[New] `no-extraneous-dependencies`: Add `considerInParents` option

[luxaritas]

Resolves https://github.com/import-js/eslint-plugin-import/issues/1913

I am not particularly happy with the implementation (in particular, the error passing and "requireExact" handling to match the current behavior feel misplaced and/or mishandled), but it's at the point that I feel it would be useful to get feedback (both on how the options are handled and the implementation).

[codecov[bot]]

Codecov Report

Base: 95.27% // Head: 95.23% // Decreases project coverage by -0.04% :warning:

Coverage data is based on head (a686917) compared to base (c3d14cb). Patch coverage: 98.07% of modified lines in pull request are covered.

:exclamation: Current head a686917 differs from pull request most recent head 61eddb2. Consider uploading reports for the commit 61eddb2 to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2481      +/-   ##
==========================================
- Coverage   95.27%   95.23%   -0.05%     
==========================================
  Files          68       66       -2     
  Lines        2944     2793     -151     
  Branches      998      940      -58     
==========================================
- Hits         2805     2660     -145     
+ Misses        139      133       -6     

Impacted Files	Coverage Δ
utils/readPkgUp.js	96.42% <95.23%> (+14.61%)	:arrow_up:
src/rules/no-extraneous-dependencies.js	100.00% <100.00%> (+1.78%)	:arrow_up:
utils/parse.js	98.14% <0.00%> (-0.13%)	:arrow_down:
src/rules/dynamic-import-chunkname.js	96.77% <0.00%> (-0.11%)	:arrow_down:
src/rules/order.js	99.10% <0.00%> (-0.09%)	:arrow_down:
src/rules/no-cycle.js	97.91% <0.00%> (-0.05%)	:arrow_down:
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)
src/rules/consistent-type-specifier-style.js
... and 4 more

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

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

[luxaritas]

FYI, I have seen these comments (thanks! good points). Hopefully I'll have time within the next week or two to address them.

[luxaritas]

Sorry for the delay! Rewrote this with your feedback and it is significantly simpler. :) As a bonus, this should also improve caching, since it's no longer necessary to touch the file system to look for the closest package.json. Let me know if there's anything further I can improve!

[luxaritas]

Resolved all the feedback - let me know if you want me to squash.

[WangHoHan]

What is the status of this PR?

[luxaritas]

Waiting for rereview as far as I'm concerned

[luxaritas]

Interesting idea to use set. I'd be interested to see that benchmarked given the small sizes and additional spread when creating the cache key, though I wouldn't immediately know how to do that myself. :) I can certainly add a few more tests.

[ljharb]

yeah lol for such a small array it probably doesn't matter, but might as well use the proper data structure :-p

[luxaritas]

@ljharb Can you help me understand why two of the tests I added are failing? It looks like this rule is trying to resolve any found import, even if it isn't present, and returns successfully if it isn't found? That.. can't be right, can it?

[ljharb]

Hmm, no, I don't know why it's failing :-/ i just rebased it so results are fresh.