[Merged by Bors] - Impl AsRef+AsMut for Res, ResMut, and Mut #2189
Conversation
alice-i-cecile
added
C-Code-Quality
|
Of particular note: this is likely to make sharing logic between components and resources easier. For example I would like to try to use this to create EventReaders that can be stored in both system and query parameters in #2116. |
| impl<'w, T: Component> AsMut<T> for ResMut<'w, T> { | ||
| fn as_mut(&mut self) -> &mut T { | ||
| &mut *self | ||
| } | ||
| } | ||
|
|
There was a problem hiding this comment.
could you put this impl next to impl<'w, T: Component> AsRef<T> for ResMut<'w, T> below?
There was a problem hiding this comment.
Good call, not sure how that got mixed up
|
It may be worth marking these |
I don't mind doing it, but I'm curious - is rustc not usually aggressive enough to make that necessary? |
Soo here's where I get a little confused, as yes I do believe the compiler is smart enough to inline this. However across the code base I've seen |
|
Because of the generics it would be possible for rustc to inline it even without Functions can only be inlined when they are codegened as part of the same crate. This only happens for functions from upstream crates when either |
|
@bjorn3 okay, done! |
|
Awesome looks good to me! |
NathanSWard
added
the
S-Ready-For-Final-Review
|
bors r+ |
This can save users from having to type `&*X` all the time at the cost of some complexity in the type signature. For instance, this allows me to accommodate @jakobhellermann's suggestion in #1799 without requiring users to type `&*windows` 99% of the time.
|
Pull request successfully merged into main. Build succeeded: |
bors
bot
changed the title
This can save users from having to type `&*X` all the time at the cost of some complexity in the type signature. For instance, this allows me to accommodate @jakobhellermann's suggestion in bevyengine#1799 without requiring users to type `&*windows` 99% of the time.
This can save users from having to type
&*Xall the time at the cost of some complexity in the type signature. For instance, this allows me to accommodate @jakobhellermann's suggestion in #1799 without requiring users to type&*windows99% of the time.