4658: Fix problem with format string tokenization r=matklad a=ruabmbua
Fixed by just not handling closing curlybrace escaping.
Closes https://github.com/rust-analyzer/rust-analyzer/issues/4637
Co-authored-by: Roland Ruckerbauer <roland.rucky@gmail.com>
4580: Fix invoking cargo without consulting CARGO env var or standard installation paths r=matklad a=Veetaha
Followup for #4329
The pr essentially fixes [this bug](https://youtu.be/EzQ7YIIo1rY?t=2189)
cc @lefticus
Co-authored-by: veetaha <veetaha2@gmail.com>
As per matklad, we now pass the responsibility for finding the binary to the frontend.
Also, added caching for finding the binary path to reduce
the amount of filesystem interactions.
4651: Use first match branch in case of type mismatch, not last r=kjeremy a=flodiebold
The comment says this was intentional, but I do agree with #4304 that it makes
more sense the other way around (for if/else as well).
Fixes#4304.
Co-authored-by: Florian Diebold <florian.diebold@freiheit.com>
4641: Upgrade Chalk r=matklad a=flodiebold
Chalk newly added TypeName::Never and Array; I implemented the conversion for
Never, but not Array since that expects a const argument.
Co-authored-by: Florian Diebold <flodiebold@gmail.com>
4592: fix textedit range returned for completion when left token is a keyword r=bnjjj a=bnjjj
close#4545
Co-authored-by: Benjamin Coenen <5719034+bnjjj@users.noreply.github.com>
4596: Strip leading underscores of argument names in function/method r=matklad a=kuy
Closes#4510
### Goal
When I select a function/method from completions, I get a snippet that doesn't contain leading underscores of argument names.
### Solution
- Option 1: All signatures don't contain underscores
- Option 2: Keep same signature, but inserted snippet doesn't contain underscores
I choose Option 2 because I think that leading underscores is a part of "signature". Users should get correct signatures. On the other hand, trimming underscores is an assist by IDE.
### Other impls.
rls: Complete argument names with underscores (same as actual ra)
IntelliJ Rust: Doesn't complete argument names
VSCode (TypeScript): Doesn't complete argument names
### Working example
Co-authored-by: Yuki Kodama <endflow.net@gmail.com>
4625: Partially fix displaying inlay hints in Github PR diff views r=matklad a=Veetaha
See the comment in https://github.com/rust-analyzer/rust-analyzer/issues/4608#issuecomment-63424257
It partially fixes the left side of diff view (the one where old code is displayed), but the diff editor with new code changes still has `file` scheme and will proceed displaying inlay hints...
4629: Fix the `should_panic` snippet r=matklad a=eminence
Closes#4628
Co-authored-by: veetaha <veetaha2@gmail.com>
Co-authored-by: Andrew Chin <achin@eminence32.net>
4534: Add call postfix completion r=matklad a=vain0x
To make it easier to wrap an expression with Ok/Some/Rc::new etc.
Note I agree with conclusion of the discussion in #1431 that adding many completions is not the way to go. However, this PR still could be justified due to versatility of use.
Co-authored-by: vain0x <vainzerox@gmail.com>
The line separator is moved below the function signature to split
regions between the docs. This is very similar to how IntelliJ
displays tooltips. Adding an additional separator between the module
name and function signature currently has rendering issues.
Fixes#4594
Alternative to #4615
4602: Add boolean literal semantic token type to package.json r=matklad a=lnicola
Closes#4583.
CC @GrayJack
4603: Add self keyword semantic token type r=matklad a=lnicola
Not sure if this is warranted a new token type or just a modifier.
---
CC #4583, @GrayJack
Co-authored-by: Laurențiu Nicola <lnicola@dend.ro>
* Done at line 243: "Add validation of `crate` keyword not appearing in the middle of the symbol path"
* Already happened: "Remove validation of unterminated literals (it is already implemented in `tokenize()`)"
* Happens in `unescape()`: "Add validation of character literal containing only a single char"
* Missing: "raw string literals and raw byte string literals"
4555: VSCode: added patchelf after download for NixOS support r=matklad a=cab404
This adds Nix support, and fixes#4542
4575: Use Chalk's built-in representations for fn items and pointers r=matklad a=flodiebold
The `TypeName::FnDef` was just added; the function pointer variant has existed for a while, I just forgot about it because it's special (because fn pointers can be higher-ranked over lifetimes).
