8745: Support goto_type_definition for types r=matklad a=Veykril
I'm unsure if the approach of lowering an `ast::Type` to a `hir::Type` is a good idea, it seems fine to me at least.
Fixes#2882
Co-authored-by: Lukas Tobias Wirth <lukastw97@gmail.com>
8280: Borrow text of immutable syntax node r=iDawer a=iDawer
In https://github.com/rust-analyzer/rowan/pull/101 `rowan::SyntaxNode::green` returns `Cow<'_, GreenNodeData>`. It returns borrow of green node of immutable syntax tree node.
Using this we can return borrowed text from `ast::Name::text`.
~~However now it allocates in case of mutable syntax trees.~~ (see next comment)
The idea comes from https://github.com/rust-analyzer/rowan/pull/100#issuecomment-809330325
Co-authored-by: Dawer <7803845+iDawer@users.noreply.github.com>
8674: fix for #8664: Emit folding ranges for multi-line where clauses r=matklad a=m5tfi
#8664
I added a test that assert folding multi-line where clauses while leaving single lined one. Please, let me know if the code needs further improvements.
Co-authored-by: m5tfi <72708423+m5tfi@users.noreply.github.com>
8711: Only resolve selected assist r=matklad a=SomeoneToIgnore
Part of https://github.com/rust-analyzer/rust-analyzer/issues/8700
Now resolves only the assist that was selected out of the list, while before the whole assist list was resolved despite a single popup selection.
Co-authored-by: Kirill Bulatov <mail4score@gmail.com>
8693: Ensure that only one cache priming task can run at a time r=matklad a=Bobo1239
Fixes#8632.
Co-authored-by: Boris-Chengbiao Zhou <bobo1239@web.de>
8692: Fix panic caused by new Try trait definition r=flodiebold a=flodiebold
The new Try trait definition caused a query cycle for us. This adds recovery for that cycle, but also fixes the cause, which is that we went through the supertraits when resolving `<T as Trait>::Assoc`, which isn't actually necessary. I also rewrote `all_super_trait_refs` to an iterator before I realized what the actual problem was, so I kept that.
Fixes#8686.
Co-authored-by: Florian Diebold <flodiebold@gmail.com>
According to the spec we should return ServerNotInitialized if the server is waiting for an initialize request and something else comes in.
Upgrading to lsp-server 0.5.1 will do this and retry until the initialize request comes in.
Fixes#8581
At the moment,the popup is just a bazillion of Cargo's "Compiling this\nCompiling that",
which is not that useful.
--quiet still displays error, which is what we needc
Attempting to rename an element of a tuple field would previously
replace the type with the new name, which doesn't make sense; now it
fails instead.
The check is done in both `prepare_rename` and `rename` so that the case
is caught before the user is prompted for a new name. Some other
existing failure cases are also now additionally checked in
`prepare_rename`.
8591: Remove SyntaxRewriter usage in insert_use in favor of mutable syntax trees r=matklad a=Veykril
Unfortunately changing `insert_use` to not use `SyntaxRewriter` creates a lot of changes since so much relies on that. But on the other hand this should be the biggest usage of `SyntaxRewriter` I believe.
8638: Remove SyntaxRewriter::from_fn r=Veykril a=Veykril
Co-authored-by: Lukas Wirth <lukastw97@gmail.com>
8620: Remove unnecessary braces for extracted block expression r=Veykril a=brandondong
This change addresses the first bullet point of https://github.com/rust-analyzer/rust-analyzer/issues/7839.
Specifically, when extracting block expressions, remove the unneeded extra braces inside the generated function.
Co-authored-by: Brandon <brandondong604@hotmail.com>
8570: Flycheck tries to parse both Cargo and Rustc messages. r=rickvanprim a=rickvanprim
This change allows non-Cargo build systems to be used for Flycheck provided they call `rustc` with `--error-format=json` and emit those JSON messages to `stdout`.
