Commit Graph

6217 Commits

Author SHA1 Message Date
Edwin Cheng
07de8ea165 invert if 2020-05-05 23:01:07 +08:00
Edwin Cheng
cf821cc332 Change union resolve types only 2020-05-05 01:19:54 +08:00
Edwin Cheng
e921195d93 Change favor_types to has_constructor 2020-05-05 01:15:27 +08:00
Edwin Cheng
5ae18f4f81 Remove unused import 2020-05-05 00:34:54 +08:00
Edwin Cheng
204c8de3bf Add test for issue 2020-05-05 00:29:47 +08:00
Edwin Cheng
af001677d1 Fix test 2020-05-05 00:17:30 +08:00
Edwin Cheng
8d96db3417 Favor types for Record in name resolution 2020-05-05 00:17:22 +08:00
bors[bot]
6a48a94d47
Merge #4268
4268: Support auto-import in macro r=SomeoneToIgnore a=edwin0cheng

Fixed: #3854

Co-authored-by: Edwin Cheng <edwin0cheng@gmail.com>
2020-05-03 22:52:10 +00:00
bors[bot]
baa35b02b2
Merge #4280
4280: Add documents owner for ImplDef and SourceFile r=matklad a=edwin0cheng

When working on #3182, I found that `ImplDef` and `SourceFile` do not implemet `DocCommentsOwer` trait, and I tested it in `cargo doc` that `impl` could has some doc-comments.

I am not so sure about `SourceFile` case, but in theory if that file is a crate root, the doc comment of it should represent the whole crate documentation, right ?

Co-authored-by: Edwin Cheng <edwin0cheng@gmail.com>
2020-05-03 10:46:07 +00:00
Edwin Cheng
f90fbaf6a6 Add documents owner for ImplDef and SourceFile 2020-05-03 18:00:27 +08:00
bors[bot]
76c2f4ef49
Merge #4278
4278: Log panics in apply_document_changes r=matklad a=lnicola

This doesn't necessarily help (because of https://github.com/rust-analyzer/rust-analyzer/issues/4263#issuecomment-623078531), but maybe we could leave it in there for a while in case it catches something.

Co-authored-by: Laurențiu Nicola <lnicola@dend.ro>
2020-05-03 09:48:44 +00:00
Laurențiu Nicola
074d1ac2f7 Log panics in apply_document_changes 2020-05-03 12:46:20 +03:00
bors[bot]
682c079043
Merge #4276
4276: Don't count start of non-ASCII characters as being inside of them r=matklad a=lnicola

I'm still not sure that `utf16_to_utf8_col` is correct for code points from Supplementary Planes. These have two UTF-16 code units, and I feel we're not going to count them correctly.

Fixes the crash in https://github.com/rust-analyzer/rust-analyzer/issues/4263#issuecomment-622988258.

Co-authored-by: Laurențiu Nicola <lnicola@dend.ro>
2020-05-03 08:57:02 +00:00
Laurențiu Nicola
16d3bb9106 Don't count start of non-ASCII characters as being inside of them 2020-05-03 09:54:39 +03:00
Kirill Bulatov
2fd054f276 Fix derive argument parsing 2020-05-02 23:47:51 +03:00
Kirill Bulatov
9f0ed7ed92 Separate macros completion from other 2020-05-02 22:33:23 +03:00
Edwin Cheng
2034002413 Support auto-import in macro 2020-05-03 03:24:55 +08:00
Kirill Bulatov
fee74851b0 Propose custom derives in completion 2020-05-02 22:24:27 +03:00
Kirill Bulatov
767bff89ed Complete standard derives 2020-05-02 21:41:02 +03:00
Kirill Bulatov
c4b32d1534 Fix the extension method 2020-05-02 21:41:02 +03:00
bors[bot]
89e1f97515
Merge #4207 #4253
4207: Add unwrap block assist #4156 r=matklad a=bnjjj

close issue #4156 

4253: Remove `workspaceLoaded` setting r=matklad a=eminence

The `workspaceLoaded` notification setting was originally designed to
control the display of a popup message that said:

    "workspace loaded, {} rust packages"

This popup was removed and replaced by a much sleeker message in the
VSCode status bar that provides a real-time status while loading:

    rust-analyzer: {}/{} packages

This was done as part of #3587

The change in this PR simply renames this setting from `workspaceLoaded` to
`progress` to better describe what it actually controls.  At the moment,
the only type of progress message that is controlled by this setting is
the initial load messages, but in theory other messages could also be
controlled by this setting.


