I believe the interesting twist is the use of a GADT to represent the parse result. I see this in this line of history: "Higher-Order Functions for Parsing", G

I believe the interesting twist is the use of a GADT to represent the parse result. I see this in this line of history: "Higher-Order Functions for Parsing", Graham Hutton - parser combinators before monadic parsing "Monadic parser combinators", Graham Hutton and Erik Meijer - the discovery of monadic parsing, which is now the standard for parser combinators