We *could* also make `FnPtr` a separate `Ty` variant instead of a `TypeCtor` variant, which would make the conversion code a bit less special-casey, but it doesn't seem worth doing right now.
Co-authored-by: Vladimir Serov <me@cab404.ru>
Co-authored-by: Cabia Rangris <me@cab404.ru>
Co-authored-by: Florian Diebold <florian.diebold@freiheit.com>
Function pointers can be 'higher-ranked' over lifetimes, which is why they're
not an application type in Chalk, but since we don't model lifetimes it doesn't
matter for us yet.
4570: Use Chalk's built-in impls r=matklad a=flodiebold
This contains two changes:
- Chalk has begun adding built-in representations of primitive types; use these in our type conversion logic. There's one somewhat 'iffy' part here, namely references; we don't keep track of lifetimes, but Chalk does, so it will expect a lifetime parameter on references. If we didn't provide that, it could cause crashes in Chalk code that expects the lifetime, so I rather hackily add an (always the same) lifetime placeholder during conversion. I expect that we'll fully switch to using Chalk's types everywhere before we add lifetime support, so I think this is the best solution for now.
- let Chalk know about well-known traits (from lang items), so it can apply its built-in impls.
Before:
```
Total expressions: 181485
Expressions of unknown type: 2940 (1%)
Expressions of partially unknown type: 2884 (1%)
Type mismatches: 901
Inference: 37.821210245s, 0b allocated 0b resident
Total: 53.399467609s, 0b allocated 0b resident
```
After:
```
Total expressions: 181485
Expressions of unknown type: 2923 (1%)
Expressions of partially unknown type: 2879 (1%)
Type mismatches: 734
Inference: 39.157752509s, 0b allocated 0b resident
Total: 54.110767621s, 0b allocated 0b resident
```
(I will start splitting up `chalk.rs` in a separate PR, since it's getting pretty big...)
Co-authored-by: Florian Diebold <flodiebold@gmail.com>
Co-authored-by: Florian Diebold <florian.diebold@freiheit.com>
4571: KISS SourceChange r=matklad a=matklad
The idea behind requiring the label is a noble one, but we are not
really using it consistently anyway, and it should be easy to retrofit
later, should we need it.
bors r+
Co-authored-by: Aleksey Kladov <aleksey.kladov@gmail.com>
The idea behind requiring the label is a noble one, but we are not
really using it consistently anyway, and it should be easy to retrofit
later, should we need it.
4516: LSP: Two stage initialization r=kjeremy a=kjeremy
Fills in server information.
Derives CodeAction capabilities from the client. If code action literals
are unsupported we fall back to the "simple support" which just sends back
commands (this is already supported in our config). The difference being
that we did not adjust our server capabilities so that if the client was
checking for `CodeActionProvider: "true"` in the response that would have failed.
Part of #144Fixes#4130 (the specific case called out in that issue)
Co-authored-by: kjeremy <kjeremy@gmail.com>
This also changes our handiling of snippet edits on the client side.
`editor.insertSnippet` unfortunately forces indentation, which we
really don't want to have to deal with. So, let's just implement our
manual hacky way of dealing with a simple subset of snippets we
actually use in rust-analyzer
4506: Make `find_path_inner` a query r=matklad a=jonas-schievink
This eliminates the remaining performance problems in the "Implement default members" assist (at least those that I've found).
Closes https://github.com/rust-analyzer/rust-analyzer/issues/4498
Co-authored-by: Jonas Schievink <jonasschievink@gmail.com>
4526: Use a flat play icon instead of the blue emoji with test code lens r=kjeremy a=aloucks
@lnicola
Restores this commit:
55e914a2a1
That was effectively wiped out by this code formatting commit:
dc217bdf903d445256fe
Co-authored-by: Aaron Loucks <aloucks@cofront.net>
Fills in server information.
Derives CodeAction capabilities from the client. If code action literals
are unsupported we fall back to the "simple support" which just sends back
commands (this is already supported in our config). The difference being
that we did not adjust our server capabilities so that if the client was
checking for `CodeActionProvider: "true"` in the response that would have failed.
