Suggest removing Box::new instead of unboxing it

[camsteffen]

No description provided.

[rustbot]

r? @jackh726

rustbot has assigned @jackh726.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[fmease]

Does your code crash on this?

fn main() {
    let _: String = Box::new(String::new()❩;
}

I've replaced the final ) (U+0028) with ❩ (U+2769). CC #128790.

[camsteffen]

Hmm, I didn't want to use the argument span since it could be a parent or child expansion...

[fmease]

If you want to keep basing things off of expr.span, consider using one of SourceMap::{start,end,next}_point or similar.

Otherwise, shouldn't you be able to use things like find_ancestor_inside{,_same_ctxt}?

[camsteffen]

Fixed

[camsteffen]

Oh I saw your next comment after a refresh. end_point is good!

Otherwise, shouldn't you be able to use things like find_ancestor_inside{,_same_ctxt}?

That would work if the arg is a child expansion, but not if it's a parent. -- Box::new($x)