format arrays in inline tables, optionally sort dependency feature lists

[syphar]

resolves #89, resolves #36

related

• cargo metadata docs about dependencies
• example how this would look like, applied to the docs.rs codebase

[syphar]

is there a way to get quicker feedback on any issues with CI?

right now I have to try a fix, and then wait for approval of the CI run, and then again try ?

( I don't own a windows machine)

[jplatte]

If there is any small refactor you can extract into a separate PR, I can merge it ahead of this one. Once you have a PR merged, I no longer need to approve CI runs 😅

Unfortunately I can't access repo settings do loosen rhe approval-required setting.

[syphar]

If there is any small refactor you can extract into a separate PR, I can merge it ahead of this one. Once you have a PR merged, I no longer need to approve CI runs 😅

perhaps #121 ?

Unfortunately I can't access repo settings do loosen rhe approval-required setting.

😅

[jplatte]

Oh yeah, merged!

[syphar]

Oh yeah, merged!

perfect, CI works now directly.

I'll ping you when CI is fixed.

[syphar]

This is a quickfix to make the tests pass.

I don't know much about git and its behaviour around line endings on windows, and also didn't fully understand the formatting code around line-endings.

With most other combinations I tried, the output was always LF, even on windows, no matter if I did set clrf: Some(true). Super odd.

Do you have a hint? or is this expected and I misunderstood the intended behaviour?

[syphar]

@jplatte anything missing here to make progress?

I'm happy to add whatever is needed.

Thanks!

( I'm aware this is really only a nice-to-have thing ;) )

[jplatte]

I want to look into the weird windows line ending issue but am very low on time. cc @thomaseizinger in case he can help out.

[syphar]

@jplatte I think I found the issue, it seems like i should have used the assert_eq test helper.

[jplatte]

Hm, I wonder why we need that helper in some tests but not others. Anyways, not related to this PR.