Make sure that all file loading happens via SourceMap
That way, callers don't need to repeat "let's add this to sm manually
for tracking dependencies" trick.
It should make it easier to switch to using `FileLoader` for binary
files in the future as well
cc #62948
r? @petrochenkov
Suggest Rust 2018 on `<expr>.await` with no such field
When type checking a field projection (`fn check_field`) to `<expr>.await` where `<expr>: τ` and `τ` is not a primitive type, suggest switching to Rust 2018. E.g.
```
error[E0609]: no field `await` on type `std::pin::Pin<&mut dyn std::future::Future<Output = ()>>`
--> $DIR/suggest-switching-edition-on-await.rs:31:7
|
LL | x.await;
| ^^^^^ unknown field
|
= note: to `.await` a `Future`, switch to Rust 2018
= help: set `edition = "2018"` in `Cargo.toml`
= note: for more on editions, read https://doc.rust-lang.org/edition-guide
```
Fixes https://github.com/rust-lang/rust/issues/63533
This PR also performs some preparatory cleanups in `fn check_field`; the last 2 commits are where the suggestion is introduced and tested respectively.
r? @varkor
Continue refactoring resolve and hygiene
The general goal is addressing FIXMEs from the previous PRs.
Merging similar data structures (+ prerequisites for such merging), accounting for the fact that all `ExpnId`s have associated data in `HygieneData` now (less `Option`s).
Also, some renaming.
This should be the last renaming session in this area, I think.
r? @matthewjasper
Traces already contain module info without that.
It's easy to forget to call `finalize_*` on a module.
In particular, macros enum and trait modules weren't finalized.
By happy accident macros weren't placed into those modules until now.
Test HRTB issue accepted by compiler
Hi! First Rust PR, so if anything needs changing just let me know and I'll take care of it right away.
Closes#50301 which was marked E-needstest
Reduce the genericity of closures in the iterator traits
By default, closures inherit the generic parameters of their scope,
including `Self`. However, in most cases, the closures used to implement
iterators don't need to be generic on the iterator type, only its `Item`
type. We can reduce this genericity by redirecting such closures through
local functions.
This does make the closures more cumbersome to write, but it will
hopefully reduce duplication in their monomorphizations, as well as
their related type lengths.
Rollup of 11 pull requests
Successful merges:
- #62984 (Add lint for excess trailing semicolons)
- #63075 (Miri: Check that a ptr is aligned and inbounds already when evaluating `*`)
- #63490 (libsyntax: cleanup and refactor `pat.rs`)
- #63507 (When needing type annotations in local bindings, account for impl Trait and closures)
- #63509 (Point at the right enclosing scope when using `await` in non-async fn)
- #63528 (syntax: Remove `DummyResult::expr_only`)
- #63537 (expand: Unimplement `MutVisitor` on `MacroExpander`)
- #63542 (Add NodeId for Arm, Field and FieldPat)
- #63543 (Merge Variant and Variant_)
- #63560 (move test that shouldn't be in test/run-pass/)
- #63570 (Adjust tracking issues for `MaybeUninit<T>` gates)
Failed merges:
r? @ghost
move test that shouldn't be in test/run-pass/
We no longer test `src/test/run-pass/`; the proper way now is `// run-pass` in `src/test/ui/`
r? @petrochenkov
handle elision in async fn correctly
We now always make fresh lifetimne parameters for all elided
lifetimes, whether they are in the inputs or outputs. But then
we generate `'_` in the case of elided lifetimes from the outputs.
Example:
```rust
async fn foo<'a>(x: &'a u32) -> &u32 { .. }
```
becomes
```rust
type Foo<'a, 'b> = impl Future<Output = &'b u32>;
fn foo<'a>(x: &'a u32) -> Foo<'a, '_>
```
Fixes#63388
syntax: account for CVarArgs being in the argument list.
Fixes#63430 by testing for `1` argument (the `CVarArgs` itself) instead of `0`.
Note that the error has basically been impossible to trigger since the change that caused #63430, so perhaps we need an audit of untested errors.
Also, this check probably belongs in AST validation/HIR lowering, but I'd rather fix it in place for now.
r? @petrochenkov cc @dlrobertson
When encountering a boxed value as expected and a stack allocated value
that could be boxed to fulfill the expectation, like in the following
snippet, suggest `Box::new` wrapping.
