P. Wadler. The marriage of e#ects and monads. In International Conference on Functional Programming, ICFP'98, pages 63--74, Baltimore, Maryland, 1998.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....sequence corresponds to the body of the main method in a Java program [9, 12.1.4] For example, here are some typical MJ class definitions. Another approach for lazy functional programming with state is the use of monads [21] Wadler has shown that e#ects systems can easily be adapted to monads [22]. 4 Program p : cd 1 . cd n ; s Class definition cd : class C extends C 1 . fd k cnd fd : C f ; cnd : C(C 1 x 1 , C j x j ) super(e 1 , e k ) s 1 . s n md : # m(C 1 x 1 , C n x n ) s 1 . s k Return type # : C void Expression e ....

P. Wadler. The marriage of e#ects and monads. In International Conference on Functional Programming, 1998.

....of class definitions plus a sequence of statements, s, to be evaluated. This term corresponds to the body of the main method in Java. Another approach for lazy functional programming with state is the use of monads [11] Wadler has shown that e#ects systems can easily be adapted to monads [12]. Program p : cd 1 . cd n ; s Class definition cd : class C extends C fd cnd md fd : C f ; cnd : C (C 1 var 1 , C j var j ) md : # m(C 1 var 1 , C n var n ) Expression e : var Variable null Null e.f Field access (C)e Cast pe ....

P. Wadler. The marriage of e#ects and monads. In International Conference on Functional Programming, 1998.

....not I O Or I O but not concurrency The answer is technically, yes of course. There is a long history of research into so called e ect systems, that track what kind of e ects an expression can have [21] Such e ect systems can be expressed in a monadic way, or married with a monadic type system [51]. However, the overhead on the programmer becomes greater, and I do not know of any language that uses such a system . An interesting challenge remains, to devise a more re ned system that is still practical; there is some promising work in this direction [6, 51, 45, 5] Meanwhile I argue that ....

....married with a monadic type system [51] However, the overhead on the programmer becomes greater, and I do not know of any language that uses such a system . An interesting challenge remains, to devise a more re ned system that is still practical; there is some promising work in this direction [6, 51, 45, 5]. Meanwhile I argue that a simple pure or impure distinction o ers an excellent cost bene t tradeo . Some smart compilers use type based e ect systems to guide their optimisers, but that is di erent from the programmer visible type system. 56 8 Summary We have surveyed Haskell s monadic I O ....

Wadler, P. The marriage of eects and monads. In ICFP98 [18], pp. 63-74.

....technique for specifying the operational semantics of programming languages (see [WF94] while for computationally irrelevant term constructors it su#ces to give local simplification rules, that can be applied non deterministically (because they are semantic preserving) Remark 2. 1In [Wad99] Wadler adopts a similar style, that distinguishes pure from monadic reduction. However, his pure reduction is a deterministic strategy, while simplification is non deterministic. In this respect, our approach is related to the Cham [BB90] simplification corresponds to heating and computation to ....

Philip Wadler. The marriage of e#ects and monads. In the International Conference on Functional Programming (ICFP '98), volume 34(1) of ACM SIGPLAN Notices, pages 63--74. ACM, June 1999.

....form the free algebra on the values for a monad which encapsulates the computational e ects present in the programming language. This idea fed through into the functional programming community where it has led to techniques in structuring compilers for both functional and non functional languages [CH97, Gor94, LH96, LHJ95, JW93, Wad95, Wad97, Wad98]. Less surprisingly the closely related theory of sketches has been used to give accounts of the construction of datatypes [BW90] These applications depend on looking at monads on categories other than Set, and for these categories something interesting happens. Equational logic can be soundly ....

Philip Wadler. The marriage of eects and monads. In International Conference on Functional Programming, Baltimore, September 1998. 21

....and e ects to perform all memory management operations in their ML compiler. More recently, researchers have begun to investigate the relationships between these three fundamental mechanisms. For example, Chen and Hudak [5] have discovered a connection between linear types and monads and Wadler [38] has recently presented a correspondence between monads and e ect systems. In this paper, we ll in the third side of the triangle by exploring the synergy between linear types and region, type and e ect systems, speci cally for the purpose of exploring new techniques in safe, static memory ....

Philip Wadler. The marriage of eects and monads. In ACM International Conference on Functional Programming, pages 63-74, Baltimore, September 1998.

....prove that it would induce the same observational congruence of the pure CBN interpreter. 13 E ect Masking and Monadic Encapsulation. SS99] shows that runST implements a cheap form of e ect masking (see [LG88, TJ92] thus extending the relation between e ects and monads established in [Wad98]. More precisely they give a translation from a type system with e ects and regions (EML) to one with runST (MML) It seems plausible that the translation given in [SS99] can be adapted so that the target language (MML) uses our run with construct instead of runST . Indeed, replacing runST by ....

P. Wadler. The marriage of eects and monads. In International Conference on Functional Programming, pages 63-74, Baltimore, September 1998. ACM.

....concurrency The answer is technically, yes of course. There is a long history of research into so called e ect systems, that track what kind of e ects an expression can have [Jouvelot and Gi ord, 1991] Such e ect systems can be expressed in a monadic way, or married with a monadic type system [Wadler, 1998]. However, the overhead on the programmer becomes greater, and I do not know of any language that uses such a system 8 . An interesting challenge remains, to devise a more re ned system that is still practical; there is some promising work in this direction [Benton et al. 1998, Wadler, 1998, ....

