From an academic point of view, most parser combinator libraries are basically LL parsers with backtracking. This is eminently practical (especially if the bac

From an academic point of view, most parser combinator libraries are basically LL parsers with backtracking. This is eminently practical (especially if the backtracking part is done well), but from an academic point of view that means they're not very interesting - really just a nicer way of implementing recursive descent parsers. There is of course plenty of academic work specifically on how to implement parser combinator libraries efficiently, how to make them invertible (so they can also produce prettyprinters), provide good error messages, etc.