4501: Querify `importable_locations_in_crate` r=jonas-schievink a=jonas-schievink
This brings the time needed to compute the `add_missing_impl_members` assist down from ~5 minutes to 20 seconds on my test workload (which is editing within an impl of a MIR [`MutVisitor`](https://doc.rust-lang.org/nightly/nightly-rustc/rustc_middle/mir/visit/trait.MutVisitor.html))
cc #4498
Co-authored-by: Jonas Schievink <jonasschievink@gmail.com>
4497: Create LowerCtx on the fly r=matklad a=edwin0cheng
Previously we create `LowerCtx` at the beginning of lowering, however, the hygiene content is in fact changing between macro expression expanding.
This PR change it to create the `LowerCtx` on the fly to fix above bug.
However, #4465 is not fixed by this PR, the goto-def is still not work yet. It only fixed the infer part.
Co-authored-by: Edwin Cheng <edwin0cheng@gmail.com>
4493: Provide builtin impls of Fn traits for fn-pointers r=flodiebold a=hban
Meant to be, but isn't actually a fix for #2880.
Consider this snippet:
```rust
use std::marker::PhantomData;
use std::ops::Deref;
struct Lazy<T, F/* = fn() -> T*/>(F, PhantomData<T>);
impl<T, F> Lazy<T, F> {
pub fn new(f: F) -> Lazy<T, F> {
Lazy(f, PhantomData)
}
}
impl<T, F: FnOnce() -> T> Deref for Lazy<T, F> {
type Target = T;
fn deref(&self) -> &T { todo!() }
}
fn test() {
let lazy1: Lazy<u32, _> = Lazy::new(|| 0u32);
let r1 = lazy1.to_string();
fn make_u32_fn() -> u32 { todo!() }
let make_u32_fn_ptr: fn() -> u32 = make_u32_fn;
let lazy2: Lazy<u32, _> = Lazy::new(make_u32_fn_ptr);
let r2 = lazy2.to_string();
}
```
* On current master:
* When type default is commented-out, `r1` is correctly inferred, `r2` in _{unknown}_.
* When type default is not commented-out, both `r1` and `r2` are _{unknown}_.
* With this PR:
* When type default is commented-out, both `r1` and `r2` are correctly inferred.
* When type default is not commented-out, both `r1` and `r2` are _{unknown}_.
Well, it's a improvement at least. I guess this thing with type defaults is a different problem.
I also tried add Fn impls for fn items, but wasn't successful. So this PR only adds those impls for fn pointers.
Co-authored-by: Hrvoje Ban <hban@users.noreply.github.com>
4489: Memory allocation optimization r=matklad a=simonvandel
I did some profiling using DHAT, and this was what I could easily optimize without much knowledge of the codebase.
This speeds up analysis-stats on rust-analyser by ~4% on my local machine.
**Benchmark**
➜ rust-analyzer-base git:(master) hyperfine --min-runs=2 '/home/simon/Documents/rust-analyzer/target/release/rust-analyzer analysis-stats .' '/home/simon/Documents/rust-analyzer-base/target/release/rust-analyzer analysis-stats .'
Benchmark #1: /home/simon/Documents/rust-analyzer/target/release/rust-analyzer analysis-stats .
Time (mean ± σ): 49.621 s ± 0.317 s [User: 48.725 s, System: 0.792 s]
Range (min … max): 49.397 s … 49.846 s 2 runs
Benchmark #2: /home/simon/Documents/rust-analyzer-base/target/release/rust-analyzer analysis-stats .
Time (mean ± σ): 51.764 s ± 0.045 s [User: 50.882 s, System: 0.756 s]
Range (min … max): 51.733 s … 51.796 s 2 runs
Summary
'/home/simon/Documents/rust-analyzer/target/release/rust-analyzer analysis-stats .' ran
1.04 ± 0.01 times faster than '/home/simon/Documents/rust-analyzer-base/target/release/rust-analyzer analysis-stats .'
Co-authored-by: Simon Vandel Sillesen <simon.vandel@gmail.com>
4484: Allow calling dyn trait super trait methods without the super trait in scope r=flodiebold a=flodiebold
This also removes some vestiges of the old impl trait support which I think aren't currently in use.
Co-authored-by: Florian Diebold <flodiebold@gmail.com>
4472: Fix path resolution for module and function with same name r=hasali19 a=hasali19
This fixes#3970 and also fixes completion for the same issue.
