fix(parser/html): correctly lex tag names and attribute names

[dyc3]

Summary

This PR fixes the lexing of html tag names and attribute names to allow all the characters defined in the HTML spec.

This should fix the errors we were seeing in our ecosystem CI.

Test Plan

See: https://html.spec.whatwg.org/#elements-2

Also added some unit tests.

[codspeed-hq]

CodSpeed Performance Report

Merging #3853 will degrade performances by 8.08%

_{Comparing 09-10-fix_parser_html_correctly_lex_tag_names_and_attribute_names (663a823) with main (83e5cef)}

Summary

⚡ 2 improvements
❌ 1 regressions
✅ 104 untouched benchmarks

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Benchmarks breakdown

	Benchmark	main	09-10-fix_parser_html_correctly_lex_tag_names_and_attribute_names	Change
⚡	big5-added_15586211152145260264.json[cached]	828.7 µs	434.5 µs	+90.75%
❌	db_17847247775464589309.json[uncached]	15.4 ms	16.8 ms	-8.08%
⚡	eucjp_1600564308684076393.json[uncached]	1,027.4 µs	959.3 µs	+7.1%