Logic Programming with Monads and.pdf

Logic Programming with Monads andComprehensionsYves Bekkers  Paul Tarau yAbstractWe give a logical reconstruction of all-solution predicates in terms oflist comprehensions in Prologs and we describe a variety of logic pro-gramming constructs in terms of monads and monad morphisms. Novelmonad structures are described for lazy function lists, clause unfoldingsand a monad morphism based embedding of Prolog in Prolog is given.Keywords: computing paradigms, logic programming, monads, listcomprehensions, all solution predicates, Prolog, higher-order unication,lazy function lists.1 IntroductionMonads and comprehensions, have been successfully used in functional program-ming as a convenient generalization of various structurally similar programmingconstructs starting with simple ones like list processing and ending with intri-cate ones like CPS transformations and state transformers. We refer to the workof Philip Wadler [16, 17] for a long list of powerful examples and to Moggi [10]for the categorist sources of the concept of monad.All-solution predicates are the logic programming siblings of list comprehen-sions in functional programming languages.Let us point out an interesting (although possibly episodical) parallel be-tween these two programming paradigms. When doubts have been raised byGoguens thesis [6] on the usefulness of higher order constructs for functionalprogramming, this seemed deja vu for logic programmers familiar with War-rens paper [18]. After emulating some basic -calculus constructs in terms ofrst-order Horn-clause logic (i.e. what is also known as pure Prolog) Warrenhas shown that most of the expressiveness of higher-order constructs can beemulated in rst order logic programming.All-solution predicates have been recognized by Warren as a notable excep-tion and have been present in logic programming ever since. However, in theUniversite de Rennes I and IRISA/INRIA Rennes, bekkers@irisa.fryDepartement dInformatique, Universite de Moncton, ta