Co-authored-by: James Leitch <rickvanprim@gmail.com>
reading both stdout & stderr is a common gotcha, you need to drain them
concurrently to avoid deadlocks. Not sure why I didn't do the right
thing from the start. Seems like I assumed the stderr is short? That's
not the case when cargo spams `compiling xyz` messages
8524: Fix extract function with partial block selection r=matklad a=brandondong
**Reproduction:**
```rust
fn foo() {
let n = 1;
let mut v = $0n * n;$0
v += 1;
}
```
1. Select the snippet ($0) and use the "Extract into function" assist.
2. Extracted function is incorrect and does not compile:
```rust
fn foo() {
let n = 1;
let mut v = fun_name(n);
v += 1;
}
fn fun_name(n: i32) {}
```
3. Omitting the ending semicolon from the selection fixes the extracted function:
```rust
fn fun_name(n: i32) -> i32 {
n * n
}
```
**Cause:**
- When `extraction_target` uses a block extraction (semicolon case) instead of an expression extraction (no semicolon case), the user selection is directly used as the TextRange.
- However, the existing function extraction logic for blocks requires that the TextRange spans from start to end of complete statements to work correctly.
- For example:
```rust
fn foo() {
let m = 2;
let n = 1;
let mut v = m $0* n;
let mut w = 3;$0
v += 1;
w += 1;
}
```
produces
```rust
fn foo() {
let m = 2;
let n = 1;
let mut v = m let mut w = fun_name(n);
v += 1;
w += 1;
}
fn fun_name(n: i32) -> i32 {
let mut w = 3;
w
}
```
- The user selected TextRange is directly replaced by the function call which is now in the middle of another statement. The extracted function body only contains statements that were fully covered by the TextRange and so the `* n` code is deleted. The logic for calculating variable usage and outlived variables for the function parameters and return type respectively search within the TextRange and so do not include `m` or `v`.
**Fix:**
- Only extract full statements when using block extraction. If a user selected part of a statement, extract that full statement.
8527: Switch introduce_named_lifetime assist to use mutable syntax tree r=matklad a=iDawer
This extends `GenericParamsOwnerEdit` trait with `get_or_create_generic_param_list` method
Co-authored-by: Brandon <brandondong604@hotmail.com>
Co-authored-by: Dawer <7803845+iDawer@users.noreply.github.com>
8565: Fill match arms assist: add remaining arms for tuple of enums r=iDawer a=iDawer
Fix for #8493
However, the assist is still flaky and does not use `hir_ty::diagnostics::match_check`
Co-authored-by: Dawer <7803845+iDawer@users.noreply.github.com>
8540: Prevent being able to rename items that are not part of the workspace r=Veykril a=Veykril
This change causes renames that happen on items coming from crates outside the workspace to fail. I believe this should be the right approach, but usage of cargo's workspace might not be entirely correct for preventing these kinds of refactoring from touching things they shouldn't. I'm not entirely sure?
cc #6623, this is one of the bigger footguns when it comes to refactoring, especially in combination with import aliases people tend to rename items coming from a crates dependency which this prevents.
Co-authored-by: Lukas Wirth <lukastw97@gmail.com>
8467: Adds impl Deref assist r=jhgg a=jhgg
This PR adds a new `generate_deref` assist that automatically generates a deref impl for a given struct field.
Check out this gif:
--
I have a few Q's:
- [x] Should I write more tests, if so, what precisely should I test for?
- [x] I have an inline question on line 65, can someone provide guidance? :)
- [x] I can implement this for `ast::TupleField` too. But should it be a separate assist fn, or should I try and jam both into the `generate_deref`?
- [x] I want to follow this up with an assist on `impl $0Deref for T {` which would automatically generate a `DerefMut` impl that mirrors the Deref as well, however, I could probably use some pointers on how to do that, since I'll have to reach into the ast of `fn deref` to grab the field that it's referencing for the `DerefMut` impl.
Co-authored-by: jake <jh@discordapp.com>
8560: Escape characters in doc comments in macros correctly r=jonas-schievink a=ChayimFriedman2
Previously they were escaped twice, both by `.escape_default()` and the debug view of strings (`{:?}`). This leads to things like newlines or tabs in documentation comments being `\\n`, but we unescape literals only once, ending up with `\n`.
