-
-
Notifications
You must be signed in to change notification settings - Fork 721
perf(linter): refactor rules to take advantage of node type skipping #14519
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
perf(linter): refactor rules to take advantage of node type skipping #14519
Conversation
How to use the Graphite Merge QueueAdd either label to this PR to merge it via the merge queue:
You must have a Graphite account in order to use the merge queue. Sign up using this link. An organization admin has enabled the Graphite Merge Queue in this repository. Please do not merge from GitHub as this will restart CI on PRs being processed by the merge queue. This stack of pull requests is managed by Graphite. Learn more about stacking. |
31c1afa to
8496ad0
Compare
CodSpeed Performance ReportMerging #14519 will not alter performanceComparing Summary
Footnotes
|
8496ad0 to
83303b9
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Pull Request Overview
This PR refactors several ESLint rules to leverage node type skipping, improving linter performance by avoiding unnecessary traversals of every file. Instead of checking symbols globally, rules now target specific AST node types first.
Key Changes:
- Rules converted from
run_on_symboltorunmethod with node type filtering - Symbol-based rules refactored to find node types before checking symbols
- Code style improvements (formatting, closure definitions)
Reviewed Changes
Copilot reviewed 6 out of 7 changed files in this pull request and generated 2 comments.
Show a summary per file
| File | Description |
|---|---|
| no_inner_declarations.rs | Extracted common logic into check_rule helper function |
| no_import_assign.rs | Changed from symbol-based to import declaration-based traversal |
| no_func_assign.rs | Changed from symbol-based to function node-based traversal |
| no_extra_label.rs | Converted sequential if-let chains to match statement |
| no_ex_assign.rs | Changed from symbol-based to catch parameter node-based traversal |
| no_empty_pattern.rs | Converted tuple matching to explicit match arms |
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
Merge activity
|
…14519) Updating rules to take advantage of the node type skipping so that they don't need to be run on every file. I forgot to split this out, so this is a composite PR of several rules, which I'll try to avoid in the future. Some symbol-based rules have been refactored to search for the node type first and then look for symbols based on that. For example, if there are no imports in the file, we can skip `no-import-assign` instead of looping over all symbols and looking for an import.
83303b9 to
4440516
Compare
## [1.23.0] - 2025-10-13 ### 🚀 Features - 452d128 linter: Node/no_process_env (#14536) (Kenneth Skovhus) - a9e1261 linter: Implement react-refresh/only-export-components (#14122) (Rintaro Itokawa) - a8d542b linter/no-duplicate-imports: Support `allowSeparateTypeImports` option (#14527) (camc314) - 75a529c linter: Add run info about what functions rules implement (#13839) (camchenry) - 4fbdef2 language_server: Support `fmt.configPath` configuration (#14502) (Sysix) - 4d3ce2e language_server: Autodetect root `.oxfmtrc.json` (#14466) (Sysix) ### 🐛 Bug Fixes - fb4a20d language_server: Add whitespace for `// oxlint-disable-next-line` fix (#14356) (Sysix) - 6fce7f4 oxlint/changelog: Remove duplicate changelog entries (#14528) (camc314) - 74e52f3 linter/plugins: Resolve JS plugins with ESM condition names (#14541) (magic-akari) - 9965676 linter/exhaustive-deps: Ignore empty strings passed to `additionalHooks` option (#14464) (Redas) - d0a8e3e linter: `vue/no_required_prop_with_default` called `Option::unwrap()` on a `None` value (#14491) (Sysix) - 1192ea2 linter: Correct `nextjs/no-typos` path handling (#14480) (shulaoda) - 2796b16 tsgolint: Pipe tsgolints stderr (#14477) (camc314) ### 🚜 Refactor - 6440cde linter: Remove lifetime of `Message` (#14481) (Sysix) - f599bef language_server: Move `Message` to `DiagnosticReport` transformation to one place (#14447) (Sysix) - 4f301de napi/parser, linter/plugins: Improve formatting of generated code (#14554) (overlookmotel) - 68c0252 napi/parser, linter/plugins: Shorten generated raw transfer deserializer code (#14553) (overlookmotel) - 20e884e linter: Store `LintService` in `LintRunner` (#14471) (Sysix) - 9e9c5ba linter: Simplify built-in lint plugin checks (#14518) (camchenry) - 1986e0f linter/no-ex-assign: Use let-else chain (#14526) (camc314) - ce9bcf0 linter/no-func-assign: Use let-else chain (#14525) (camc314) - a9cea7c language_server: Use `FxHashSet` for `ServerLinter::extended_paths` (#14517) (Sysix) - ea5838e linter/no-import-assign: Use let-else chain (#14524) (camc314) - f977700 language_server: Pass `LintOptions` to `Worker::refresh_server_linter` (#14510) (Sysix) - 54b001f linter/no-new-require: Improve diagnostic message clarity (#14511) (shulaoda) - 650ea68 linter: Improve nextjs/no-typos rule (#14476) (shulaoda) ### 📚 Documentation - 198f2e9 linter: Fix code example for `branches-sharing-code` (#14514) (camc314) - d776a17 linter: Improve `nextjs/no-typos` rule documentation (#14470) (shulaoda) ### ⚡ Performance - b44a30e language_server: Transform `Message` to `DiagnosticReport` with one call (#14448) (Sysix) - 31766fd linter/plugins: Provide `loc` via prototype (#14552) (overlookmotel) - aec0c08 linter: Allow analyzing node types in match blocks with guards (#14459) (camchenry) - 9044187 linter: Skip running node-specific rule if file contains no relevant nodes (#14457) (camchenry) - 422f54e linter: Only run rule run functions if implemented (#14454) (camchenry) - 2c228ae liner: Use top-level match for `no_obj_calls` (#14523) (camchenry) - c49d891 linter: Use match for `no_negated_condition` (#14522) (camchenry) - e222fc2 linter: Use match for `no_multi_assign` (#14521) (camchenry) - 4440516 linter: Refactor rules to take advantage of node type skipping (#14519) (camchenry) - 78261d6 linter: Refactor `no-invalid-fetch-options` to be more easily analyzed (#14458) (camchenry) ### 🧪 Testing - 8d8881d linter/plugins: Expand tests for module resolution of plugins (#14559) (overlookmotel) - 33b6cde language_server: Add basic tests for `WorkspaceWorker::did_change_configuration` (#14531) (Sysix) - bfe1ecd language_server: Add tests for `WorkspaceWorker::init_watchers` (#14516) (Sysix) Co-authored-by: camc314 <18101008+camc314@users.noreply.github.com>

Updating rules to take advantage of the node type skipping so that they don't need to be run on every file. I forgot to split this out, so this is a composite PR of several rules, which I'll try to avoid in the future.
Some symbol-based rules have been refactored to search for the node type first and then look for symbols based on that. For example, if there are no imports in the file, we can skip
no-import-assigninstead of looping over all symbols and looking for an import.