DSTify BytesContainer

[japaric]

• The BytesContainer::container_into_owned_bytes method has been removed
• Methods that used to take BytesContainer implementors by value, now take them by reference. In particular, this breaks some uses of Path:

Path::new("foo")// Still works
path.join(another_path) -> path.join(&another_path)

[breaking-change]

---

Re: container_into_owned_bytes, I've removed it because

• Nothing in the whole repository uses it
• Takes self by value, which is incompatible with unsized types (str)

The alternative to removing this method is to split BytesContainer into BytesContainer for Sized? and SizedBytesContainer: BytesContainer + Sized, where the second trait only contains the container_into_owned_bytes method. I tried this alternative in another branch and it works, but it seemed better not to create a new trait for an unused method.

Re: Breakage of Path methods

We could use the idea that @alexcrichton proposed in #18457 (add blanket impl BytesContainer for &T where T: BytesContainer + keep taking T: BytesContainer by value in Path methods) to avoid breaking any code.

r? @aturon
cc #16918

[rust-highfive]

Warning

• These commits modify unsafe code. Please review it carefully!

[alexcrichton]

I'm fine with removing container_into_owned_bytes, and I would personally vote for the impl BytesContainer for &T where T: BytesContainer route as it's probably wanted anyway (&String is certainly a container of bytes!). This is a pretty common trait where the impl may be worth it.

[aturon]

Agreed with @alexcrichton. Can you update with the proposed change for Path ergonomics, and then I'll re-r?

[japaric]

Addressed Path ergonomics and squashed.

@aturon re-r?

[japaric]

Rebased. @alexcrichton or @aturon re-r?

[japaric]

@aturon / @alexcrichton Could you tell bors to retry?

The previous attempt failed because one of the bots was continuously failing an udp-related test. But, I think the bots are properly working now.

[aturon]

@japaric done!