Feature: enable auto fixable linters

[dsonck92]

This MR will enable linters and apply their suggested changes. Any linters that need manual decisions are for now excluded.

[loboda4450]

Hello @dsonck92,

honestly I do not think that including linting within the same pipelines is a good idea. Imo it should be splitted into separate CIs.

[dsonck92]

@loboda4450 I don't quite follow. Each module has a separate pipeline. Or do you mean you want to have the linting only on MRs and excluding them on others? I'm currently not entirely up to date with the naming in Github actions as I'm using Gitlab CI myself.

But feel free to tweak the CIs in a way you think works best. I'm curious about your proposal

[loboda4450]

I allowed myself to change them on your branch, so as not to build on each PR :D

[dsonck92]

I will cherry-pick that to the fix-module-names PR as I want to correct the module names first. This makes this PR more reliable as well (there seems to be fighting linters on how to sort the imports) I hope

[dsonck92]

@loboda4450 I realized that the old pipelines were still in the build-* files, I think this was not intended?

[loboda4450]

@dsonck92 obviously, I was in a rush and forgot to remove them :)

[dsonck92]

@bouassaba the godot linter is an autofix capable linter, but it's complaining about your style of comments as they don't follow go standards. I will delay this as it's a larger change (applying documentation to everything public). Also as note to self.

[bouassaba]

@bouassaba the godot linter is an autofix capable linter, but it's complaining about your style of comments as they don't follow go standards. I will delay this as it's a larger change (applying documentation to everything public). Also as note to self.

Is it complaining about the Swag comments used to generate OpenAPI specs?

[dsonck92]

Ahhhhh, that's where they are coming from. Yeah, I think it would have been nice if that project used the go: prefix convention here.

Well, in particular it just wants a dot at the end of a comment sentence

[bouassaba]

Ahhhhh, that's where they are coming from. Yeah, I think it would have been nice if that project used the go: prefix convention here.

Well, in particular it just wants a dot at the end of a comment sentence

What if we try to add the dot and do the adjustments required by the linter, then see if the Swag generation is still working? For generating the OpenAPI with redocly, I have it documented on the README.md of each service that has it.

[dsonck92]

A convention I used in the past was to add a trailing dot at the end of the comment, which should prevent the dot from interfering with any of the instruction lines

[dsonck92]

I think these are all the automatic linters. The rest need some consideration, and especially wsl will completely go crazy on the formatting as it doesn't like all the compressed code. I do think, based on the code at work, that adopting it makes for pretty readable code. But sometimes it's a bit tricky to understand and definitely opinionated.

[bouassaba]

I'm all in for code readability - totally agree 👍 I will be happy to follow the recommendations suggested by the tools.

[bouassaba]

I had a look through the diff, the automatic linters are doing a pretty good job.

[bouassaba]

@dsonck92 I'm done with the new Go based webdav, can we also add the linters for it?

[dsonck92]

@bouassaba I think this includes everything now. There are some interesting linters we should enable for webdav though.