Reviewer notes:

* If we didn't like the idea of causing minor breaking to user's config, we could keep the setting name as `workspaceLoaded`
* I think we can now close both #2719 and #3176 since the notification dialog in question no longer exists (actually I think you can close those issues even if you reject this PR 😄 )

Co-authored-by: Benjamin Coenen <5719034+bnjjj@users.noreply.github.com>
Co-authored-by: Andrew Chin <achin@eminence32.net>
2020-05-02 12:44:55 +00:00
Aleksey Kladov
359d3be308 Fix parsing of blocks without { 2020-05-02 14:35:17 +02:00
Benjamin Coenen
fdf86637bf Merge branch 'master' of github.com:rust-analyzer/rust-analyzer 2020-05-02 13:39:05 +02:00
Benjamin Coenen
6d5f3922f7 Add unwrap block assist #4156
Signed-off-by: Benjamin Coenen <5719034+bnjjj@users.noreply.github.com>
2020-05-02 12:31:11 +02:00
bors[bot]
fb8fb65131
Merge #4234
4234: Support local_inner_macros r=jonas-schievink a=edwin0cheng

This PR implements `#[macro_export(local_inner_macros)]` support. 

Note that the rustc implementation is quite [hacky][1] too. :)

[1]: 614f273e93/src/librustc_resolve/macros.rs (L456)

Co-authored-by: Edwin Cheng <edwin0cheng@gmail.com>
2020-05-02 10:30:49 +00:00
Benjamin Coenen
0b40876b99 Merge branch 'master' of github.com:rust-analyzer/rust-analyzer 2020-05-02 12:25:04 +02:00
Benjamin Coenen
eea21738ab Add unwrap block assist #4156
Signed-off-by: Benjamin Coenen <5719034+bnjjj@users.noreply.github.com>
2020-05-02 12:20:39 +02:00
bors[bot]
75bc024946
Merge #4257
4257: ast::EffectExpr r=matklad a=matklad

closes #4230 

Co-authored-by: Aleksey Kladov <aleksey.kladov@gmail.com>
2020-05-02 09:57:45 +00:00
Aleksey Kladov
b73dbbfbf2 Add missing members generates indented blocks 2020-05-02 11:53:07 +02:00
bors[bot]
235728319f
Merge #4256
4256: Improve formatting of analyzer status text r=flodiebold a=eminence

The old formatting had everything on 1 line, making it quite hard to read:

    requests:    1 textDocument/documentSymbol         2ms    2 textDocument/codeAction             2ms    3 rust-analyzer/inlayHints            20ms    4 textDocument/foldingRange           108ms    6 textDocument/codeLens               66ms    5 textDocument/semanticTokens/range   76ms    8 rust-analyzer/inlayHints            195ms    7 textDocument/semanticTokens         250ms    9 textDocument/semanticTokens/range   108ms


It now looks like this:

```
requests:
*   1 textDocument/documentSymbol         11ms
    2 textDocument/codeAction             15ms
    3 rust-analyzer/inlayHints            4ms
    5 textDocument/foldingRange           3ms
    4 textDocument/semanticTokens/range   45ms
    6 textDocument/codeLens               182ms
    8 rust-analyzer/inlayHints            124ms
    7 textDocument/semanticTokens         127ms
    9 textDocument/documentHighlight      2ms
   10 textDocument/codeAction             3ms
```

Co-authored-by: Andrew Chin <achin@eminence32.net>
2020-05-02 09:45:37 +00:00
Aleksey Kladov
b58dfd24f1 Add smoke test for decorated blocks 2020-05-02 11:27:28 +02:00
Aleksey Kladov
623faefcda Cleanup inline tests 2020-05-02 11:21:39 +02:00
Aleksey Kladov
642a3392d9 Update test data 2020-05-02 11:21:39 +02:00
Aleksey Kladov
4f2134cc33 Introduce EffectExpr 2020-05-02 11:21:39 +02:00
Aleksey Kladov
247d32cbfd Test/check the whole package
Closes #4255
2020-05-02 10:57:37 +02:00
Edwin Cheng
edf0b4c152 Test whether it is bang macro properly 2020-05-02 10:16:26 +08:00
Edwin Cheng
291d03949b Add test in name resolutions 2020-05-02 10:06:17 +08:00
Edwin Cheng
443762cda9
Update crates/ra_hir_expand/src/builtin_derive.rs
Fix spacing

