Improve print_type.
#144625
Improve print_type.
#144625
Conversation
`ty::Param`s have a name, might as well print it.
|
|
rustbot
added
S-waiting-on-review
|
Note: the output of |
| Ok(()) | ||
| } | ||
|
|
||
| ty::Param(param_ty) => { |
There was a problem hiding this comment.
Can you add a comment here? All the other sections in this match have a nice comment.
| | ty::Dynamic(_, _, _) | ||
| | ty::UnsafeBinder(_) => self.pretty_print_type(ty), | ||
|
|
||
| // Placeholders (all printed as `_` to uniformize them). |
There was a problem hiding this comment.
This does sound like the _ was deliberate -- but I don't have a strong opinion either way.
|
|
||
| fn main() { | ||
| assert_eq!(Bar(()).foo(), "issue_61894::Bar<_>::foo::f"); | ||
| assert_eq!(Bar(()).foo(), "issue_61894::Bar<M>::foo::f"); |
There was a problem hiding this comment.
What is confusing about this is that there is no M in scope when printing the name. If M shadows the name of some other type, this could become misleading.
There was a problem hiding this comment.
yea, iirc printing _ was on purpose, because it can't have been relevant for the type name if it's still an uninstantiated param. If there are two impl<M> Bar<M> (e.g. with different bounds, which is where such "dupliate impl blocks" are common), then the type names would still be the same, so adding the M does not really bring any benefit
rustbot
removed
the
S-waiting-on-review
4 participants
ty::Params have a name, might as well print it.r? @oli-obk