## [0.10.1] - 2024-10-21
### Features
- af25752 linter: Add `unicorn/prefer-math-min-max` (#6621) (Brian Liu)
- 5095f02 linter: Added fixer for duplicate prefix in valid title jest
rule (#6699) (Tapan Prakash)
- e9976d4 linter: Add title whitespace fixer for jest valid title rule
(#6669) (Tapan Prakash)
- 45f02d5 linter: Add `unicorn/consistent-empty-array-spread` (#6695)
(Brian Liu)
- 01a35bb linter/eslint: Show ignore patterns in `eslint/no-unused-vars`
diagnostic messages (#6696) (DonIsaac)
### Bug Fixes
- 1bcd707 editor: Update config sent to language server (#6724)
(Nicholas Rayburn)
- ce25c45 linter: Panic in `disable-directives` (#6677) (dalaoshu)
- a5de230 linter/import: `import/no-duplicates` handles namespace
imports correctly (#6694) (DonIsaac)
- b0b6ac7 linter/no-cond-assign: False positive when assignment is in
body statement (#6665) (camchenry)
### Performance
- 6a76ea8 linter/no-unused-vars: Use default IgnorePattern when /^_/ is
provided as a pattern (#6697) (DonIsaac)
### Refactor
- d6609e9 linter: Use `run_on_jest_node` for existing lint rules (#6722)
(camchenry)
- 97195ec linter: Add `run_on_jest_node` to run rules on only jest nodes
(#6721) (camchenry)
- 155fe7e linter: Allow `Semantic` to be passed for collecting Jest
nodes (#6720) (camchenry)
- ad8f281 linter: Use iter for collecting jest nodes (#6719) (camchenry)
- dc19a8f linter: Use iterator for collecting jest imports (#6718)
(camchenry)
- 29c1447 linter: `jest/valid-title` fixer to use `Span::shrink` method
(#6703) (Tapan Prakash)
- 2eb984a linter: Add missing `should_run` implementations (#6666)
(camchenry)
- 23f88b3 linter/import: Better diagnostic messages for
`import/no-duplicates` (#6693) (DonIsaac)
---------
Co-authored-by: Boshen <1430279+Boshen@users.noreply.github.com>
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
Syncs up the options from the language server `Options` struct with the
values that the extension supplies.
I tend to avoid `toJSON` because it will be used by `JSON.stringify`
when called and it's not well known.
There's no reason that `LanguageServerConfig` and `Config` have to both
exist, but I think two types is easier to understand.
---------
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
Follow-on after #6404. Add `Derive::snake_name` method which defaults to `Self::trait_name().to_case(Case::Snake)`, but can be overridden. This allows moving the "special case" code for `ESTree` filename into `estree.rs`.
- closes https://github.com/oxc-project/oxc/issues/6038
Migrates all simple iterations over Jest nodes to use the `run_on_jest_node` method. Rules which require more custom data (such as getting nodes in order or storing data about counts, duplicates, etc.) have not been migrated.
Some simple benchmarking shows that this is ~5% faster on `vscode` when the Jest/Vitest rules are NOT enabled (due to being able to skip them now). And it is up to 8% on `vscode` when the Jest/Vitest plugins are enabled.
```
hyperfine -i -N --warmup 3 --runs 30 './oxlint-jest-iter --deny=all --silent ./vscode' './oxlint-main --deny=all --silent ./vscode'
Benchmark 1: ./oxlint-jest-iter --deny=all --silent ./vscode
Time (mean ± σ): 2.348 s ± 0.104 s [User: 16.922 s, System: 0.641 s]
Range (min … max): 2.141 s … 2.544 s 30 runs
Benchmark 2: ./oxlint-main --deny=all --silent ./vscode
Time (mean ± σ): 2.476 s ± 0.042 s [User: 17.768 s, System: 0.668 s]
Range (min … max): 2.430 s … 2.598 s 30 runs
Summary
./oxlint-jest-iter --deny=all --silent ./vscode ran
1.05 ± 0.05 times faster than ./oxlint-main --deny=all --silent ./vscode
```
```
hyperfine -i -N --warmup 3 --runs 30 './oxlint-jest-iter --deny=all --jest-plugin --vitest-plugin --silent ./vscode' './oxlint-main --deny=all --jest-plugin --vitest-plugin --silent ./vscode'
Benchmark 1: ./oxlint-jest-iter --deny=all --jest-plugin --vitest-plugin --silent ./vscode
Time (mean ± σ): 2.728 s ± 0.118 s [User: 19.782 s, System: 0.786 s]
Range (min … max): 2.580 s … 3.078 s 30 runs
Benchmark 2: ./oxlint-main --deny=all --jest-plugin --vitest-plugin --silent ./vscode
Time (mean ± σ): 2.939 s ± 0.051 s [User: 21.259 s, System: 0.859 s]
Range (min … max): 2.848 s … 3.064 s 30 runs
Summary
./oxlint-jest-iter --deny=all --jest-plugin --vitest-plugin --silent ./vscode ran
1.08 ± 0.05 times faster than ./oxlint-main --deny=all --jest-plugin --vitest-plugin --silent ./vscode
```
- part of https://github.com/oxc-project/oxc/issues/6038
Adds a new `run_on_jest_node` method for running lint rules which allows only Jest/Vitest nodes to be linted. This simplifies a number of Jest rules by removing the need to iterate/collect Jest nodes inside of them. Now, they can simply run on the Jest nodes that are passed to them directly. This also saves time by skipping the running of the rule if it is not a test file or the Jest/Vitest plugins are not enabled.
