Should &'a Value as &'a Trait work?

[pnkfelix]

Inspired by rust-dev posting:
https://mail.mozilla.org/pipermail/rust-dev/2013-November/007054.html

Here is the relevant bit of code from the post:

trait T {}

fn f<'a, V: T>(v: &'a V) -> &'a T {
    v as &'a T
}

We need to ensure that for an expression <source> as <target> that any borrowed pointers in the type of are not obscured (#5723) by the cast.

A �collection of conditions sufficient to enforce this are listed in a comment in librustc/middle/kind.rs that I think is apropos here:

https://github.com/mozilla/rust/blob/master/src/librustc/middle/kind.rs#L488

However, there are probably other conditions that would also suffice that we might add to that set.

In particular, I do not see anything immediately wrong with the example from that mailing list post; the type-expression &'a V should ensure that V does not contain any lifetimes that are shorter than 'a, and therefore it should be safe, when <V: T> to cast v: &'a V to a &'a T.

(It could be that I have misinterpreted the constraints imposed by `&'a V.)

cc: @nikomatsakis

[nikomatsakis]

Basically yes. Not sure of status of code, have to recheck.

[pnkfelix]

((there was a comment here on a potentially related bug; however. further exploration/reduction led me to conclude that the other bug was orthogonal to this one. Therefore I have transcribed this comment into the description for #10902 and removed it from here.))

[treeman]

Code snippet from OP now compiles.

$ rustc -v
rustc 0.12.0-pre (f6a7ab40e 2014-08-29 08:21:26 +0000)

Can close.