Fix E0191 suggestion for empty dyn trait args

[qaijuang]

Fixes #155578.

[rustbot]

HIR ty lowering was modified

cc @fmease

[rustbot]

r? @jackh726

rustbot has assigned @jackh726.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

Why was this reviewer chosen?

The reviewer was selected based on:

• Owners of files modified in this PR: compiler
• compiler expanded to 72 candidates
• Random selection from 18 candidates

[Kivooeo]

Suggested change

	let code = if let Some(snippet) = snippet.strip_suffix('>') {
	// Don't insert a leading comma for empty generic args like `dyn Trait<>`.
	let separator = if snippet.trim_end().ends_with('<') { "" } else { ", " };
	// The user wrote `Trait<'a>` or similar and we don't have a term we can suggest,
	// but at least we can clue them to the correct syntax `Trait<'a, Item = /* ... */>`
	// while accounting for the `<'a>` in the suggestion.
	format!("{}, {}>", snippet, bindings.join(", "))
	format!("{snippet}{separator}{}>", bindings.join(", "))
	} else if in_expr_or_pat {
	// The user wrote `Trait`, so we don't have a term we can suggest, but at least we
	let code = if let Some(snippet) = snippet.strip_suffix("<>") {
	// Empty generics
	format!("{snippet}<{}>", bindings.join(", "))
	} else if let Some(snippet) = snippet.strip_suffix('>') {
	// Non-empty generics
	format!("{snippet}, {}>", bindings.join(", "))
	} else if in_expr_or_pat {
	// ...
	};

Here is prototype that I haven't tested, but, something like this should be easier to read and follow

View changes since the review

[fmease]

Could you please look for an existing test file that exercises this diagnostic and add the test case there? Of course, you should keep the reference to the GH issue in a comment above it.

This is super niche and doesn't warrant its own file.

View changes since the review