better error message for trailing doc comment in traits (issue #56766) #57333
Conversation
|
rust-highfive
added
the
S-waiting-on-review
lcnr
changed the title
lcnr
changed the title
lcnr
changed the title
estebank
added
S-waiting-on-author
lcnr
force-pushed
the
doc-comment
branch
2 times, most recently
from
164ca62 to
101e91e
Compare
|
@estebank Done. It seems like I somehow automatically created |
|
☔ The latest upstream changes (presumably #57755) made this pull request unmergeable. Please resolve the merge conflicts. |
| @@ -5940,6 +5940,14 @@ impl<'a> Parser<'a> { | |||
| self.expect(&token::OpenDelim(token::Brace))?; | |||
| let mut trait_items = vec![]; | |||
| while !self.eat(&token::CloseDelim(token::Brace)) { | |||
| if let token::DocComment(_) = self.token { | |||
| if self.look_ahead(1, |tok| tok == &token::Token::CloseDelim(token::Brace)) { | |||
| self.span_fatal_err(self.span, Error::UselessDocComment).emit(); | |||
There was a problem hiding this comment.
It'd be nice if you used struct_span_err(...).emit() instead, and then consume the doc (as you're doing here with self.bump(). That way the parser will "succeed" and continue to type checking in the face of this.
There was a problem hiding this comment.
I tried doing that, but I was unable to find a way to emit an Error without using span_fatal_err. All other methods of Parsercurrently use a &str instead. I might look deeper into the diagnostics system once I have the time. Might take a while though 😢
There was a problem hiding this comment.
I think there's struct_span_err_with_code which will make you duplicate some of the work. That being said 1) a very small amount of duplication is potentially ok and 2) it might be very easy to unify all uses in a single method.
Also, take your time. We're here to help if you need us.
| } | ||
|
|
||
| fn main() { | ||
|
|
There was a problem hiding this comment.
Add a let _: usize =""; line to make sure the compiler recovers from the incorrect doc comment.
|
Ping from triage @lcnr: What is the status of this PR? |
|
I currently don't have much time and will probably want to improve the error functions of |
|
I don't think that I can finish this in the near future. |
closes #56766
r? @estebank