Make lexical_region_resolve public for Clippy

[camsteffen]

I have a locally working fix for rust-lang/rust-clippy#4002 that involves lexical_region_resolve::resolve. I'm doing my best to understand region checking and how it should integrate with Clippy, so please tell me if I'm totally on the wrong path. Basically I need to construct and evaluate a TypeOutlives obligation.

[rust-highfive]

r? @petrochenkov

(rust-highfive has picked a reviewer for you, use r? to override)

[camsteffen]

I wonder if the Cilppy suggestion should compile though? Isn't any A::b 'static as long as it is a function pointer?

[petrochenkov]

If something is public only for clippy, then it probably should have a comment saying that it's public only for clippy.

Regarding other questions, r? @jackh726 maybe.

[camsteffen]

After looking at this again, I think this is wrong. But there may be an opportunity to improve rustc coercion. I found out that take(G::f as fn(Arg)) works.

[jackh726]

I still need to take some time to look at this. I'll try to do that soon.

[jackh726]

@camsteffen do you still need this to be public? If so, do you have a branch I can look at?

[jackh726]

If you still need this, I think r=me with an appropriate comment

[camsteffen]

I decided not to go this route. I just opened rust-lang/rust-clippy#7661 with a workaround for the Clippy bug - I just ignore associated functions on type parameters. I'm fairly certain this is an issue with rustc coercion so I don't think it makes sense to add this complexity to Clippy to work around a rustc bug.

[camsteffen]

See https://github.com/rust-lang/rust-clippy/blob/ae2a95fadc098b326f2114fc5ee6f55ee0db81a8/clippy_lints/src/eta_reduction.rs#L113 and https://github.com/rust-lang/rust-clippy/blob/ae2a95fadc098b326f2114fc5ee6f55ee0db81a8/tests/ui/eta.rs#L248