macros: subdiagnostic derive

[davidtwco]

Add a new macro, #[derive(SessionSubdiagnostic)], which can be applied to structs that represent subdiagnostics, such as labels, notes, helps or suggestions.

#[derive(SessionSubdiagnostic)] can be used with the existing #[derive(SessionDiagnostic)]. All diagnostics implemented using either derive are translatable, and this new derive should make it easier to port existing diagnostics to using these derives.

For example, consider the following subdiagnostic types...

#[derive(SessionSubdiagnostic)]
pub enum ExpectedIdentifierLabel<'tcx> {
    #[label(slug = "parser-expected-identifier")]
    WithoutFound {
        #[primary_span]
        span: Span,
    }
    #[label(slug = "parser-expected-identifier-found")]
    WithFound {
        #[primary_span]
        span: Span,
        found: String,
    }
}

#[derive(SessionSubdiagnostic)]
#[suggestion_verbose(slug = "parser-raw-identifier")]
pub struct RawIdentifierSuggestion<'tcx> {
    #[primary_span]
    span: Span,
    #[applicability]
    applicability: Applicability,
    ident: Ident,
}

...and the corresponding Fluent messages:

parser-expected-identifier = expected identifier

parser-expected-identifier-found = expected identifier, found {$found}

parser-raw-identifier = escape `{$ident}` to use it as an identifier

These can be emitted using the new subdiagnostic function on Diagnostic...

diag.subdiagnostic(ExpectedIdentifierLabel::WithoutFound { span });
diag.subdiagnostic(RawIdentifierSuggestion { span, applicability, ident });

...or as part of a larger #[derive(SessionDiagnostic)]:

#[derive(SessionDiagnostic)]
#[error(slug = "parser-expected-identifier")]
pub struct ExpectedIdentifier {
    #[primary_span]
    span: Span,
    token_descr: String,
    #[subdiagnostic]
    label: ExpectedIdentifierLabel,
    #[subdiagnostic]
    raw_identifier_suggestion: Option<RawIdentifierSuggestion>,
}

sess.emit_err(ExpectedIdentifier { ... });

r? @oli-obk
cc @pvdrz

[rust-highfive]

Some changes occurred in src/tools/clippy.

cc @rust-lang/clippy

[estebank]

One thing I noticed is that you might not be supporting MultiSpans. I think they might be tricky to handle nicely at the macro level (you only have access to tokens, not types), and making it only MultiSpan would be limiting. 🤔

[oli-obk]

@bors r+

[bors]

📌 Commit dca8861 has been approved by oli-obk

[bors]

⌛ Testing commit dca8861 with merge 683c582...

[bors]

☀️ Test successful - checks-actions
Approved by: oli-obk
Pushing 683c582 to master...

[rust-timer]

Finished benchmarking commit (683c582): comparison url.

Summary:

• Primary benchmarks: 😿 relevant regression found
• Secondary benchmarks: 🎉 relevant improvements found

	Regressions 😿 (primary)	Regressions 😿 (secondary)	Improvements 🎉 (primary)	Improvements 🎉 (secondary)	All 😿 🎉 (primary)
count1	1	0	0	2	1
mean2	0.6%	N/A	N/A	-0.4%	0.6%
max	0.6%	N/A	N/A	-0.5%	0.6%

If you disagree with this performance assessment, please file an issue in rust-lang/rustc-perf.

@rustbot label: -perf-regression

Footnotes

1. number of relevant changes ↩

2. the arithmetic mean of the percent change ↩