Commit Graph

597 Commits

Author SHA1 Message Date
Laurențiu Nicola
36567eb9be Bump deps 2021-05-30 14:48:10 +03:00
cynecx
54d60fdee9 hir_ty: use async ret type for inference inside async bodies 2021-05-29 18:17:45 +02:00
Jonas Schievink
31588aea04 Remove fragment kind knowledge from builtin macros 2021-05-29 17:17:08 +02:00
Jonas Schievink
26b4777e1f Move hir_ty incremental test to its own file 2021-05-27 20:21:52 +02:00
Jonas Schievink
33debc4065 Update salsa 2021-05-27 15:05:41 +02:00
bors[bot]
f3aaae6555
Merge #9007
9007: Internal: `clippy::redundant_clone` fixes r=lnicola a=lnicola

bors r+

Co-authored-by: Laurențiu Nicola <lnicola@dend.ro>
2021-05-26 15:36:14 +00:00
Laurențiu Nicola
8206939fed clippy::redundant_clone fixes 2021-05-26 18:34:50 +03:00
bors[bot]
5587d0a3e3
Merge #8973
8973: internal: move diagnostics to hir r=matklad a=matklad



Co-authored-by: Aleksey Kladov <aleksey.kladov@gmail.com>
2021-05-25 20:33:21 +00:00
Aleksey Kladov
5c9f31d4c2 internal: move diagnostics to hir
The idea here is to eventually get rid of `dyn Diagnostic` and
`DiagnosticSink` infrastructure altogether, and just have a `enum
hir::Diagnostic` instead.

The problem with `dyn Diagnostic` is that it is defined in the lowest
level of the stack (hir_expand), but is used by the highest level (ide).

As a first step, we free hir_expand and hir_def from `dyn Diagnostic`
and kick the can up to `hir_ty`, as an intermediate state. The plan is
then to move DiagnosticSink similarly to the hir crate, and, as final
third step, remove its usage from the ide.

One currently unsolved problem is testing. You can notice that the test
which checks precise diagnostic ranges, unresolved_import_in_use_tree,
was moved to the ide layer. Logically, only IDE should have the infra to
render a specific range.

At the same time, the range is determined with the data produced in
hir_def and hir crates, so this layering is rather unfortunate. Working
on hir_def shouldn't require compiling `ide` for testing.
2021-05-25 17:49:59 +03:00
Lukas Wirth
28ca371755 Consider trait to be in scope for trait-impl 2021-05-25 16:16:29 +02:00
bors[bot]
33fdd512e3
Merge #8987
8987: Fix lowering of FnOnce() without return type r=flodiebold a=flodiebold

This should result in an implicit `-> ()`, not leaving out the binding.

Co-authored-by: Florian Diebold <flodiebold@gmail.com>
2021-05-25 13:25:15 +00:00
Florian Diebold
7c6f764ad6 Hide -> () in Fn traits 2021-05-25 15:23:52 +02:00
Lukas Wirth
3ee4e6c54c Fix type inference not working for new Try trait 2021-05-25 14:59:54 +02:00
Florian Diebold
35c948ff4a Fix lowering of FnOnce() without return type
This should result in an implicit `-> ()`, not leaving out the binding.
2021-05-25 14:29:53 +02:00
Florian Diebold
c33ee36d2a Minor test fixes / new tests 2021-05-25 13:07:18 +02:00
Florian Diebold
b26a472ccb Fix type mismatch caused by macros
MacroStmts should be completely transparent, but it prevented
coercion. (I should maybe give `infer_expr` and `infer_expr_inner`
better names.)
2021-05-25 11:15:02 +02:00
Aleksey Kladov
45112aa8c0 internal: rename hypothetical -> speculative
Lets steal this good naming from Roslyn before I forget about it yet
again.
2021-05-24 22:21:25 +03:00
bors[bot]
c139eff5b8
Merge #8963
8963: Bump deps r=matklad a=lnicola



Co-authored-by: Laurențiu Nicola <lnicola@dend.ro>
2021-05-24 15:34:52 +00:00
Laurențiu Nicola
74f1b21b08 Bump chalk 2021-05-24 16:27:24 +03:00
Jonas Schievink
8ebb8d29e1 internal: intern TypeBounds
Doesn't save much memory (~2 mb), but interning things is generally a
good pattern to follow
2021-05-24 15:13:23 +02:00
Florian Diebold
4a6cdd776d Record method call substs and use them in call info 2021-05-23 18:24:21 +02:00
Florian Diebold
a5d85a6356 Add test for #8931 and better checking 2021-05-23 12:52:41 +02:00
Florian Diebold
a88c7c04d2 Add last remaining module docstring 2021-05-22 16:39:56 +02:00
Florian Diebold
63614aafad Resolve any lifetime variables to 'static after inference
Chalk's unification can sometimes create lifetime variables, which we
currently don't really deal with, but at least we don't want to leak
them outside of inference.

