Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 96.23%. Comparing base (95716bc) to head (b0459f9).
Report is 1 commits behind head on main.

@@            Coverage Diff             @@
##             main     #565      +/-   ##
==========================================
- Coverage   96.94%   96.23%   -0.71%     
==========================================
  Files         121      121              
  Lines       20340    20352      +12     
  Branches     4842     4816      -26     
==========================================
- Hits        19718    19586     -132     
- Misses        619      762     +143     
- Partials        3        4       +1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

I don't have an idea of how to do the incremental breaking changes yet. On one hand I want to get them out one-by-one as soon as possible for users to test out, but on the other hand, we can't just do breaking changes for every single rule. Do you have any idea on that? Is that possible to still keep the original behavior and then break the default options all together (when we finish the migration of all rules)?

Maybe we could make a factory function to create two version of rules with different defaults, use it to create two different entry?

After careful consideration, I think that this PR contains too much content. Maybe I should split it up and submit.
Here's what I think about the changes.

• update report location
• don't remove comments while auto-fixing
• support ImportExpression in ts version
• auto-fix optional chain when the option is never

Some safe changes (or maybe fixes)

• allow f?. (), f ?.(), f ?. () when the option is always

Since the core problem is the difference in the optional chain, I think maybe we can just add an option for the user to choose, which can also get a better auto-fix.
Example: @stylistic/function-call-spacing: ["error", "always", { "optionalChain": { "before": true, "after": true } }]

• don't auto-fix while there is a line break and the allowNewlines option is falsy

There are some syntax errors in the issue description, so we can do without porting it.
eslint/eslint#7787

Completing these changes one by one may reduce the breaking changes.

Oh yeah that would do. We could merge those non-breaking PRs and ship them, while we could mark all the needed breaking PRs draft (and keep them as small as possible), and finally merge then when we do majors.

Updated the description at the top, this PR should be merged after #606