For a bit deeper intro to PEGs in Janet, the Janet for Mortals guide [1] is pretty incredible here (and also uses HTML as an example).

[1] https://janet.guide/pegular-expressions/

Reminds me of raku (perl6) grammars https://docs.raku.org/language/grammars

The part I really like about pegs in janet is it works on bytes.

Are pegs easier for syntax highlighting or LSPs?

https://janetdocs.org/ does (server side) syntax highlighting through hiccup-lexer which converts janet-peg's tokens to hiccup: https://github.com/sogaiu/janet-peg

The LSP used on vscode works with PEGs also: https://github.com/CFiggers/janet-lsp/blob/ca526aee35c60bcf5...

They're not really the same kind of thing. LSP is a carrier format, but it doesn't know or care how you parse things, whether it's with PEGs or an earley parser or shift reduce or whatever

Yes I know, but are they making LSP server construction easier?