Should fix #8919.
2021-05-22 14:27:22 +02:00
Florian Diebold
ef558c97d0 Clean up visibilities 2021-05-21 19:51:21 +02:00
Florian Diebold
67f1a08fd8 Some remaining cleanups 2021-05-21 18:23:03 +02:00
Florian Diebold
e9d1550001 Fix test after rebase 2021-05-21 18:23:03 +02:00
Florian Diebold
6e5637983c Record type mismatches for failed coercions in match etc. 2021-05-21 18:23:03 +02:00
Florian Diebold
556c9cebdb Refactor expectation handling
So as to not use `TyKind::Error` as "no expectation".
2021-05-21 18:23:03 +02:00
Florian Diebold
99c73537fa Remove TypeVariableTable 2021-05-21 17:49:09 +02:00
Florian Diebold
9716c0b949 Deal with goals arising from unification 2021-05-21 17:49:09 +02:00
Florian Diebold
4bd446f5b3 Get rid of resolve_ty_as_possible
Instead use shallow resolving where necessary.
2021-05-21 17:49:07 +02:00
Florian Diebold
a78f0076ab Make resolve_ty_shallow return Ty 2021-05-21 17:48:34 +02:00
Florian Diebold
1250ddc5cf Rework obligation handling
We can't do the easy hack that we did before anymore, where we kept
track of whether any inference variables changed since the last time we
rechecked obligations. Instead, we store the obligations in
canonicalized form; that way we can easily check the inference variables
to see whether they have changed since the goal was canonicalized.
2021-05-21 17:48:34 +02:00
Florian Diebold
a3d9cac690 Fix another panic 2021-05-21 17:48:34 +02:00
Florian Diebold
29266ada04 Improve debug printing without TLS 2021-05-21 17:48:34 +02:00
Florian Diebold
7c423f5b88 Fix panic 2021-05-21 17:48:34 +02:00
Florian Diebold
8397734cfe Fix HIR expecting errors to unify with anything 2021-05-21 17:48:34 +02:00
Florian Diebold
b939a6dcac Fix if/else coercion 2021-05-21 17:48:34 +02:00
Florian Diebold
212f0477f2 Make diverging type variables work again
Chalk doesn't know about the `diverging` flag, so we need to instead
propagate it before fully resolving the variables.
2021-05-21 17:48:34 +02:00
Florian Diebold
32fc944263 Fix handling of diverging branches in match coercion
Fixes #7626.
2021-05-21 17:48:34 +02:00
Florian Diebold
a09079f27a Fix coercion of two closures to a function pointer
Fixes #8604.
2021-05-21 17:48:34 +02:00
Florian Diebold
afa6be2435 Update tests with expected changes 2021-05-21 17:48:34 +02:00
Florian Diebold
3379a52dff Make type resolver a proper folder, make it resolve consts as well 2021-05-21 17:48:34 +02:00
Florian Diebold
278f5b043d Fix fallback to bound vars in unify 2021-05-21 17:48:34 +02:00
Florian Diebold
aebcf7b5d4 Better Debug impl for InternedWrapper 2021-05-21 17:48:34 +02:00
Florian Diebold
0f7f1f0705 Temporary fix for unknown expectations 2021-05-21 17:48:34 +02:00
Florian Diebold
4ca1981c91 Fix warnings & format 2021-05-21 17:48:34 +02:00
Florian Diebold
693582946f Rewrite coercion using the new unification 2021-05-21 17:48:33 +02:00
Florian Diebold
84074cb185 Remove our unification code, use Chalk's instead 2021-05-21 17:48:33 +02:00
Dawer
e2b1c69f74 Check patterns for type match recursively. 2021-05-19 11:08:59 +05:00
Dawer
472317c008 internal: Record mismatches of pattern types. 2021-05-19 11:08:59 +05:00
bors[bot]
a57bd59f35
Merge #8813
8813: Get some more array lengths! r=lf- a=lf-

This is built on #8799 and thus contains its changes. I'll rebase it onto master when that one gets merged. It adds support for r-a understanding the length of:

* `let a: [u8; 2] = ...`
* `let a = b"aaa"`
* `let a = [0u8; 4]`

I have added support for getting the values of byte strings, which was not previously there. I am least confident in the correctness of this part and it probably needs some more tests, as we currently have only one test that exercised that part (!).

