Jason Newcomb [Mon, 9 Aug 2021 18:18:53 +0000 (14:18 -0400)]
Fix tracking of which locals would need to be captured in a closure.
* Captures by sub closures are now considered
* Copy types are correctly borrowed by reference when their value is used
* Fields are no longer automatically borrowed by value
* Bindings in `match` and `let` patterns are now checked to determine how a local is captured
bors [Thu, 12 Aug 2021 20:20:58 +0000 (20:20 +0000)]
Auto merge of #7560 - xFrednet:7289-configuration-for-every-type-lint, r=camsteffen
Use `avoid-breaking-exported-api` configuration in types module
This PR empowers our lovely `avoid-breaking-exported-api` configuration value to also influence the emission of lints inside the `types` module.
(That's pretty much it, not really a change worthy of writing a fairy tale about. Don't get me wrong, I would love to write a short one, but I sadly need to study now).
---
Closes: rust-lang/rust-clippy#7489
changelog: The `avoid-breaking-exported-api` configuration now also works for [`box_vec`], [`redundant_allocation`], [`rc_buffer`], [`vec_box`], [`option_option`], [`linkedlist`], [`rc_mutex`]
changelog: [`rc_mutex`]: update the lint message to comply with the normal format
---
r? `@camsteffen,` as you implemented the configuration value
cc: `@flip1995,` as we've discussed this change in rust-lang/rust-clippy#7308
bors [Thu, 12 Aug 2021 08:02:44 +0000 (08:02 +0000)]
Auto merge of #7516 - lf-:unwrap-or-default, r=xFrednet
Add `unwrap_or_else_default` lint
---
*Please write a short comment explaining your change (or "none" for internal only changes)*
changelog: Add a new [`unwrap_or_else_default`] style lint. This will catch `unwrap_or_else(Default::default)` on Result and Option and suggest `unwrap_or_default()` instead.
bors [Thu, 12 Aug 2021 07:47:07 +0000 (07:47 +0000)]
Auto merge of #7556 - F3real:no_effect_inclusive_range, r=flip1995
No effect inclusive range
I noticed during last PR that range expression is `ExprKind::Struct` while inclusive range is `ExprKind::Call` which was why it was not handled. This PR adds check for this case.
changelog: [`no_effect]` Report inclusive range in no_effect lint
bors [Mon, 9 Aug 2021 14:04:36 +0000 (14:04 +0000)]
Auto merge of #7544 - r00ster91:patch-1, r=flip1995
Clean up examples in new lint suggestion template
I'm pretty sure they meant to write `bounds checking` when they wrote `bounce checking` but I could be wrong. After that I thought I could improve it further and ended up with this.
bors [Mon, 9 Aug 2021 13:48:47 +0000 (13:48 +0000)]
Auto merge of #7542 - LeSeulArtichaut:cleanups, r=flip1995
Cleanup usage of `span_to_snippet` and `LintContext::sess`
- avoid using `SourceMap::span_to_snippet` directly and use `clippy_utils::source::snippet_opt` instead
- don't use `LintContext::sess()` on `EarlyContext`s which have a `sess` field directly available, saving the import of `LintContext`
bors [Wed, 4 Aug 2021 09:02:48 +0000 (09:02 +0000)]
Auto merge of #7529 - giraffate:update_node_version, r=flip1995,llogiq
Use node v12.x in CI
In https://github.com/rust-lang/rust-clippy/pull/7528#issuecomment-892281632, remark-cli version has been updated recently and it requires to use ESM module: https://github.com/remarkjs/remark/releases/tag/14.0.0. Node version in CI was v10.24.1, it seems to be old.
bors [Mon, 2 Aug 2021 20:09:11 +0000 (20:09 +0000)]
Auto merge of #7525 - xFrednet:7172-update-define-conf-macro, r=flip1995
Updated `define_Conf!` to support multi-line doc comments
Updated the `define_Conf!` macro to support multi-line doc comments for readability. This also enables configuration documentation to have multiple paragraphs.
