Change scalafmt.conf

[toddburnside]

Use file from lucuma-ui

[bundlemon]

BundleMon

Unchanged files (4)

Status	Path	Size	Limits
✅	index.(hash).js	9.5MB	-
✅	vendor.(hash).js	1.5MB	-
✅	index.(hash).css	767.2KB	-
✅	internal-0.(hash).js	619.41KB	-

No change in files bundle size

Final result: ✅

[github-actions]

Visit the preview URL for this PR (updated for commit cdf8e09):

https://explore-gemini--pr1133-change-scalafmt-conf-h9j5el8j.web.app

_{(expires Tue, 05 Oct 2021 22:11:46 GMT)}

_{🔥 via Firebase Hosting GitHub Action 🌎}

[rpiaggio]

I missed the comment in lucuma-ui and I see now that we lost aligning =>s in match/cases and <- in fors. Or there seems to be a low threshold to keep them aligned. It'd be cool to just get those back.

[toddburnside]

I missed the comment in lucuma-ui and I see now that we lost aligning =>s in match/cases and <- in fors. Or there seems to be a low threshold to keep them aligned. It'd be cool to just get those back.

Yeah. It's a bit odd. For example, in ConnectionsStatus.scala some, but not all of the => are aligned.

[cquiroz]

LGTM

[toddburnside]

Interestingly, it only seems to be possible to get the alignment of the => and <- with align.preset = most, not with any of the individual settings. Of course, that also brings back the aligned =, even for definitions separated by new lines.

Also, I like the align.arrowEnumeratorGenerator = true setting, which gives

instead of

This is something we had previously with the most preset, but can also be set independently.

I wonder if we should wait for the fix to scalameta/scalafmt#2775 in scalameta/scalafmt#2753 and go back to the most preset? The PR has been merged, and they seem to be making a new release every few hours.

[cquiroz]

I also prefer the alignment inside the for loops

We can certainly wait for upstream fixes

[toddburnside]

I also prefer the alignment inside the for loops

We can certainly wait for upstream fixes

The failure to sometimes not align the => and <- had to do with multiline statements. align.multiline = true needed to be set, but that also caused the extreme alignment of the = even across spaces. There have been 2 PRs merged to fix the multiline issues.

Unless someone objects, I'll close this PR and see what happens when a new version of scalafmt-core is released. I would like to compare what happens with align.preset = most vs what is in this PR with the addition of align.multiline = true and align.arrowEnumeratorGenerator = true.

[toddburnside]

Just a note for future work. If we use align.preset = most I think we can get rid of all (or at least most) of the align.tokens settings as they are included in most.

One reason to favor setting the tokens explicitly vs the most preset is to not get surprised by future additions to the most preset. On the other hand, I guess using most would alert us to any new alignment options, desirable or not.

[toddburnside]

Waiting on new version of scalafmt.