Use ast_map in ty::get_attrs to enable querying attributes for any type of node.

[Munksgaard]

This is more flexible and less error-prone. get_attrs and
get_attrs_opt can be used on many more items than the old get_attrs
could.

This is all courtesy of @huonw, and directly taken from here:
https://github.com/rust-lang/rust/pull/22348/files#diff-0f85fcb07fb739876892e633fa0e2be6R5575

Also thanks to @Manishearth for pointing it out to me.

[rust-highfive]

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @Aatch (or someone else) soon.

If any changes to this PR are deemed necessary, please add them as extra commits. This ensures that the reviewer can see what has changed since they last reviewed the code. The way Github handles out-of-date commits, this should also make it reasonably obvious what issues have or haven't been addressed. Large or tricky changes may require several passes of review and changes.

Please see CONTRIBUTING.md for more information.

[eddyb]

I wonder if this should just return Borrowed(&[]) instead of None when it doesn't find the item.

Oh, and, that match is already done in ast_map - but the method is with_attrs and takes a closure.
It doesn't have to anymore (it can just return &'ast [Attribute] like the function you changed), but it wasn't updated.
I have done that in a local branch - I would prefer if you prevented code duplication by reusing that method (btw, the FIXME inside doesn't really seem significant anymore).

[Manishearth]

So I guess we should just replace the contents of with_attrs with the upper half of this function, and call it internally?

Change with_attrs to be get_attrs(&self, NodeId) -> &'tcx [Attribute]?

[brson]

@eddyb have your points been addressed?

[eddyb]

This shouldn't mention the closure anymore.

[eddyb]

Also, was there nothing using this function? 0 fallout?

[Munksgaard]

Nope, nothing at all, but I'd venture that it's still a useful function for plugins and such.

[eddyb]

r? @eddyb
@brson there's two nits (one of which should prevent tests for succeeding), can r=me with them fixed.

[Munksgaard]

Well now the pull request isn't really aptly named anymore, but the commits are cleaner.

[eddyb]

@bors rollup r+

[bors]

@bors r=eddyb caf6f17

[bors]

⌛ Testing commit caf6f17 with merge 698c100...

[bors]

☀️ Test successful - auto-linux-32-nopt-t, auto-linux-32-opt, auto-linux-64-nopt-t, auto-linux-64-opt, auto-linux-64-x-android-t, auto-mac-32-opt, auto-mac-64-nopt-t, auto-mac-64-opt, auto-win-32-nopt-t, auto-win-32-opt, auto-win-64-nopt-t, auto-win-64-opt