I've also added the `metadata-collector-lint` feature to the CI build and testing tasks. (I would think that we want this, now that we officially switched over)
---
Now a small informal explanation what this PR changes (just for fun):
* *Once upon a time there was a monster. It was handsome, supportive and happy to dig through the best source code it has ever seen. Spanning over hundreds of lines and reading over complete crates, it was purely marvels!*
*However, there was one region in its territory that wasn't clean and well formatted like the rest. That was the mighty `define_Conf` macro. The monster would have cleaned it up a long time ago but, the previous ruler of this kingdom a powerful Python script was prohibiting it. But now that the old king was slain in the great battle of rust-1.54.0, everything was possible again. Our lovely monster was now able to grab its cleaning equipment and get some nice formatting into the now unprotected area.*
*Said and done! Let this day go down in history!*
(I'll be the first to admit, that I'm a bit lost today. And I'm also procrastinating on some other discussions where I have some catching up to do... Oh, well, this was fun)
bors [Mon, 2 Aug 2021 05:49:17 +0000 (05:49 +0000)]
Auto merge of #87535 - lf-:authors, r=Mark-Simulacrum
rfc3052 followup: Remove authors field from Cargo manifests
Since RFC 3052 soft deprecated the authors field, hiding it from
crates.io, docs.rs, and making Cargo not add it by default, and it is
not generally up to date/useful information for contributors, we may as well
remove it from crates in this repo.
Rollup merge of #87385 - Aaron1011:final-enable-semi, r=petrochenkov
Make `SEMICOLON_IN_EXPRESSIONS_FROM_MACROS` warn by default
This PR makes the `SEMICOLON_IN_EXPRESSIONS_FROM_MACROS` lint warn by default.
To avoid showing a large number of un-actionable warnings to users, we only enable the lint for macros defined in the same crate. This ensures that users will be able to fix the warning by simply removing a semicolon.
In the future, I'd like to enable this lint unconditionally, and eventually make it into a hard error in a future edition. This PR is a step towards that goal.
rfc3052: Remove authors field from Cargo manifests
Since RFC 3052 soft deprecated the authors field anyway, hiding it from
crates.io, docs.rs, and making Cargo not add it by default, and it is
not generally up to date/useful information, we should remove it from
crates in this repo.
Auto merge of #7509 - xFrednet:patch-1, r=flip1995
Updated config value name in change log for 1.54
Hey, I've noticed that the change log for 1.54 references the avoid_breaking_exported_api configuration with its Rust name, not the name it actually has in the config file. This should fix it for everyone, as the blog post references our change log on the master branch. Just a very small PR xD
Auto merge of #7498 - xFrednet:changelog-1-55, r=flip1995
Updated changelog for 1.55
This has again been a bit of work, but I'm happy to notice that my English is still improving, and I'm getting faster at these things. That's a very nice side effect of contributing and getting feedback on reviews :blush:
Moving on, there were a few things that I was unsure about:
* The PR rust-lang/rust#86717 changes an old entry in the change log, is this worth mentioning? I've left it out for now.
* The stabilization of `cargo clippy --fix` is quite awesome and important IMO. It sadly gets a bit lost in the *Other* entry, as it's the last one. Do we maybe want to move it somewhere else or change the headline order for this release?
* I've listed the introduction of the new `suspicious` group under the *Moves and Deprecations* section. Is this alright, or should it be moved to the *Other* section as well?
* Last but definitely not least, some fun! I've used the :tada: emoji in the `cargo clippy --fix` entry, is this okay?
Sorry for the bombardment of questions xD
---
The PR already includes the entries for the new metadata collection and website updates. These are not merged yet, but should probably be to make this correct. This might also require the commit hashes to be updated (Not sure on this, though). It would actually be super fitting to get this into this release as we also stabilize `--fix`. TODOs:
* [x] Merge metadata collection PRs:
1. #7279
2. #7298
3. #7420 (Hope to not get any merge conflicts)
Auto merge of #7502 - flip1995:rollup-y3ho3w0, r=flip1995
Rollup of 3 pull requests
Successful merges:
- #7279 (Adapting the lint list to Clippy's new metadata format)
- #7298 (Switch CI to new metadata collection)
- #7420 (Update lint documentation to use markdown headlines)
Philipp Krones [Wed, 28 Jul 2021 13:06:27 +0000 (15:06 +0200)]
Rollup merge of #7420 - xFrednet:7172-update-lint-documentation, r=flip1995
Update lint documentation to use markdown headlines
This PR updates all lint documentation to use markdown headlines. It additionally removed the *Known problems* section for lints without any problems. I've double-checked all automatic replacements, but a second pair of eyes is definitely appreciated!
I wasn't sure when you wanted to switch to the new metadata collection tomorrow, I therefore prepared this PR today. And that's it this is a standalone PR to keep the other related PRs reviewable.
changelog: none
r? `@flip1995`
cc: #7172
Note: This should be merged with the other metadata collection related PRs.
