| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|\ \ \ \ \ \ \
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
9292: fix: Code: clean-up #8951 r=wxb1ank a=wxb1ank
#8951 was a major change in the VS Code extension and caused quite a few problems. This PR is a catch-all for bugs and improvements in the new code.
This should fix:
- #9284
- [this unreported bug](https://github.com/rust-analyzer/rust-analyzer/pull/8951/files#r651570446)
- ...and one or two uncaught exceptions I just found
The original lack of testing was my own fault, but this area of the VS Code API is also tricky for a couple reasons:
- The [FileSystem](https://github.com/rust-analyzer/rust-analyzer/pull/8951/files#r651570446) API does not list or warn about any exceptions, but [FileSystemProvider](https://github.com/rust-analyzer/rust-analyzer/pull/8951/files#r651570446) (which `FileSystem` is a wrapper of, AFAICT) does.
- At first glance, [Uri.path](https://github.com/rust-analyzer/rust-analyzer/pull/8951/files#r651570446) *looks* like it works for FS operations. It does not, at least, on Windows. You need to use `Uri.fsPath`.
I only use Windows, so I need people on macOS, Linux, and (possibly) NixOS to test this.
Co-authored-by: wxb1ank <[email protected]>
|
| | | | | | | | |
|
| | | | | | | | |
|
|\ \ \ \ \ \ \ \
| |_|_|/ / / / /
|/| | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
9295: fix: Highlight unsafe trait refs as unsafe only in impl blocks and definitions r=Veykril a=Veykril
bors r+
Co-authored-by: Lukas Wirth <[email protected]>
|
|/ / / / / / / |
|
|\ \ \ \ \ \ \
| | |/ / / / /
| |/| | | / /
| |_|_|_|/ /
|/| | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
9291: minor: fix typo r=matklad a=matklad
bors r+
🤖
Co-authored-by: Aleksey Kladov <[email protected]>
|
| | | | | | |
|
|\| | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
9290: internal: enforce no #[ignore] and no #[should_panic] r=matklad a=matklad
bors r+
🤖
Co-authored-by: Aleksey Kladov <[email protected]>
|
| | | | | | |
|
|/ / / / /
| | | | |
| | | | |
| | | | | |
See the style.md for motivation
|
|\ \ \ \ \
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
9288: internal: cleanup tests r=matklad a=matklad
bors r+
🤖
Co-authored-by: Aleksey Kladov <[email protected]>
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
* ensure standard, non-indented style (should add this check to
`fixture` some day)
* removed a couple of ignores
|
|\| | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
9285: minor: add a test variation r=matklad a=matklad
bors r+
🤖
Co-authored-by: Aleksey Kladov <[email protected]>
|
|/ / / / / |
|
|\ \ \ \ \
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
9278: internal: document that we don't #[ignore] tests r=matklad a=matklad
Co-authored-by: Aleksey Kladov <[email protected]>
|
| | | | | | |
|
|\ \ \ \ \ \
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
9282: fix: Use objects instead of bools for markers in package.json r=lnicola a=lnicola
Addresses https://github.com/rust-analyzer/rust-analyzer/issues/9174#issue-913980665.
Co-authored-by: Laurențiu Nicola <[email protected]>
|
| | |_|_|/ /
| |/| | | | |
|
|\ \ \ \ \ \
| |/ / / / /
|/| | / / /
| | |/ / /
| |/| | |
| | | | |
| | | | |
| | | | | |
9281: fix: Code: use `Uri.toString()` for URLs r=lnicola a=wxb1ank
I believe this should fix #9280. Testing is needed but I wanted to quickly push a hotfix.
Co-authored-by: wxb1ank <[email protected]>
|
|/ / / / |
|
|\| | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
9277: internal: more natural order of sources for TypeParam r=matklad a=matklad
bors r+
🤖
Co-authored-by: Aleksey Kladov <[email protected]>
|
| | | |
| | | |
| | | |
| | | |
| | | | |
We usually use first (left) variant of `Either` for "usual" case, and
use right for odd things. For example, pat source is Pat | SelfParam.
|
|\| | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
9276: internal: refactor diagnostics more r=matklad a=matklad
bors r+
🤖
Co-authored-by: Aleksey Kladov <[email protected]>
|
| | | | |
|
| | | | |
|
|\ \ \ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
9275: feat: Support goto type for field expressions and patterns r=Veykril a=Veykril
bors r+
Co-authored-by: Lukas Wirth <[email protected]>
|
| | | | | |
|
|\ \ \ \ \
| |_|/ / /
|/| | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
9273: internal: don't pull tracing via chalk r=matklad a=matklad
bors r+
🤖
Co-authored-by: Aleksey Kladov <[email protected]>
|
|/ / / / |
|
|\ \ \ \
| |/ / /
|/| | |
| | | |
| | | |
| | | |
| | | |
| | | | |
9272: internal: move diagnostics to a dedicated crate r=matklad a=matklad
bors r+
Co-authored-by: Aleksey Kladov <[email protected]>
|
| | | | |
|
| | | | |
|
| | | |
| | | |
| | | |
| | | |
| | | | |
It seems that any crate can be made better by flattening the modules
down to a single layer?
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| |/ / |
|
|\ \ \
| |/ /
|/| |
| | |
| | |
| | |
| | |
| | | |
9270: minor: Add assoc type in trait bound completion test r=Veykril a=Veykril
bors r+
Co-authored-by: Lukas Wirth <[email protected]>
|
| | | |
|
|\ \ \
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
9265: internal: prepare to move rename to base_db r=matklad a=matklad
bors r+
🤖
Co-authored-by: Aleksey Kladov <[email protected]>
|
|/ / /
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
It's better to handle magical cases upper in the stack, because it
allows for better re-use of the general implementation below. So, we
pull the `self` case up here.
The end goal is to put `Definition::rename` to the `ide_db`, because
it's a generally re-usable functionality useful for different ide
features, alongside with the search which is already there.
|
|\ \ \
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
9263: fix: don't use display-related functionality where semantics matters r=matklad a=matklad
bors r+
🤖
Co-authored-by: Aleksey Kladov <[email protected]>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
NavigationTarget is strictly a UI-level thing -- it describes where the
cursor should be placed when the user presses goto definition. It
doesn't make any semantic guaratees about rage and focus range and, as
such, is not suitable for driving renames.
|
|\ \ \ \
| |_|/ /
|/| | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
9262: fix: Don't keep a trailing `self` token in import paths after `unmerge_use` r=Veykril a=Veykril
Fixes #9213
bors r+
Co-authored-by: Lukas Wirth <[email protected]>
|
|/ / / |
|
|\ \ \
| |/ /
|/| |
| | |
| | |
| | |
| | |
| | | |
8951: internal: migrate to vscode.FileSystem API r=matklad a=wxb1ank
I encountered an error where `bootstrap()` attempts to create a directory with the path `C:\C:\...`. I couldn't find this reported anywhere else. Using the `vscode.FileSystem` API instead of the `fs` one works here. I assume the latter automatically prepends `C:\` to paths whereas the former does not. I don't know if this suggests `vscode.FileSystem` should be used in more places for consistency.
Co-authored-by: wxb1ank <[email protected]>
|
| | | |
|
| | | |
|