> Think about it: the Haskell community, consisting largely of academic researchers concerned with mathematics, logics, theorem proving, pure functional programming, semantics-preserving compilation, type systems, and so on, is unable to implement their own specification. This is like a man wearing a tuxedo, walking around with toilet paper stuck to his shoe. For fifteen years! It would be funny if the same people weren’t convinced that monadic parser combinators solve all problems in parsing. As it is, I find it maddening. > Back to the point at hand. Since no one implements the spec, it doesn’t make sense to see whether the spec is context sensitive; that would be like flogging a dead horse, and, moreover, a dead horse that was never actually alive in the first place. > I’ll use GHC as my example. Here is the parser (which of course is not written using monadic parser combinators), and here is the lexer.