9321: Inline generics in const and function trait completions r=Veykril a=RDambrosio016
This PR does a couple of things:
- moves path_transform from ide_assists to ide_db to be shared by both assists and completions
- when completing a const or a function for a trait, it will "inline" any generics in those associated items instead
of leaving the generic's name. For example:
```rust
trait Foo<T> {
const BAR: T;
fn foo() -> T;
}
struct Bar;
impl Foo<u32> for Bar {
// autocompletes to this
fn foo() -> u32;
// and not this (old)
fn foo() -> T;
// also works for associated consts and where clauses
const BAR: u32 = /* */
}
```
Currently this does not work for const generics, because `PathTransform` does not seem to account for them. If this should work on const generics too, `PathTransform` will need to be changed. However, it is uncommon to implement a trait only for a single const value, so this isnt a huge concern.
Co-authored-by: rdambrosio <rdambrosio016@gmail.com>
9301: internal: Start refactoring ide_completion tests r=Veykril a=Veykril
Our current completion test infra resovles around usually just checking a specific `CompletionKind` which is suboptimal. We only see what we want to see in tests with this causing us to miss a lot of incorrect completions we are doing. Instead we should test for different cursor locations for all kinds(sans the magic kind maybe? not sure yet). This way we will also see potential duplicate completions that merely different in their kind.
Also since most completion submodules complete things in tests of other modules due to the tests overlapping it makes more sense to group these tests differently which implies moving them to a new module. Exceptions for this might be stuff like attribute completion as these cannot currently interfere.
I only wrote a few tests to check for completions in `ItemList` position so far and I already found a few incorrect/irrelevant completions as these haven't been tested properly due to them being hidden by the `CompletionKind` filtering.
I think `CompletionKind` doesn't really seem to be beneficial to me as to I can't think of a occasion where we would want to only check a specific completion kind.
Co-authored-by: Lukas Wirth <lukastw97@gmail.com>
