use flake8 as part of Travis/GitHub Actions in
numerous projects
and have found it valuable in code style alignment.  We will need to
address E501
(max-line-length), however this is configurable (like many flake8 settings).

I also like your approach of tackling flake8 errors iteratively.
Perhaps we can update
[1] with a checklist of all the error types so if anyone wants to jump
they can see the
state of things at a given point in time.

Big +1 from me.

..Tom

[1] https://github.com/mapproxy/mapproxy/issues/478


On Wed, Jan 13, 2021 at 2:36 AM Julien R <julien.rebetez at picterra.ch> wrote:
>
> Hello,
>
> I would like to propose adding linting to the mapproxy codebase and enforce it on the CI. The benefits of linting are mostly that they make reviews easier because reviewers don't have to point stylistic issues.
>
> Johannes Weskamm originally proposed to introduce linting in this github issue:
> https://github.com/mapproxy/mapproxy/issues/478
>
> I personally have experience with using flake8 professionally for multiple years and it works great. I started experimenting with setting up flake8 on the CI here:
> https://github.com/mapproxy/mapproxy/pull/486
>
> If this is accepted, the plan would be to:
> - Introduce flake8 on CI, but disable all errors at first
> - Enable errors one by one in separate commits
>
> This is to avoid submitting a massive PR touching everything. See the attached 'mapproxy_flake8_stats.txt' file for statistics about the "errors" found by flake8.
>
> Best regards,
> Julien
>
> --
> Julien Rebetez
> Lead Software & Machine Learning Engineer
> Picterra SA
> Rue de la Mouline 8, 1022 Chavannes, Switzerland
> www.picterra.ch
> _______________________________________________
> MapProxy-dev mailing list
> MapProxy-dev at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/mapproxy-dev