[order] Sort type imports

[aaronadamsCA]

Here's a preliminary PR that resolves #2339. It's my first contribution, so please bear with me! 🙂

• Type imports are now sorted before regular imports, because this appears to be the overwhelming ecosystem standard.
• To reduce ambiguity, error messages can now refer to "type import" instead of just "import".
• I included typeof support just in case (though I don't know much about Flow).
• Tests were updated to reflect revised behaviour, and two new tests were added.

I also did some opportunistic code cleanup:

• Some test messages seemed to reflect old non-deterministic rule behaviour with "before" vs. "after" errors. Given this is now predictable, I made the tests easier to understand and maintain.
• Coding style toward the end of the test file was all over the place, so I cleaned it up.
• I made a minor performance improvement in fixOutOfOrder.

Let me know if there's anything I should undo here, or separate into separate PRs.

[codecov]

Codecov Report

Base: 95.16% // Head: 94.98% // Decreases project coverage by -0.17% ⚠️

Coverage data is based on head (cff27b1) compared to base (395e26b).
Patch coverage: 92.00% of modified lines in pull request are covered.

❗ Current head cff27b1 differs from pull request most recent head c4f3cc4. Consider uploading reports for the commit c4f3cc4 to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2398      +/-   ##
==========================================
- Coverage   95.16%   94.98%   -0.18%     
==========================================
  Files          66       66              
  Lines        2831     2731     -100     
  Branches      951      916      -35     
==========================================
- Hits         2694     2594     -100     
  Misses        137      137              

Impacted Files	Coverage Δ
src/rules/order.js	98.48% <92.00%> (-0.66%)	⬇️
src/rules/newline-after-import.js	95.31% <0.00%> (-0.85%)	⬇️
src/rules/no-cycle.js	97.67% <0.00%> (-0.29%)	⬇️
src/rules/no-unused-modules.js	97.43% <0.00%> (-0.24%)	⬇️
utils/parse.js	98.14% <0.00%> (-0.13%)	⬇️
src/rules/dynamic-import-chunkname.js	96.77% <0.00%> (-0.11%)	⬇️
src/rules/export.js	100.00% <0.00%> (ø)
src/rules/no-restricted-paths.js	100.00% <0.00%> (ø)
src/rules/no-anonymous-default-export.js	100.00% <0.00%> (ø)
... and 2 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[stropho]

Hmm wondering what is this waiting for :) Anyway, if I was the maintainer, I would appreciate keeping the test coverage on the same level or higher.

A nice non-blocking improvement could be providing and option to sort type imports before/after the regular import. On the other hand that could be a different PR..

[ljharb]

Thanks, this is great!

We need some more test coverage, and it needs to be rebased (there’s a conflict).

[ljharb]

Regarding separate PRs, i don’t think that’s necessary, but it would be nice when rebasing to group the “cleanup” stuff into its own commits, separate from the “fix and updated tests” commits.

[stropho]

Hey @aaronadamsCA , how are you? What's the status? As this feature is very useful to me right now, I'm willing to help :) If you give me access to your fork, I could move this PR a bit further ;)

[ljharb]

@stropho its only been 4 days over a holiday weekend, give it some time :-)

[aaronadamsCA]

Back after 3 weeks off, I'll take a look this week, thanks!

[ljharb]

@aaronadamsCA this needs a rebase, and it's not a straightforward one. any chance you'll get to it soon?

[aaronadamsCA]

Sorry, my contributions fell behind for a while due to offline things, so thank you for completing this!

[stropho]

And now, finally publishing it after a while would be great :)