Philipp Krones [Wed, 28 Jul 2021 13:06:26 +0000 (15:06 +0200)]
Rollup merge of #7298 - flip1995:ci-switch-to-monster, r=xFrednet,flip1995
Switch CI to new metadata collection
r? `@xFrednet`
Things we have to keep in mind:
- This removes the template files and the scripts used for deployment from the checkout. This was added in #5517. I don't think we ever needed those there. Not sure though.
- ~~As a result, we can't remove the python scripts yet. We have to wait until this hits a stable Clippy release.~~ I'll just break the next stable deploy and do it by hand once.
- This should be merged together with #7279. Me and `@xFrednet` will coordinate the switch
- ...?
I still have to try out some things:
- [x] Is it worth caching? Yes
- [x] ~~Is it worth to do a release build?~~ Nope
- [x] Does it actually work? With a few changes, yes
- [ ] ...?
changelog: Clippy now uses a lint to generate its documentation :tada:
Philipp Krones [Wed, 28 Jul 2021 13:06:26 +0000 (15:06 +0200)]
Rollup merge of #7279 - xFrednet:7172-adapt-website-to-new-format, r=flip1995
Adapting the lint list to Clippy's new metadata format
This is close to the end of a long living project to rewrite the lint metadata collection for Clippy. Progress on this has been tracked in #7172. This PR adds one of the last missing puzzle pieces, the actual display of all the changes that have been done in the background. A preview can be seen here: [Clippy's lint list](https://xfrednet.github.io/rust-clippy/master/index.html)
The styling has been discussed on [zulip](https://rust-lang.zulipchat.com/#narrow/stream/257328-clippy/topic/Styling.20feedback.20for.20Clippy's.20lint.20list/near/239601067) but is still open to suggestion.
Side note: It has been fun working on the website where we don't have unit tests and everything is just tested by playing around. However, it's good that this chaos is contained into this one part of Clippy. :bug:
---
Closes: #1303 Closes: #4310
This actually closes fewer things than I thought it would...
changelog: Reworked Clippy's [website](https://rust-lang.github.io/rust-clippy/master/index.html):
changelog: * Added applicability information about lints
changelog: * Added a link to jump to the specific lint implementation
changelog: * Adapted some styling and improved loading time
xFrednet [Tue, 18 May 2021 22:08:31 +0000 (00:08 +0200)]
Adapting the lint list to Clippy's new metadata format
Changes included:
- Minimum adaption to the new `lints.json` format
- Fixing filtering for the new `lints.json` format; hardcoding the
lint groups in the index
- Recreating the original doc styling for the new format
- Fixed sytax highlighting for rust,ignore code blocks
- Fixed markdown table extraction in the metadata collector and
fixed lint level output
- Adding the additional information row for lints
- Changed the website title to Clippy's lint list
- Flexing the website for mobile users
- Added (?) references for lint levels and groups
- Making deprecated lints look dead
- Removed JS code block language extraction in favor of a rust
implementation `rust-clippy#7352`
- Added the suspicious lint group to the lint list
- Remove trailing whitespaces from index.html
- Fix code highlighting
- Use default value if the docVersion is empty
Co-authored-by: Philipp Krones <hello@philkrones.com>
flip1995 [Mon, 31 May 2021 15:46:20 +0000 (17:46 +0200)]
Update deploy CI
This updates all the deploy scripts and the deploy workflow.
The deploy workflow now runs the metadata collector to collect the lint
documentation. It also changes the files that are checked out in the
deploy workflow from master and adds an explanation why we have to do
this.
Auto merge of #7492 - nfejzic:improve_help, r=Manishearth
Explain flags missing in cargo check in --help
This commit closes #7389. As stated in the issue, `cargo clippy --help`
provides explanation for some flags and states that the rest are same
as in `cargo check --help`, even though some clippy specific flags
exist.
This commit extends the `cargo clippy --help` with two additional flags,
- `cargo clippy --fix`
- `cargo clippy --no-deps`
If there are more flags which are not present in `cargo check --help`
please bring these to my attention, I will include these aswell.
For now, I noticed only the two flags mentioned above.
changelog: `cargo clippy --help` now explains additional flags missing in `cargo check --help`.
Nadir Fejzic [Tue, 27 Jul 2021 09:13:42 +0000 (11:13 +0200)]
Handle `--no-deps` flag same as `--fix` flag.
As proposed in the pull request thread, there is some inconsistency in
handling the `--no-deps` flag which requires `--` before it, and
`--fix` flag which does not.
In this commit the `--no-deps` flag does not need the `--` anymore.
However, it can still be used that way: `cargo clipyy -- --no-deps`.