-
Notifications
You must be signed in to change notification settings - Fork 12.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Fix suggestion for removing &mut from &mut macro!(). #85939
Changes from all commits
File filter
Filter by extension
Conversations
Jump to
Diff view
Diff view
There are no files selected for viewing
Original file line number | Diff line number | Diff line change |
---|---|---|
|
@@ -16,6 +16,7 @@ use rustc_span::Span; | |
use super::method::probe; | ||
|
||
use std::fmt; | ||
use std::iter; | ||
|
||
impl<'a, 'tcx> FnCtxt<'a, 'tcx> { | ||
pub fn emit_coerce_suggestions( | ||
|
@@ -577,12 +578,19 @@ impl<'a, 'tcx> FnCtxt<'a, 'tcx> { | |
// We have `&T`, check if what was expected was `T`. If so, | ||
// we may want to suggest removing a `&`. | ||
if sm.is_imported(expr.span) { | ||
if let Ok(src) = sm.span_to_snippet(sp) { | ||
if let Some(src) = src.strip_prefix('&') { | ||
// Go through the spans from which this span was expanded, | ||
// and find the one that's pointing inside `sp`. | ||
// | ||
// E.g. for `&format!("")`, where we want the span to the | ||
// `format!()` invocation instead of its expansion. | ||
if let Some(call_span) = | ||
iter::successors(Some(expr.span), |s| s.parent()).find(|&s| sp.contains(s)) | ||
{ | ||
if let Ok(code) = sm.span_to_snippet(call_span) { | ||
return Some(( | ||
sp, | ||
"consider removing the borrow", | ||
src.to_string(), | ||
code, | ||
Comment on lines
-585
to
+593
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. In another diagnostic, I emitted (potentially multiple) suggestions for removing the parts that need to be removed, instead of a single suggestion that contains a copy of (part of) the original source. Is there a guideline for which to prefer? Basically: Most diagnostics seem to use A, which surprises me a bit. The suggestion for A can get quite large for large expressions. There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. The later didn't exist until "recently" (2019?), so there's a historical component. There's also a bug that needs to be fixed on rustc's side to allow rustfix apply multipart suggestions, which it can't today. Even with that caveat, I push for multipart suggestions in all new diagnostics when possible/necessary. There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Thanks!
That should be fixed by rust-lang/rustfix#195 There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Awesome! That should also allow us to add a |
||
Applicability::MachineApplicable, | ||
)); | ||
} | ||
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@estebank What do you think about adding this as a function on
Span
(e.g.span.find_parent_within(span)
or something)?I have a feeling more diagnostics can/should use this, but I haven't written enough diagnostics yet to know if that's true.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I agree with your assessment and we should have something like that (bikeshed notwithstanding).