Co-authored-by: Hasan Ali <git@hasali.co.uk>
4448: Generate configuration for launch.json r=vsrs a=vsrs
This PR adds two new commands: `"rust-analyzer.debug"` and `"rust-analyzer.newDebugConfig"`. The former is a supplement to the existing `"rust-analyzer.run"` command and works the same way: asks for a runnable and starts new debug session. The latter allows adding a new configuration to **launch.json** (or to update an existing one).
If the new option `"rust-analyzer.debug.useLaunchJson"` is set to true then `"rust-analyzer.debug"` and Debug Lens will first look for existing debug configuration in **launch.json**. That is, it has become possible to specify startup arguments, env variables, etc.
`"rust-analyzer.debug.useLaunchJson"` is false by default, but it might be worth making true the default value. Personally I prefer true, but I'm not sure if it is good for all value.
----
I think that this PR also solves https://github.com/rust-analyzer/rust-analyzer/issues/3441.
Both methods to update launch.json mentioned in the issue do not work:
1. Menu. It is only possible to add a launch.json configuration template via a debug adapter. And anyway it's only a template and it is impossible to specify arguments from an extension.
2. DebugConfigurationProvider. The exact opposite situation: it is possible to specify all debug session settings, but it is impossible to export these settings to launch.json.
Separate `"rust-analyzer.newDebugConfig"` command looks better for me.
----
Fixes#4450Fixes#3441
Co-authored-by: vsrs <vit@conrlab.com>
Co-authored-by: vsrs <62505555+vsrs@users.noreply.github.com>
4273: Trigger add_vis assist on paths/record fields as well r=flodiebold a=TimoFreiberg
Resolves#4037.
- [x] Function defs
- [x] ADT defs
- [x] Enum variants
- [x] Consts
- [x] Statics
- [x] Traits
- [x] Type aliases
- [x] Modules
- [x] Record fields (using different implementation)
- [x] struct fields
- [x] enum variant fields
- ❌ union fields (`Semantics::resolve_record_field` seems to not work for union fields, so I think this can be handled in a future PR)
- [x] More tests?
- [x] Improve test fixture code and documentation a bit (see [Zulip](https://rust-lang.zulipchat.com/#narrow/stream/185405-t-compiler.2Fwg-rls-2.2E0/topic/resolve_path.20between.20fixture.20files))
Co-authored-by: Timo Freiberg <timo.freiberg@gmail.com>
4445: Correctly fill default type parameters r=flodiebold a=montekki
Fixes#3877
So, basically even if the parameters are omitted from the `impl` block, check the parameters in `trait` if they have a default type, and if they do go from `hir` to `ast::TypeArg`. I've added a helper for that but I am not sure that it's a proper way to go from `hir` to `ast` here.
Co-authored-by: Fedor Sakharov <fedor.sakharov@gmail.com>
4405: Make some stuff public so that they can be reused by other tools r=pksunkara a=pksunkara
So, my little experiment of building a code analysis tool using rust-analyzer is successful. I am going to proceed to build the tool now. This PR makes the needed things public.
I know there were some things about trying to change stuff regarding loading workspaces, which would make it more easier for other tools to reuse. But, until then, it should be okay using this `load_cargo` fn.
Btw, if I were publish my tool, I would need the `ra` crates to be released. Since @matklad told me that he doesn't want to care about breaking stuff, I would propose the following.
Every monday, during the weekly release, we release a new pre v1 minor version of all the crates. That way, we don't need to care about breaking stuff but still have rust-analyzer on crates.io.
I made https://github.com/pksunkara/cargo-workspaces to help release workspace crates easily.
So, coming week, we start with `0.1.0`, then week after that, we release `0.2.0` and then `0.3.0` etc.. until we decide on `1.0.0` which is probably when the compiler team also starts using the crates. There is no limit to the minor versions (we can even have `0.150.0` or `0.1500.0`), so I don't see anything wrong with this strategy.
Co-authored-by: Pavan Kumar Sunkara <pavan.sss1991@gmail.com>
4434: add more specific match postfix for Result and Option r=matklad a=bnjjj
In order to have the same behavior than `if let` and `while let`
Co-authored-by: Benjamin Coenen <5719034+bnjjj@users.noreply.github.com>
