mirror of
https://github.com/oxc-project/oxc.git
synced 2026-09-14 19:36:11 +08:00
621808eb31
Lazily measure non-ASCII graphemes instead of allocating an eager boundary table. Criterion single_unicode_tab: 1.1575 µs → 596.36 ns (-48.6%). ASCII cases remain neutral or better. Kibana output is unchanged; end-to-end performance is neutral without type-aware linting. AI-assisted.