This refactors the Jest node collection to not be rule-specific by allowing `Semantic` to be passed instead of the LintContext. This prepares the code to be used outside of a particular rule context.
Adds a new method for iterating over jest nodes instead of collecting them all at once. The jest node collection now uses the iterator internally for consistency between the two approaches. This should save time and memory anywhere we need to iterate over Jest nodes, as we don't need to collect everything at once.
Change `pub fn` to `pub(crate) fn` for all internal methods in isolated declarations. I would appreciate someone who is more familiar with this code informing me if I incorrectly privatized methods that should be public.
Uses an iterator for getting reference IDs now, so we don't have to collect all of them at once. Also removes the checking around whether there are any `JEST_METHOD_NAMES`: this occurs in `collect_ids_referenced_to_global` anyway (will return no elements in the iterator if there are not any).
This was originally in #6404 but I pulled it out to reduce that PR's diff.
Add ability (currently unused) to output arbitrary text files from `oxc_ast_tools` generators.
Follow-on after #6404.
`ArrayExpressionElement` and `Elision` are not used in the TS types, because `ArrayExpression` has an override for the field it uses.
002289b4b1/crates/oxc_ast/src/ast/js.rs (L293-L302)
Prevent these TS type defs being emitted by introducing a new `#[estree(custom_ts_def)]` attr, to go with `#[estree(custom_serialize)]`.
Use `IgnorePattern::Default` even when `/^_/` is explicitly provided as a regex. This gives better diagnostic messages and lets us use `starts_with('_')` in more cases.
Inform users what pattern an unused variable must match in diagnostic help messages.
```
⚠ eslint(no-unused-vars): Variable 'a' is declared but never used. Unused variables should start with a '_'.
╭─[no_unused_vars.tsx:1:5]
1 │ var a=10
· ┬
· ╰── 'a' is declared here
╰────
help: Consider removing this declaration.
```
Supported fixer for white space in title for valid title rule.
---------
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
Closes#6660
- Value (i.e. not `import type`) namespace imports are now grouped correctly
- When grouping, consider _all_ import entries, not just the first one. Fixes a false negative when inlined `type` imports are used, e.g. `import { a, type b } from 'foo'`
## What This PR Does
- Include the offending module in the diagnostic's message
- Add a help message
- Add label text to the first module request
It also includes these minor refactors:
- Move `check_request` closure into a separate function
- Move diagnostics creation to a separate function (same as every other rule)
`oxc_allocator::Vec` is intended for storing AST types in the arena. `Vec` is intended to be non-drop, because all AST types are non-drop. If they were not, it would be a memory leak, because those types will not have `drop` called on them when the allocator is dropped.
However, `oxc_allocator::Vec` is currently a wrapper around `allocator_api2::vec::Vec`, which *is* drop. That unintentionally makes `oxc_allocator::Vec` drop too.
This PR fixes this by wrapping the inner `allocator_api2::vec::Vec` in `ManuallyDrop`. This makes `oxc_allocator::Vec` non-drop.
The wider consequence of this change is that the compiler is now able to see that loads of other types which contain `oxc_allocator::Vec` are also non-drop. Once it can prove that, it can remove a load of code which handles dropping these types in the event of a panic. This probably also then allows it to make many further optimizations on that simplified code.
Strictly speaking, this PR is a breaking change. If `oxc_allocator::Vec` is abused to store drop types, then in some circumstances this change could produce a memory leak where there was none before. However, we've always been clear that only non-drop types should be stored in the arena, so such usage was always a bug.
#6622 fixes the only place in Oxc where we mistakenly stored non-drop types in `Vec`.
The change to `oxc_prettier` is because compiler can now deduce that `Doc` is non-drop, which causes clippy to raise a warning about using `then` instead of `then_some`.
As follow-up, we should:
1. Wrap other `allocator_api2` types (e.g. `IntoIter`) in `ManuallyDrop` too, so compiler can prove they are non-drop too (or reimplement `Vec` ourselves - #6488).
2. Introduce static checks to prevent non-drop types being used in `Box` and `Vec`, to make memory leaks impossible, and detect them at compile time.
