New lint ignored_unit_patterns

[samueltardieu]

This idea comes from #11238. I've put the lint in pedantic as it might trigger numerous positives (three in Clippy itself).

changelog: [ignored_unit_patterns]: new lint

[rustbot]

r? @giraffate

(rustbot has picked a reviewer for you, use r? to override)

[Jarcho]

I think it would be better for the name to mention that it's about pattern matching. Maybe ignored_unit_patterns.

[samueltardieu]

I think it would be better for the name to mention that it's about pattern matching. Maybe ignored_unit_patterns.

Name and PR updated

[samueltardieu]

@Centri3 Thanks for your review, I think all points have been addressed.

[Centri3]

LGTM, I'll wait until @giraffate reviews as well though I think this should be good to go. Thanks!

[Centri3]

Oh yeah btw, I think in the future we can extend this to other patterns, like so: https://play.rust-lang.org/?version=stable&mode=debug&edition=2021&gist=55acf7c246efb174d34dfe0c4e177425, in a general pedantic lint for unnecessary wildcards. Though it's probably debatable on whether this is better in that case, it does prevent the type changing as well which may or may not be a good thing.

[samueltardieu]

I think in the future we can extend this to other patterns

Yeah, the first step could be to extend it to other singleton types (such as been done for `()`).

[giraffate]

Hmm, I'm wondering if restriction would be a good category. It looks like a strict coding style.

I would like hear @Centri3's opinion.

[Centri3]

pedantic is fine imo for the current lint, the only real strictness I think that may be undesirable in some cases is that the type cannot change, but it's opt in so it should be ok.

[giraffate]

@bors r=Centri3,giraffate

Thanks!

[bors]

📌 Commit f9a6dfa has been approved by Centri3,giraffate

It is now in the queue for this repository.

[bors]

⌛ Testing commit f9a6dfa with merge 1eb254e...

[bors]

☀️ Test successful - checks-action_dev_test, checks-action_remark_test, checks-action_test
Approved by: Centri3,giraffate
Pushing 1eb254e to master...

[cloneable]

This check also catches _: () as function parameter. Is this intended? Based on the error help I'm not sure it is.

error: matching over `()` is more explicit
 --> src/lib.rs:8:12
  |
8 | pub fn moo(_: ()) {}
  |            ^ help: use `()` instead of `_`: `()`
  |
  = help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#ignored_unit_patterns
  = note: `-D clippy::ignored-unit-patterns` implied by `-D clippy::pedantic`

[Centri3]

_: () is pat: ty, so yes, but I'm not sure if it's desirable

[cloneable]

Ok, I filed #11403 where this can be discussed, but feel free to close if all is good.