Commit Graph

4788 Commits

Author SHA1 Message Date
bors[bot]
5a99184967
Merge #1680
1680: Correctly infer match with early return r=flodiebold a=SomeoneToIgnore

Fixes #1505

Co-authored-by: Kirill Bulatov <mail4score@gmail.com>
Co-authored-by: Florian Diebold <flodiebold@gmail.com>
2019-08-27 18:29:00 +00:00
Kirill Bulatov
4adfdea1ad Small fixes 2019-08-26 23:00:35 +03:00
Kirill Bulatov
590aed2eec Remove redundant tests 2019-08-26 23:00:27 +03:00
Florian Diebold
c256a72b03 Fix 'missing Ok in tail expr' validation
Because of the coercion change, the type mismatch now only happens on the block
expression.
2019-08-26 22:44:50 +03:00
Florian Diebold
e37b6c5837 Make infer_block not unify; add back calculate_least_upper_bound 2019-08-26 22:44:50 +03:00
Kirill Bulatov
44386d5373 An attempt to add the coercion logic for Never 2019-08-26 22:44:50 +03:00
Kirill Bulatov
89f3cc587d Properly coerce never types 2019-08-26 22:44:50 +03:00
Kirill Bulatov
8b612251fd Remove extra inference test 2019-08-26 22:44:50 +03:00
Kirill Bulatov
44cf7b34fe Fix never in if expressions 2019-08-26 22:44:50 +03:00
Kirill Bulatov
c1f47c3788 Add test marks 2019-08-26 22:44:50 +03:00
Kirill Bulatov
0ce05633a1 Fix match type inference for Never match arms 2019-08-26 22:44:50 +03:00
Kirill Bulatov
f63cfd5fca Tests 2019-08-26 22:44:50 +03:00
bors[bot]
04c2961d0c
Merge #1736
1736: Switch to `@types/vscode` and `vscode-test` r=matklad a=bkchr

The old `vscode` package is outdated and it is recommened to switch to
these two new packages. This also solves a problem of a missing `.d.ts`
for `vscode` in Nixos.

Co-authored-by: Bastian Köcher <git@kchr.de>
2019-08-26 08:07:59 +00:00
Bastian Köcher
b58f84626f Switch to @types/vscode and vscode-test
The old `vscode` package is outdated and it is recommened to switch to
these two new packages. This also solves a problem of a missing `.d.ts`
for `vscode` in Nixos.
2019-08-26 08:22:48 +02:00
bors[bot]
1fbe5ffba8
Merge #1735
1735: ⬆️ vfs r=matklad a=matklad



Co-authored-by: Aleksey Kladov <aleksey.kladov@gmail.com>
2019-08-25 10:13:37 +00:00
Aleksey Kladov
ee932d464b ⬆️ vfs 2019-08-25 13:13:08 +03:00
bors[bot]
cd433ed194
Merge #1734
1734: Strip indents and empty lines in check_apply_diagnostic_fix_from_position r=matklad a=matklad



Co-authored-by: Phil Ellison <phil.j.ellison@gmail.com>
2019-08-25 09:58:54 +00:00
Phil Ellison
14a23d1bde Strip indents and empty lines in check_apply_diagnostic_fix_from_position 2019-08-25 12:56:23 +03:00
Phil Ellison
59dd30402b Specify cursor position in ok-wrapping tests, and switch to using analysis_and_position function 2019-08-25 12:56:23 +03:00
Phil Ellison
6620949cae Simplify checking return type, add new test 2019-08-25 12:56:23 +03:00
Phil Ellison
6a04e9ce14 Fix build for Diagnostic type change 2019-08-25 12:56:23 +03:00
Phil Ellison
200470692f Cast SyntaxNodePtr to AstPtr directly 2019-08-25 12:56:23 +03:00
Phil Ellison
c8911e872e Remove reliance on expr ordering 2019-08-25 12:56:23 +03:00
Phil Ellison
4f6f3933ec cargo format 2019-08-25 12:55:56 +03:00
Phil Ellison
456e72c4e4 Change test to not rely on trait inference 2019-08-25 12:55:56 +03:00
Phil Ellison
a40e390860 Check type rather than just name in ok-wrapping diagnostic. Add test for handling generic functions (which currently fails) 2019-08-25 12:55:56 +03:00
Phil Ellison
62c2002e2b Add test that ok-wrapping handles type aliases 2019-08-25 12:55:55 +03:00
Phil Ellison
d025016f92 Mock std String and Result types in tests for ok-wrapping diagnostic 2019-08-25 12:55:55 +03:00
Phil Ellison
bacb938ab0 Add type_mismatches to InferenceResult and use this in ok-wrapping code fix 2019-08-25 12:55:55 +03:00
Phil Ellison
d00a285fa7 Initial implementation of Ok-wrapping 2019-08-25 12:55:55 +03:00
bors[bot]
866b41ddd8
Merge #1733
1733: Parse arbitrarily complex `box` patterns. r=matklad a=ecstatic-morse

This fully resolves the pattern part of #1412 by enabling the parsing of complex `box` patterns such as:

```rust
let box Struct { box i, j: box Inner(box &x) } = todo!();
```

This introduces a new `ast::BoxPat` (in the mold of `ast::RefPat`) that gets translated to `hir::Pat::Missing`.

