Open
Description
openedon Sep 5, 2023
If I have a struct like this:
struct T1 {}
struct T2 {}
pub struct Aggregate {
v1: Vec<T1>,
v2: Vec<T2>,
}
impl Aggregate {
pub fn compute<R>(&self, range: R) -> Option<u32> {
for v in self.v1.get(range)? {
output += v.compute();
}
for v in self.v2.get(range)? {
output += v.compute();
}
Some(output)
}
}The type bound for R would need to be R: SliceIndex<[T1]> + SliceIndex<[T2]>, which exposes internal implementation details.
I propose moving the <T> in SliceIndex<T> to the methods and leverage GATs:
trait SliceIndex : Sized {
type Output<T: ?Sized>: ?Sized;
fn get<T: ?Sized>(self, slice: &T) -> Option<&Self::Output<T>>;
fn get_mut<T: ?Sized>(self, slice: &mut T) -> Option<&mut Self::Output<T>>;
// ...
}Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment