fix(match_as_ref): also suggest downcasting references #15935
Conversation
rustbot
added
the
S-waiting-on-review
ada4a
force-pushed
the
15932-match_as_ref-v2
branch
2 times, most recently
from
9cde98a to
7a6f20a
Compare
| } | ||
|
|
||
| fn issue15932() { | ||
| let _: Option<&i32> = Some(0).as_mut().map(|x| x as _); |
There was a problem hiding this comment.
Why not Some(0).as_ref()?
There was a problem hiding this comment.
Just to be clear, I'm not opposed to that solution (though I did end up liking this one more) -- I explicitly opened two PRs so that you could compare the outputs and see which one you prefer
There was a problem hiding this comment.
Yeah, that we not clear to me, I get a lot of notifications from GitHub.
tests/ui/match_as_ref.fixed
Outdated
| fn issue15932() { | ||
| let _: Option<&i32> = Some(0).as_mut().map(|x| x as _); | ||
|
|
||
| let _: Option<&dyn std::fmt::Debug> = Some(0).as_mut().map(|x| x as &mut _ as &_); |
There was a problem hiding this comment.
Why is the double cast necessary here, or even useful? Some(0).as_mut().map(|x| x as &_) would work, and Some(0).as_ref().map(|x| x as &_) would be even better. Why not even Some(0).as_ref().map(|x| x as _)?
There was a problem hiding this comment.
To the double-cast part: it's because when I first tried that I got a type ambiguity error. Though now that I think about it, the error was actually caused by me trying to cast &mut u32 as &i32 (note that the types change as well). Fixed that.
As to why I don't use as_ref, see the reply above.
rustbot
added
S-waiting-on-author
ada4a
force-pushed
the
15932-match_as_ref-v2
branch
from
7a6f20a to
789d36f
Compare
|
#15934 will be merged |
3 participants
v2 of #15934
changelog: [
match_as_ref]: also suggest downcasting referencesr? @samueltardieu