Co-authored-by: Dylan MacKenzie <ecstaticmorse@gmail.com>
2019-08-25 09:47:44 +00:00
Dylan MacKenzie
c93903e9c7 Generate and bless tests 2019-08-23 16:24:30 -07:00
Dylan MacKenzie
c08ad1cf8a Handle Struct { box i } syntax
Named structs can have `box` patterns that will bind to their fields.
This is similar to the behavior of the `ref` and `mut` fields, but is at
least a little bit surprising.
2019-08-23 16:20:18 -07:00
Dylan MacKenzie
83433cd1f0 Centralize box pattern tests in patterns.rs 2019-08-23 16:19:53 -07:00
Dylan MacKenzie
1e991f2eaf Convert BoxPat to Missing in HIR 2019-08-23 16:07:11 -07:00
Dylan MacKenzie
ab51f57e1d Parse BoxPat 2019-08-23 16:07:11 -07:00
Dylan MacKenzie
e6f3790106 Add BoxPat variant 2019-08-23 15:11:34 -07:00
bors[bot]
fdece911fe
Merge #1732
1732: remove wrong config from .travis.yml r=matklad a=matklad



Co-authored-by: Aleksey Kladov <aleksey.kladov@gmail.com>
2019-08-23 16:34:54 +00:00
Aleksey Kladov
c75f5d1190 remove wrong config from .travis.yml
The right key here is target_branch, not branch.

However, this works anyway b/c gh-pages is the default.
2019-08-23 19:33:57 +03:00
bors[bot]
a832a2f7dd
Merge #1731
1731: rename pos_field -> tuple_field r=matklad a=matklad



Co-authored-by: Aleksey Kladov <aleksey.kladov@gmail.com>
2019-08-23 14:01:06 +00:00
Aleksey Kladov
5b18a4eef9 rename struct -> record, pos -> tuple 2019-08-23 16:59:50 +03:00
bors[bot]
c12dce0073
Merge #1716
1716: make ast object safe r=matklad a=matklad



Co-authored-by: Aleksey Kladov <aleksey.kladov@gmail.com>
2019-08-23 12:07:56 +00:00
Aleksey Kladov
bbcca4f735 make ast object safe 2019-08-23 15:06:47 +03:00
bors[bot]
e055cfacdf
Merge #1724
1724: Refactor fill_match_arms assist to use AstBuilder facilities r=matklad a=viorina



Co-authored-by: Ekaterina Babshukova <ekaterina.babshukova@yandex.ru>
2019-08-23 05:43:32 +00:00
Ekaterina Babshukova
e84f93cb5b refactor fill_match_arms assist 2019-08-23 00:43:12 +03:00
bors[bot]
0c35d82329
Merge #1721 #1723
1721: Impl/dyn trait r=flodiebold a=flodiebold

This adds support for `impl Trait` and `dyn Trait` types as far as possible without Chalk. So we can represent them in the type system, and handle them in method resolution, but Chalk doesn't know about them yet. There's a small temporary hack here where we bypass Chalk during method resolution, so we can handle simple method calls on them and completion works.

Fixes #1608.

1723: Make sysroot use `RUST_SRC_PATH` if set r=matklad a=bkchr



Co-authored-by: Florian Diebold <flodiebold@gmail.com>
Co-authored-by: Bastian Köcher <git@kchr.de>
2019-08-22 20:06:32 +00:00
Bastian Köcher
18c7a1ebe7 Make sysroot use RUST_SRC_PATH if set 2019-08-22 21:59:23 +02:00
bors[bot]
daf1a96ff0
Merge #1722
1722: Parse `box` keyword in patterns below top-level r=matklad a=ecstatic-morse

This extends the parser to handle patterns like `if let Some(box x) = ...` where the `box` keyword is not at the top-level. The last line of the added test caused a `ParseError`. This is a variant of #1412 which was not fixed by #1414.

~~I'm not familiar with `rust-analyzer`, otherwise I would fix this as well 😄.~~

Co-authored-by: Dylan MacKenzie <ecstaticmorse@gmail.com>
2019-08-22 19:58:35 +00:00
Florian Diebold
4768f5e717 Improve/fix type bound lowering 2019-08-22 21:58:29 +02:00
Florian Diebold
b1a40042e8 Handle impl/dyn Trait in method resolution
When we have one of these, the `Trait` doesn't need to be in scope to call its
methods. So we need to consider this when looking for method
candidates. (Actually I think the same is true when we have a bound `T:
some::Trait`, but we don't handle that yet).

At the same time, since Chalk doesn't handle these types yet, add a small hack
to skip Chalk in method resolution and just consider `impl Trait: Trait` always
true. This is enough to e.g. get completions for `impl Trait`, but since we
don't do any unification we won't infer the return type of e.g. `impl
Into<i64>::into()`.
2019-08-22 21:55:11 +02:00