Fixes #2922.

Co-authored-by: Jade <software@lfcode.ca>
2021-05-16 01:53:12 +00:00
Jade
de0ed9860d Address final feedback
* rename ConstExtension->ConstExt
* refactor a manual construction of a Const
2021-05-15 18:51:18 -07:00
Dawer
2abb4c99fe Fix false positive "Missing match arm". 2021-05-15 00:18:59 +05:00
Jade
78d6b88f21 Add more tests, refactor array lengths/consteval work
Fix #2922: add unknown length as a condition for a type having unknown.

Incorporate reviews:

* Extract some of the const evaluation workings into functions
* Add fixmes on the hacks
* Add tests for impls on specific array lengths (these work!!! 😁)
* Add tests for const generics (indeed we don't support it yet)
2021-05-14 01:39:28 -07:00
Jade
32c600664e Test lowering byte strings some more 2021-05-12 21:22:46 -07:00
Jade
e666589e63 Add support for lengths in array repeats, if they are literals
Now we will get the type of `[0u8; 4]`.
2021-05-12 21:22:46 -07:00
Jade
73023c0299 Support length for ByteStrings
I am not confident that my added byte string parsing is right.
2021-05-12 21:22:46 -07:00
Jade
8b147624ff Add lowering of array lengths in types
Now e.g.

```rust
fn a(b: [u8; 2]) {
}
```

will know about the length of b.
2021-05-12 21:22:46 -07:00
bors[bot]
312f1fe20a
Merge #8799
8799: Add basic support for array lengths in types r=flodiebold a=lf-

This recognizes `let a = [1u8, 2, 3]` as having type `[u8; 3]` instead
of the previous `[u8; _]`. Byte strings and `[0u8; 2]` kinds of range
array declarations are unsupported as before.

I don't know why a bunch of our rustc tests had single quotes inside
strings un-escaped by `UPDATE_EXPECT=1 cargo t`, but I don't think it's
bad? Maybe something in a nightly?

Co-authored-by: Jade <software@lfcode.ca>
2021-05-12 14:49:43 +00:00
Jade
f28c053c67 address review by @eddyb 2021-05-11 15:18:45 -07:00
bors[bot]
da80dfc022
Merge #8398
8398: Fix inference with conditionally compiled tails r=flodiebold a=DJMcNab

Fixes #8378

Co-authored-by: Daniel McNab <36049421+DJMcNab@users.noreply.github.com>
2021-05-11 19:01:39 +00:00
Jade
dc63fea427 Add basic support for array lengths in types
This recognizes `let a = [1u8, 2, 3]` as having type `[u8; 3]` instead
of the previous `[u8; _]`. Byte strings and `[0u8; 2]` kinds of range
array declarations are unsupported as before.

I don't know why a bunch of our rustc tests had single quotes inside
strings un-escaped by `UPDATE_EXPECT=1 cargo t`, but I don't think it's
bad? Maybe something in a nightly?
2021-05-11 05:25:19 -07:00
Lukas Wirth
41f470fea8 Correctly support SelfType when searching for usages 2021-05-08 22:34:55 +02:00
Jonas Schievink
20ae41c1a1 Reuse database in LowerCtx 2021-05-06 23:23:50 +02:00
Jonas Schievink
976a3226fe Don't store call-site text offsets in hygiene info 2021-05-06 19:59:54 +02:00
Daniel McNab
11c926fd97 Add a test for conditionally compiled tails 2021-05-03 14:14:33 +01:00
Daniel McNab
ebbcf9f458 Fix inference with conditionally compiled tails
Fixes #8378
2021-05-03 14:13:05 +01:00
Florian Diebold
c2aefd5b95 Don't look in super traits for <T as Trait>::Assoc
This isn't actually how it works, you have to specify the exact trait
that has the associated type.

Fixes #8686.
2021-04-29 20:23:02 +02:00
Florian Diebold
2d20ab7eaf Rewrite all_super_trait_refs as an iterator
Doesn't fix the bug I was trying to fix, but now that I did it anyway it
seems fine to keep.
2021-04-29 20:18:41 +02:00
Florian Diebold
b384cfcb81 Handle cycles in generic_defaults more gracefully 2021-04-29 20:00:43 +02:00
Florian Diebold
184a0d7c1e Add test for #8686 2021-04-29 20:00:21 +02:00
Comonad
78f1583bdd fix: closure unify without check ClosureId
closes #8604
2021-04-29 15:26:41 +08:00
Laurențiu Nicola
27c5e2da48 Bump chalk 2021-04-21 19:20:01 +03:00
bors[bot]
e4f7f1e1bd
Merge #8462
8462: Expand macros at type position r=jonas-schievink a=cynecx



