Adding the ability to invalidate caches to force metadata collection
This adds the discussed hack to touch `clippy_lints/src/lib.rs` to invalidate cargos cache and force metadata collection. I've decided to use the [`filetime`](https://github.com/alexcrichton/filetime) crate instead of the touch command to make is cross-platform and just cleaner in general. Looking at the maintainers I would say that it's a save crate to use xD.
---
cc: #7172 I know this ID without looking it up now... This is not good
changelog: none
r? `@flip1995`
Move `semicolon_if_nothing_returned` to `pedantic`
This moves the `semicolon_if_nothing_returned` lint to `pedantic` category.
I had done #7148, but on the master branch, and Github doesn't seem to let me change that, so here's another PR
changelog: Move [`semicolon_if_nothing_returned`] lint into `pedantic` category.
Some SpanlessHash improvements
changelog: none
* Use `mem::discriminant().hash()` instead of `stable_hash` for simple enums and then use `FxHasher` instead of `StableHasher`. We don't use any StableHash features.
* Use `UnHashMap` for maps keyed by spanless hash values.
Fix invalid syntax in `from_iter_instead_of_collect` suggestion
First attempt at contributing, hopefully this is a good start and can be improved. :)
fixes#7259
changelog: [`from_iter_instead_of_collect`] fix invalid suggestion involving "as Trait"
Fix `redundant_closure` for `vec![]` macro in a closure with arguments
fixes: #7224
changelog: Fix `redundant_closure` for `vec![]` macro in a closure with arguments
Move `needless_borrow` to style
fixes: #3742#7105 should be merged first to fix the false positive.
changelog: move `needless_borrow` from `nursery` to `style`
fix `needless_borrow` suggestion
fixes: #2610
While I'm working on this, should needless_borrow be split into two? One lint for expressions and another for patterns. In expression it only lints when the compiler inserts a dereference, but for patterns it's whenever a double reference is created. I think at least the case where a double reference is needed should be split into a new lint as it's not 'needless', it can just be done without a ref binding.
For illustration:
```rust
fn foo(x: &&str) {}
match Some("test") {
// ref binding is useless here
Some(ref x) => *x,
_ => (),
}
match Some("test") {
// ref binding is useless here
Some(ref x) => x.len(),
_ => (),
}
match Some("test") {
// double reference is needed, but could be `Some(x) => foo(&x)`
Some(ref x) => foo(x),
_ => (),
}
```
changelog: Improve the suggestion for `needless_borrow` in patterns to change all usage sites as needed.
changelog: Add lint `ref_binding_to_reference`
Early return from LintPass registration when collecting metadata
This speeds up the metadata collection by 2-2.5x on my machine. During
metadata collection other lint passes don't have to be registered, only
the lints themselves.
cc #7172
r? `@xFrednet`
changelog: none
This speeds up the metadata collection by 2-2.5x on my machine. During
metadata collection other lint passes don't have to be registered, only
the lints themselves.
Rustup
r? `@ghost`
cc `@xFrednet` There was a change to the `rustc_span::FileName` removing the `Display` impl for it. I adapted the metadata collector to compile with that change. I'm not sure if I changed the behavior with this. The path to the string is now printed relative to the `clippy_lints` dir. So for example `src/swap.rs`. I think this should be fine, but probably something to be aware of.
changelog: none
Adding the default lint level to the metadata collection
I noticed while working on the website adaption that the lint groups still had the `clippy::` prefix in the JSON output. This PR removes this prefix and adds a `level` field to each lint and with that simplifies the website display and saves performance.
The deprecated lints get are assigned to the level `none`. This is a bit different in comparison to the current lint list, but I believe that this will look better overall. Unless there is any argument against this :).
That's it just a small baby PR in comparison to the original monster ^^
---
See: #7172 for the full metadata collection to-do list or to suggest a new feature in connection to it.
---
changelog: none
r? `@flip1995`
Implement the new desugaring from `try_trait_v2`
~~Currently blocked on https://github.com/rust-lang/rust/issues/84782, which has a PR in https://github.com/rust-lang/rust/pull/84811~~ Rebased atop that fix.
`try_trait_v2` tracking issue: https://github.com/rust-lang/rust/issues/84277
Unfortunately this is already touching a ton of things, so if you have suggestions for good ways to split it up, I'd be happy to hear them. (The combination between the use in the library, the compiler changes, the corresponding diagnostic differences, even MIR tests mean that I don't really have a great plan for it other than trying to have decently-readable commits.
r? `@ghost`
~~(This probably shouldn't go in during the last week before the fork anyway.)~~ Fork happened.
Don't lint `multiple_inherent_impl` with generic arguments
fixes: #5772
changelog: Treat different generic arguments as different types in `multiple_inherent_impl`
Deny warnings in every main sub-crate
Pointed out by `@xFrednet` in https://github.com/rust-lang/rust-clippy/pull/7229#issuecomment-842978909
This enables the same (rustc) lints in every main sub-crate:
- `clippy`
- `clippy_lints`
- `clippy_utils`
- `clippy_dev`
In addition it forwards the `deny-warnings` feature to those sub-crates, so we don't miss warnings that then become a problem during sync. (I wanted to fix that before, but forgot about it, so thanks for pointing it out `@xFrednet!)`
changelog: none
This enables the same warnings that are enabled in `clippy_lints` also
in `clippy_utils` and `clippy_dev`. Then it makes sure, that the
`deny-warnings` feature is passed down to `clippy_lints` and
`clippy_utils` when compiling Clippy.
