Given the following code: (playground)
struct A<T: Sized>(T);
struct B(A<[u8]>);
The current output is:
error[E0277]: the size for values of type `[u8]` cannot be known at compilation time
--> src/lib.rs:2:10
|
1 | struct A<T: Sized>(T);
| - required by this bound in `A`
2 | struct B(A<[u8]>);
| ^^^^^^^ doesn't have a size known at compile-time
|
= help: the trait `Sized` is not implemented for `[u8]`
Ideally the output should look like:
error[E0277]: the size for values of type `[u8]` cannot be known at compilation time
--> src/lib.rs:2:10
|
1 | struct A<T: Sized>(T);
| ----- required by this bound in `A`
2 | struct B(A<[u8]>);
| ^^^^^^^ doesn't have a size known at compile-time
|
= help: the trait `Sized` is not implemented for `[u8]`
When the implicit Sized bound is added for generic type parameters, it goes at the front of the emitted predicate list. This means that its span, which is limited to the identifier of the generic, is what shows up in the error. It would be better if the error span pointed to the actual occurrence of the explicit Sized bound. Moving that implicit predicate to the end of the list would improve at least some of the error spans where there is an explicit Sized bound.
I'm working on a pull request for this issue.
@rustbot claim
@rustbot label +A-traits +A-typesystem +D-papercut
Given the following code: (playground)
The current output is:
Ideally the output should look like:
When the implicit
Sizedbound is added for generic type parameters, it goes at the front of the emitted predicate list. This means that its span, which is limited to the identifier of the generic, is what shows up in the error. It would be better if the error span pointed to the actual occurrence of the explicitSizedbound. Moving that implicit predicate to the end of the list would improve at least some of the error spans where there is an explicitSizedbound.I'm working on a pull request for this issue.
@rustbot claim
@rustbot label +A-traits +A-typesystem +D-papercut