Thank you for your suggestions!
I understand your concerns about the overall workload. While I’m not sure about the time required for the parser, I believe I can complete the printer within 100 hours, as I’ve already finished nearly half of it last year. Given this, do you think completing both within 350 hours is reasonable?
I know that LaTeX already has a new Lark implementation. Would you recommend just using Lark instead of ANTLR for implementing a Typst parser?
After GSoC, I would also be happy to work on adding Lark implementations for existing ANTLR-based parsers.