The title here seems to be inaccurate. The title of the post--based on the Google cache--is "What’s in a parser? Attoparsec rewired (2/2)". It's not actually

The title here seems to be inaccurate. The title of the post--based on the Google cache--is "What’s in a parser? Attoparsec rewired (2/2)". It's not actually about building parsers so much as comparing the new Haskell attoparsec parser to C parsers. I'm glad to see this improvement being made in the LL parsers in Haskell. They're much more natural to write than parsers with a bottom-up parser generator. I believe Happy is the major parser-generator for Haskell and since it uses YACC-like syntax for the grammar it starts to feel like I'm writing C into my Haskell code. Parser-generators always feel a bit ugly, but they usually yield the most efficient and fastest code.