v0.13.3

What's Changed

Some tokenizer performance updates with ~12% speed improvement.

• fix: Eliminate unsafe manual token position updates in CSS parser by @bartveneman in #147
• perf: optimize advance() with lazy line/column tracking by @bartveneman in #148
• perf: inline is_ident_char() in hot tokenizer loops via CHAR_IDENT bi… by @bartveneman in #149

Full Changelog: v0.13.2...v0.13.3