Incorrect suggestion for function parameter of type str

[camelid]

fn foo(bar: str) {}

(Playground)

Errors:

   Compiling playground v0.0.1 (/playground)
error[E0277]: the size for values of type `str` cannot be known at compilation time
 --> src/lib.rs:1:8
  |
1 | fn foo(bar: str) {}
  |        ^^^ doesn't have a size known at compile-time
  |
  = help: the trait `Sized` is not implemented for `str`
help: function arguments must have a statically known size, borrowed types always have a known size
  |
1 | fn foo(&bar: str) {}
  |        ^

error: aborting due to previous error

---

Applying the suggestion leads to yet more errors. The new errors do suggest using &str, but they still suggest invalid code:

error[E0308]: mismatched types
 --> src/lib.rs:1:8
  |
1 | fn foo(&bar: str) {}
  |        ^^^^-----
  |        |     |
  |        |     expected due to this
  |        expected `str`, found reference
  |        help: did you mean `bar`: `&str`
  |
  = note:   expected type `str`
          found reference `&_`

error[E0277]: the size for values of type `str` cannot be known at compilation time
 --> src/lib.rs:1:8
  |
1 | fn foo(&bar: str) {}
  |        ^^^^ doesn't have a size known at compile-time
  |
  = help: the trait `Sized` is not implemented for `str`
help: function arguments must have a statically known size, borrowed types always have a known size
  |
1 | fn foo(&bar: &str) {}
  |              ^

error: aborting due to 2 previous errors

After applying the new suggestions, I get:

error[E0277]: the size for values of type `str` cannot be known at compilation time
 --> src/lib.rs:1:9
  |
1 | fn foo(&bar: &str) {}
  |         ^^^ doesn't have a size known at compile-time
  |
  = help: the trait `Sized` is not implemented for `str`
  = note: all local variables must have a statically known size
  = help: unsized locals are gated as an unstable feature

error: aborting due to previous error

This seems likely to send new users in circles.

[camelid]

Assigning P-medium and removing I-prioritize as discussed in the prioritization working group.

[apiraino]

I've bisected to this nightly range:

+nightly-2020-07-12-x86_64-unknown-linux-gnu 346aec9b0 => old error msg
+nightly-2020-07-26-x86_64-unknown-linux-gnu d6953df14 => new error message

(I cannot bisect further because in between of that range I cannot install nightlies 🤷‍♂️ )

[hellow554]

Funny enough, when using _ as variable name, the suggestion will get displayed correctly:

fn f(a: str) {}
fn g(_: str) {}

   Compiling playground v0.0.1 (/playground)
error[E0277]: the size for values of type `str` cannot be known at compilation time
 --> src/lib.rs:1:6
  |
1 | fn f(a: str) {}
  |      ^ doesn't have a size known at compile-time
  |
  = help: the trait `Sized` is not implemented for `str`
help: function arguments must have a statically known size, borrowed types always have a known size
  |
1 | fn f(&a: str) {}
  |      ^

error[E0277]: the size for values of type `str` cannot be known at compilation time
 --> src/lib.rs:2:6
  |
2 | fn g(_: str) {}
  |      ^ doesn't have a size known at compile-time
  |
  = help: the trait `Sized` is not implemented for `str`
help: function arguments must have a statically known size, borrowed types always have a known size
  |
2 | fn g(_: &str) {}

[hellow554]

This seems to be fixed on nighlty, but not on beta...

somewhere between 5a4ab26...9d9c2c9

Fixed in 9d9c2c9

Fixed by #84313

So this can be closed @camelid 🎉

[hellow554]

But, as a sidenote: How could this ever be commited to master? I mean, it's obviosly wrong so say &a: str :/

I hope, that in the future the review process can be improved to detect such errors early on.

The error was introduced in #78152, where the test has been adapted to match the wrong syntax, but nobody has seen this :( Let's hope in the future it will get noticed right away 😊

[camelid]

The error was introduced in #78152, where the test has been adapted to match the wrong syntax, but nobody has seen this :( Let's hope in the future it will get noticed right away 😊

It looks like that PR caused a bunch of small diagnostics changes, so the reviewers probably just missed that bug.

[camelid]

Assigning myself for adding a test. @rustbot claim