This was hard to spot because CMark unescaped them (at least for `'` and `"`), but it did not do so in code blocks.
This also was the root cause of #7781. This issue was solved by using `.escape_debug()` instead of `.escape_default()`, but the real issue remained.
We can bring the `.escape_default()` back by now, however I didn't do it because it is probably slower than `.escape_debug()` (more work to do), and also in order to change the code the least.
Example (the keyword and primitive docs are `include!()`d at https://doc.rust-lang.org/src/std/lib.rs.html#570-578, and thus originate from macro):
Before:
After:
Co-authored-by: Chayim Refael Friedman <chayimfr@gmail.com>
Previously they were escaped twice, both by `.escape_default()` and the debug view of strings (`{:?}`). This leads to things like newlines or tabs in documentation comments being `\\n`, but we unescape literals only once, ending up with `\n`.
This was hard to spot because CMark unescaped them (at least for `'` and `"`), but it did not do so in code blocks.
This also was the root cause of #7781. This issue was solved by using `.escape_debug()` instead of `.escape_default()`, but the real issue remained.
We can bring the `.escape_default()` back by now, however I didn't do it because it is probably slower than `.escape_debug()` (more work to do), and also in order to change the code the least.
8510: Move cursor position when using item movers r=jonas-schievink a=jonas-schievink
This updates the cursor position when moving items around to stay in the same location within the moved node.
I changed the `moveItem` response to `SnippetTextEdit[]`, since that made more sense to me (the file was ignored by the client anyways, since the edits always apply to the current document). It also matches `onEnter`, which seems logical to me, but please let me know if this doesn't make sense.
There's still a bug in the client-side snippet code that will cause the cursor position to be slightly off when moving parameters in the same line (presumably we don't track the column correctly after deleting `$0`). Not really sure how to fix that immediately, but this PR should already be an improvement despite that bug.
8533: Fix typo in style guide r=jonas-schievink a=jonas-schievink
Fixes bold text rendering
bors r+
Co-authored-by: Jonas Schievink <jonasschievink@gmail.com>
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>
8354: Distinguishing between different operators in semantic highlighting r=matklad a=chetankhilosiya
Co-authored-by: Chetan Khilosiya <chetan.khilosiya@gmail.com>
8415: Fix faulty assertion when extracting function with macro call r=matklad a=brandondong
**Reproduction:**
```rust
fn main() {
let n = 1;
let k = n * n;
dbg!(n);
}
```
1. Select the second and third lines of the main function. Use the "Extract into function" code assist.
2. Panic occurs in debug, error is logged in release: "[ERROR ide_assists::handlers::extract_function] assertion failed: matches!(path, ast :: Expr :: PathExpr(_))".
3. Function generates successfully on release where the panic was bypassed.
```rust
fn fun_name(n: i32) {
let k = n * n;
dbg!(n);
}
```
**Cause:**
- The generated function will take `n` as a parameter. The extraction logic needs to search the usages of `n` to determine whether it is used mutably or not. The helper `path_element_of_reference` is called for each usage but the second usage is a macro call and fails the `Expr::PathExpr(_)` match assertion.
- The caller of `path_element_of_reference` does implicitly assume it to be a `Expr::PathExpr(_)` in how it looks at its parent node for determining whether it is used mutably. This logic will not work for macros.
- I'm not sure if there are any other cases besides macros where it could be something other than a `Expr::PathExpr(_)`. I tried various examples and could not find any.
**Fix:**
- Update assertion to include the macro case.
- Add a FIXME to properly handle checking if a macro usage requires mutable access. For now, return false instead of running the existing logic that is tailored for `Expr::PathExpr(_)`'s.
Co-authored-by: Brandon <brandondong604@hotmail.com>
Conceptually, using a *message* here is wrong, because this is a
"status", rather than "point in time" thing. But statuses are an LSP
extension, while messages are stable. As a compromise, send message only
for more critical `metadata` failures, and only once per state change.
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.
