the mruby parser is hard to maintain (or read) due to ruby being a contextual language and bison being a context-free parser generator.

integrating prism into mruby would unify the parsers, allowing for faster adoption of new syntactic constructs. the prism-generated ast would also result in better error reporting.

i'm asking this because i've had a long-standing plan to integrate prism, but didn't want to make a very large pr (in scope) out of nowhere