Co-authored-by: cynecx <me@cynecx.net>
2021-04-19 13:01:30 +00:00
Jonas Schievink
20c27dbdbe Collect inherent impls in unnamed consts 2021-04-19 01:06:26 +02:00
cynecx
f0507ab7c6 hir_ty: cleanups and extend infinitely_recursive_macro_type test 2021-04-18 20:18:48 +02:00
cynecx
6ed2fd233b hir_ty: keep body::Expander in TyLoweringContext 2021-04-18 19:56:13 +02:00
cynecx
7ed42a3a52 hir_def: refactor expand_macro_type and cleanups 2021-04-17 17:38:45 +02:00
cynecx
28ef7c20d7 hir_ty: deal with TypeRef::Macro in HirFormatter 2021-04-17 16:24:56 +02:00
cynecx
cf3b4f1e20 hir_ty: Expand macros at type position 2021-04-17 16:24:56 +02:00
bors[bot]
5274eb12dd
Merge #8539
8539:  fix: Do not propose inherent traits in flyimports and import assists r=flodiebold a=SomeoneToIgnore

Closes https://github.com/rust-analyzer/rust-analyzer/issues/8520

I've went with a separate method approach, since the [highlighted code](https://github.com/rust-analyzer/rust-analyzer/issues/8520#issuecomment-819856337) has not`Type` and uses `Ty` to get his data, but the code I had to change has no access to `Ty` and has `Type` only.

Co-authored-by: Kirill Bulatov <mail4score@gmail.com>
2021-04-16 17:54:47 +00:00
Jonas Schievink
543d4ef7c5 Fix primitive shadowing with inner items 2021-04-16 19:28:22 +02:00
Kirill Bulatov
739edfd5cf Exclude inherent traits from flyimports 2021-04-16 11:13:17 +03:00
Kirill Bulatov
af8a6049a5 Profile trait solving for all invocations 2021-04-14 19:11:17 +03:00
Kirill Bulatov
75a2605361 Better places for spans 2021-04-14 17:15:37 +03:00
Kirill Bulatov
3390e73816 We need to go deeper 2021-04-14 16:59:08 +03:00
Kirill Bulatov
d1fc9d727b Add a missing span 2021-04-14 15:52:56 +03:00
Jonas Schievink
31594bcd01 decl_check: follow test style guide 2021-04-13 14:13:35 +02:00
bors[bot]
9beed98f2a
Merge #8432
8432: decl_check: consider outer scopes' allows r=jonas-schievink a=lf-

Fix #8417. Also makes it less noisy about no_mangle annotated stuff the
user can do nothing about.

Note: this still is broken with bitfield! macros. A repro in an ignore
test is included here. I believe this bug is elsewhere, and I don't
think I can work around it here.

I would like help filing the remaining bug, as it does actually affect
users, but I don't know how to describe the behaviour (or even if it
is unintended).

Co-authored-by: Jade <software@lfcode.ca>
2021-04-13 12:02:26 +00:00
Jade
26d2653dd6 address review feedback 2021-04-13 01:21:03 -07:00
Florian Diebold
97d6e36dbe Remove assertion in impl collection
This condition should always be true for *valid* code, but of course
there might be invalid code or things that we can't currently resolve.

Fixes #8464.
2021-04-11 12:07:58 +02:00
bors[bot]
5b40342d2d
Merge #8465
8465: Include more info in assert r=jonas-schievink a=jonas-schievink

This helped find https://github.com/rust-analyzer/rust-analyzer/issues/8464

changelog skip

bors r+

Co-authored-by: Jonas Schievink <jonasschievink@gmail.com>
2021-04-10 23:42:26 +00:00
Jonas Schievink
5f2efae3ba Include more info in assert 2021-04-11 01:41:40 +02:00
Jonas Schievink
e2c1da36f5 Support macros in pattern position 2021-04-11 01:25:50 +02:00
Lukas Wirth
8113c3a914 Respect test style guidelines in tests::traits 2021-04-10 18:03:27 +02:00
Lukas Wirth
a15b8136ee Add test for binary op return ty with adt 2021-04-10 17:52:24 +02:00
Lukas Wirth
d9554c258b Add manual ops::Add impls to test::traits::closure_2 2021-04-10 17:16:35 +02:00
Lukas Wirth
252eb78dc3 Implement more precise binary op return type prediction 2021-04-10 16:56:32 +02:00