Co-authored-by: Jonas Schievink <jonasschievink@gmail.com>
2020-05-02 09:56:38 +08:00
Andrew Chin
65234e8828 Remove workspaceLoaded setting
The `workspaceLoaded` notification setting was originally designed to
control the display of a popup message that said:

  "workspace loaded, {} rust packages"

This popup was removed and replaced by a much sleeker message in the
VSCode status bar that provides a real-time status while loading:

  rust-analyzer: {}/{} packages

This was done as part of #3587

The new status-bar indicator is unobtrusive and shouldn't need to be
disabled.  So this setting is removed.
2020-05-01 21:04:41 -04:00
Andrew Chin
76f34a15e6 Improve formatting of analyzer status text 2020-05-01 19:13:52 -04:00
Aleksey Kladov
fd030f9450 Revert "Merge #4233"
This reverts commit a5f2b16366, reversing
changes made to c96b2180c1.
2020-05-02 01:12:37 +02:00
kjeremy
810c961961 Update crates 2020-05-01 18:26:42 -04:00
bors[bot]
3232fd5179
Merge #4220 #4240
4220: Introduce LowerCtx r=matklad a=edwin0cheng

This PR introduces `LowerCtx` for path lowering. 

After this PR, there are only 2 places remains for using deprecated `Path::from_ast`, which is related to `AstTransform` I am not familiar. I would like to change these in another PR by others ;)

related disscusiion:  https://rust-lang.zulipchat.com/#narrow/stream/185405-t-compiler.2Fwg-rls-2.2E0/topic/Path.3A.3Afrom_src

And also fixed part of https://github.com/rust-analyzer/rust-analyzer/issues/4176#issuecomment-620672930

4240: Bump deps r=matklad a=lnicola



Co-authored-by: Edwin Cheng <edwin0cheng@gmail.com>
Co-authored-by: Laurențiu Nicola <lnicola@dend.ro>
2020-05-01 20:16:25 +00:00
bors[bot]
8d02b24b25
Merge #4244
4244: Show unsafe trait in hover r=matklad a=DianaNites

Following on #2450 and #4210, for traits.
`unsafe` is the only qualifier they can have, though.

Co-authored-by: Diana <5275194+DianaNites@users.noreply.github.com>
2020-05-01 19:45:48 +00:00
bors[bot]
21588e15df
Merge #4246
4246: Validate uses of self and super r=matklad a=djrenren

This change follows on the validation of the `crate` keyword in paths. It verifies the following things:

`super`:
 - May only be preceded by other `super` segments
 - If in a `UseItem` then all semantically preceding paths also consist only of `super`

`self`
 - May only be the start of a path


Just a note, a couple times while working on this I found myself really wanting a Visitor of some sort so that I could traverse descendants while skipping sub-trees that are unimportant. Iterators don't really work for this, so as you can see I reached for recursion. Considering paths are generally small a fancy debounced visitor probably isn't important but figured I'd say something in case we had something like this lying around and I wasn't using it.

Co-authored-by: John Renner <john@jrenner.net>
2020-05-01 19:24:25 +00:00
kjeremy
99826dab15 Address comments 2020-05-01 14:12:31 -04:00
kjeremy
ee1628dba0 Mark most assists as the base "refactor" type
Most of them area. We will separate them out later but this gets them to
show up in the "refactor" menu of vscode.
2020-05-01 13:44:32 -04:00
kjeremy
e75565c73f Advertise support for all the builtin CodeActionKinds
Even thought we don't return all of these we eventually will so might as
well advertise now.
2020-05-01 13:44:32 -04:00
kjeremy
e3ee61f5e8 Filter out CodeActions if a server only support commands. 2020-05-01 13:44:32 -04:00
Diana
375dd18dc0 Fix pub(self) visibility?
Clippy complained about it and it seems wrong
2020-05-01 12:09:47 -04:00