I recently wrote a regexp to tokeninze clojure code it runs in under a single millis to parse a large amount of text. Regexps are freaking fast.
@bhauman yes, that’s what my handwritten parser does (well in multiple steps, but it could be optimized into a single regex maybe). It’s by far the fastest.
but regexes only go so far, context free grammars are more powerful (nested parens for example), so sometimes it’s very convenient to use something like Instaparse
@borkdude agreed 🙂