....[Wadler, 1998] However, the overhead on the programmer becomes greater, and I do not know of any language that uses such a system 8 . An interesting challenge remains, to devise a more re ned system that is still practical; there is some promising work in this direction [Benton et al. 1998, Wadler, 1998, Tolmach, 1998, Benton and Kennedy, 1999] Meanwhile I argue that a simple pure or impure distinction o ers an excellent cost bene t tradeo . 9 Summary We have surveyed Haskell s monadic I O system, along with three sign cant language extensions 9 . It is easy to extend a language, though Are ....

Wadler, P. (1998). The marriage of eects and monads. In [ICFP98, 1998], pages 63-74.

....prove that it would induce the same observational congruence # of the pure CBN interpreter. 13 . E#ect Masking and Monadic Encapsulation. SS99] shows that runST implements a cheap form of e#ect masking (see [LG88, TJ92] thus extending the relation between e#ects and monads established in [Wad98]. More precisely they give a translation from a type system with e#ects and regions (EML) to one with runST (MML) It seems plausible that the translation given in [SS99] can be adapted so that the target language (MML) uses our run with construct instead of runST . Indeed, replacing runST by ....

P. Wadler. The marriage of e#ects and monads. In International Conference on Functional Programming, pages 63--74, Baltimore, September 1998. ACM.

....Then it is applied to the functional translation of imperative programs, through the usual monadic call by value translation. We show the correctness of that translation. Section 4 is a brief conclusion, in which we discuss the di erences with a similar work of Wadler recently published in [14]. 2 Generalized monads This section introduces the notion of generalized monads, and some of their properties. 2.1 De nitions First, let us recall the classical de nition of a monad. De nition 1 (monad, 10, 13] A monad is de ned by a type operator , and two operations unit : star ....

....n 1g else fs; ng where fs; ng stands for a two elds record containing the values s and n. The function above is exactly what one would have written, and it is easy to establish logical properties on such interpretations. Related work. To our knowledge, the only similar work is a paper of Wadler [14], which have been developed independently. In that paper, Wadler combines the e ect discipline of Talpin and Jouvelot [12] and the notion of monads, proposing a translation from the rst to the second, its proof of correctness and type reconstruction algorithms. The monadic operator appears as ....

P. Wadler. The marriage of eects and monads. In International Conference on Functional Programming, pages 63-74, Baltimore, September 1998. ACM.

....since our interpretation does not use a global store but local stores instead, which only mention subsets of the references. In [8] we proposed a generalization of the notion of monad, parameterized by an abstract notion of e ect. A similar framework has been proposed independently by P. Wadler [27]. The idea is to make the monadic operator dependent over an e ect, leading to a family of type operators . In our case, an e ect is a pair ( and ( would be de ned by ( def = b b Then the monadic operator star itself is naturally parameterized by ....

....cations. One of the main interests is a direct interpretation of functional features, usually painfully handled in other frameworks. Our interpretation is based on a static analysis of the programs e ects and on the use of a generalized notion of monad, similar to the one proposed by P. Wadler in [27]. We claim that this interpretation can be easily extended to other programming features (exceptions, input output and so forth) by an adequate extension of the notion of e ect and of the corresponding monad. Another main aspect of our approach is its modularity. Indeed, we de ned an extended ....

P. Wadler. The marriage of eects and monads. In International Conference on Functional Programming, pages 63-74, Baltimore, September 1998. ACM.

....however, we will still be able to write concrete programs in ML notation, with implicit call by value sequencing. We further re ne this language with a type system for keeping track of e ects, very similar to Tolmach s and (to a lesser degree) Wadler s intermediate languages for ML [Tol98, Wad98] However, we will use e ect types prescriptively, to de ne a new language, rather than descriptively, to analyze an already given one. The raw syntax is thus as follows: j b j 1 e 2 j 1 j 1 2 j 0 j 1 2 E : x j c 1 ; n j x : E j E1 E2 j ( j V E j (E1 ; E2) j ....

Philip Wadler. The marriage of eects and monads. In International Conference on Functional Programming, 1998.

....family of related analyses as type annotation (quali er) systems. ABHR99] describes the Dependency Core Calculus (DCC) and provides translations into DCC from several dependency based type quali er systems such as [VS97] DCC is one example of a calculus based on monads. Recent work [Kie98, Wad98] has explored the connection between monads and e ect systems [LG88] Some e ect systems can also be expressed as type quali er systems. However, the exact connection between monads, e ect systems, and type quali ers is unclear. Other frameworks choose a di erent design point by providing more ....

Philip Wadler. The Marriage of Eects and Monads. In ICFP'98 [ICF98], pages 63-74.

No context found.

P. Wadler. The marriage of e#ects and monads. In International Conference on Functional Programming, ICFP'98, pages 63--74, Baltimore, Maryland, 1998.

No context found.

Philip Wadler. The marriage of e#ects and monads. In the International Conference on Functional Programming (ICFP '98), volume 34(1) of ACM SIGPLAN Notices, pages 63--74. ACM, June 1999. 12

No context found.

P.Wadler. The marriage of e#ects and monads. #unpublished manuscript#, Mar. 1998.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC