8443: Rewrite `#[derive]` removal code to be based on AST r=jonas-schievink a=jonas-schievink
We now remove any `#[derive]` before and including the one we want to expand, in the `macro_arg` query.
The same infra will be needed by attribute macros (except we only remove the attribute we're expanding, not any preceding ones).
Part of https://github.com/rust-analyzer/rust-analyzer/issues/8434 (doesn't implement the cfg-expansion yet, because that's more difficult)
8446: Undo path resolution hack for extern prelude r=jonas-schievink a=jonas-schievink
Reverts the change made in https://github.com/rust-analyzer/rust-analyzer/pull/7959
We don't populate the extern prelude for block DefMaps anymore,
so this is unnecessary
bors r+
Co-authored-by: Jonas Schievink <jonasschievink@gmail.com>
8445: `hir_ty` cleanup r=flodiebold a=flodiebold
Move lots of things around within `hir_ty`. Most notably, all the Chalk-related stuff moves from within `traits/` to the top-level, since Chalk isn't purely a "traits thing" anymore.
Co-authored-by: Florian Diebold <flodiebold@gmail.com>
8444: Shrink `unlinked-file` diagnostic to 3 characters r=jonas-schievink a=jonas-schievink
Fixes https://github.com/rust-analyzer/rust-analyzer/issues/8442
(the diagnostic fires intentionally on `#[cfg]`d modules, but with this won't cover the whole file)
bors r+
Co-authored-by: Jonas Schievink <jonasschievink@gmail.com>
8406: Improve indexing of impls r=flodiebold a=flodiebold
Store impls for e.g. &Foo with the ones for Foo instead of the big "other" bucket. This can improve performance and simplifies the HIR impl search a bit.
Co-authored-by: Florian Diebold <flodiebold@gmail.com>
Store impls for e.g. &Foo with the ones for Foo instead of the big
"other" bucket. This can improve performance and simplifies the HIR impl
search a bit.
8440: Fix crash on syn involving lifetimes returned by Chalk r=flodiebold a=flodiebold
If we get lifetime variables back in autoderef, just immediately replace them by static lifetimes for now. Method resolution doesn't really deal correctly with new variables being introduced (this needs to be fixed more properly).
This fixes `rust-analyzer analysis-stats --with-deps` crashing in the RA repo.
Co-authored-by: Florian Diebold <flodiebold@gmail.com>
If we get lifetime variables back in autoderef, just immediately replace
them by static lifetimes for now. Method resolution doesn't really deal
correctly with new variables being introduced (this needs to be fixed
more properly).
This fixes `rust-analyzer analysis-stats --with-deps` crashing in the RA
repo.
8429: 8425: Added documentation for on enter covering //! doc comments. r=jonas-schievink a=chetankhilosiya
Also added passing test case.
Co-authored-by: Chetan Khilosiya <chetan.khilosiya@gmail.com>
8433: Intern lots of things r=jonas-schievink a=flodiebold
This uses the new interning infrastructure for most type-related things, where it had a positive effect on memory usage and performance. In total, this gives a slight performance improvement and a quite good memory reduction (1119MB->885MB on RA, 1774MB->1188MB on Diesel).
Co-authored-by: Florian Diebold <flodiebold@gmail.com>
8428: Use named fields in `MacroCallKind` r=jonas-schievink a=jonas-schievink
bors r+
changelog skip
Co-authored-by: Jonas Schievink <jonasschievink@gmail.com>
8426: Track recursion limit when expanding custom derive r=jonas-schievink a=jonas-schievink
You can write a custom derive that expands to itself:
```rust
#[proc_macro_derive(Derive)]
pub fn derive(item: TokenStream) -> TokenStream {
let mut out: TokenStream = "#[derive(Derive)]".parse().unwrap();
out.extend(item);
out
}
```
rustc reports a recursion limit error, but rust-analyzer used to spin in name resolution and eventually fail with "name resolution is stuck". This makes it fail fast by respecting the recursion depth of the invocation.
bors r+
Co-authored-by: Jonas Schievink <jonasschievink@gmail.com>
8419: Move hir_ty to Chalk IR r=flodiebold a=flodiebold
Closes#8313.
There's some further cleanups to do:
- we're still using our `TypeWalk` in lots of places (not for mutating/folding though, just for walking)
- we're still using our own canonicalization and unification and our `InferenceTable`
- ~`ToChalk` still exists and gets called, it's just the identity in most cases now (I'll probably clean those up before merging this)~
8423: Bump lsp-types and syn r=kjeremy a=kjeremy
This lsp-types now supports a default InsertTextMode for completion and a per-completion item commit_characters
Co-authored-by: Florian Diebold <flodiebold@gmail.com>
Co-authored-by: kjeremy <kjeremy@gmail.com>
