perf(linter/no-unused-vars): do not construct Regex for default ignore pattern

[camchenry]

Profiling shows that constructing the Regex for no-unused-vars takes up a good chunk of time for small files. Since the default pattern for ignored variables is just ^_, we can more simply implement this with .starts_with('_').

I decided to create a new Option-like enum that has three states: Some and None which are like the typical Option states, and another Default state which is a special case. The Default state represents that matching code should use the default pattern (if applicable). I decided not to keep using Option, because I didn't want to overload the None state by confusing whether or not it was the default state, or the user explicitly overrided it as None.

The theory here is that most users are not going to pass a custom ignore pattern, so we can avoid needing to create a Regex in most cases, which greatly speeds up the initialization for this rule.

[graphite-app]

Your org has enabled the Graphite merge queue for merging into main

Add the label “0-merge” to the PR and Graphite will automatically add it to the merge queue when it’s ready to merge. Or use the label “hotfix” to add to the merge queue as a hot fix.

You must have a Graphite account and log in to Graphite in order to use the merge queue. Sign up using this link.

[camchenry]

• perf(linter/no-unescaped-entities): add fast path to check if char should be replaced #6594
• perf(linter/no-unused-vars): do not construct Regex for default ignore pattern #6590 👈
• main

This stack of pull requests is managed by Graphite. Learn more about stacking.

Join @camchenry and the rest of your teammates on Graphite

[codspeed-hq]

CodSpeed Performance Report

Merging #6590 will not alter performance

_{Comparing 10-15-perf_linter_no-unused-vars_do_not_construct_regex_for_default_ignore_pattern (ee73f56) with main (9f555d7)}

Summary

✅ 30 untouched benchmarks

[camchenry]

codspeed shows +0% but it's closer to ~1% locally:

hyperfine --warmup 100 --shell=none --runs 9000 './oxlint-main --silent ./RadixUIAdoptionSection.jsx' './oxlint-unused-regex --silent ./RadixUIAdoptionSection.jsx'
Benchmark 1: ./oxlint-main --silent ./RadixUIAdoptionSection.jsx
  Time (mean ± σ):       3.4 ms ±   0.1 ms    [User: 1.1 ms, System: 1.3 ms]
  Range (min … max):     3.2 ms …   3.9 ms    9000 runs

Benchmark 2: ./oxlint-unused-regex --silent ./RadixUIAdoptionSection.jsx
  Time (mean ± σ):       3.4 ms ±   0.1 ms    [User: 1.1 ms, System: 1.3 ms]
  Range (min … max):     3.0 ms …   4.2 ms    9000 runs

Summary
  ./oxlint-unused-regex --silent ./RadixUIAdoptionSection.jsx ran
    1.01 ± 0.05 times faster than ./oxlint-main --silent ./RadixUIAdoptionSection.jsx

It's a smaller win than I hoped, but it seems like a good idea in principle.

[graphite-app]

Merge activity

• Oct 16, 3:58 PM EDT: A user added this pull request to the Graphite merge queue.
• Oct 16, 4:03 PM EDT: A user merged this pull request with the Graphite merge queue.