Do not suggest using a const parameter when there are bounds on an unused type parameter #93400

ChayimFriedman2 · 2022-01-28T00:08:27Z

The user wrote the bound, so it's obvious they want a type.

rust-highfive · 2022-01-28T00:08:29Z

(rust-highfive has picked a reviewer for you, use r? to override)

ChayimFriedman2 · 2022-01-28T00:44:44Z

Looks like hir::GenericParam::bounds does not include bounds from where clauses. How can check for them too? I can use GenericPredicates but I don't know how to differentiate implicit bounds from explicit bounds.

michaelwoerister · 2022-02-03T13:52:29Z

Thanks for the PR, @ChayimFriedman2! Maybe someone from the diagnostics working group can help you.

r? rust-lang/diagnostics

davidtwco · 2022-02-08T07:59:27Z

Looks like hir::GenericParam::bounds does not include bounds from where clauses. How can check for them too? I can use GenericPredicates but I don't know how to differentiate implicit bounds from explicit bounds.

You might be able to use explicit_predicates_of (it's been a while but I think this includes all predicates from the user but none of the implicit ones).

ChayimFriedman2 · 2022-02-13T11:12:29Z

Unfortunately, explicit_bounds_of() includes the implicit : Sized predicate:

rust/compiler/rustc_typeck/src/collect.rs

Lines 2179 to 2185 in 450ef86

    
           <dyn AstConv<'_>>::add_implicitly_sized( 
        
               &icx, 
        
               &mut bounds, 
        
               param.bounds, 
        
               Some((param.hir_id, ast_generics.where_clause.predicates)), 
        
               param.span, 
        
           );

I ended with rather ugly (IMHO) code that lazily lowers the types in the where clause (because there is no easy way to tell if a type is a generic parameter using HIR types). I'm not sure this is a good idea, but I don't have a better one. Do you have a better idea? Or do you think there is no reason to do it lazily?

davidtwco

I think this looks reasonable but I'm not familiar enough with typeck to know for sure, will reassign.

davidtwco · 2022-02-16T02:37:38Z

r? @jackh726

bors · 2022-02-25T03:45:33Z

☔ The latest upstream changes (presumably #93368) made this pull request unmergeable. Please resolve the merge conflicts.

…used type parameter The user wrote the bound, so it's obvious they want a type.

jackh726 · 2022-02-26T02:13:27Z

@bors r+

bors · 2022-02-26T02:13:29Z

📌 Commit 4520045 has been approved by jackh726

…const-with-bounds-unused-generic-param, r=jackh726 Do not suggest using a const parameter when there are bounds on an unused type parameter The user wrote the bound, so it's obvious they want a type.

…askrgr Rollup of 5 pull requests Successful merges: - rust-lang#93400 (Do not suggest using a const parameter when there are bounds on an unused type parameter) - rust-lang#93982 (Provide extra note if synthetic type args are specified) - rust-lang#94087 (Remove unused `unsound_ignore_borrow_on_drop`) - rust-lang#94235 (chalk: Fix wrong debrujin index in opaque type handling.) - rust-lang#94306 (Avoid exhausting stack space in dominator compression) Failed merges: r? `@ghost` `@rustbot` modify labels: rollup

rust-highfive assigned michaelwoerister Jan 28, 2022

rustbot added the T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. label Jan 28, 2022

rust-highfive added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Jan 28, 2022