ML Style Research Articles

Efficient compilation of algebraic effect handlers

The popularity of algebraic effect handlers as a programming language feature for user-defined computational effects is steadily growing. Yet, even though efficient runtime representations have already been studied, most handler-based programs are still much slower than hand-written code. This paper shows that the performance gap can be drastically narrowed (in some cases even closed) by means of type-and-effect directed optimising compilation. Our approach consists of source-to-source transformations in two phases of the compilation pipeline. Firstly, elementary rewrites, aided by judicious function specialisation, exploit the explicit type and effect information of the compiler’s core language to aggressively reduce handler applications. Secondly, after erasing the effect information further rewrites in the backend of the compiler emit tight code. This work comes with a practical implementation: an optimising compiler from Eff, an ML style language with algebraic effect handlers, to OCaml. Experimental evaluation with this implementation demonstrates that in a number of benchmarks, our approach eliminates much of the overhead of handlers, outperforms capability-passing style compilation and yields competitive performance compared to hand-written OCaml code as well Multicore OCaml’s dedicated runtime support.

Recovering purity with comonads and capabilities

In this paper, we take a pervasively effectful (in the style of ML) typed lambda calculus, and show how to extend it to permit capturing pure expressions with types. Our key observation is that, just as the pure simply-typed lambda calculus can be extended to support effects with a monadic type discipline, an impure typed lambda calculus can be extended to support purity with a comonadic type discipline. We establish the correctness of our type system via a simple denotational model, which we call the capability space model. Our model formalises the intuition common to systems programmers that the ability to perform effects should be controlled via access to a permission or capability, and that a program is capability-safe if it performs no effects that it does not have a runtime capability for. We then identify the axiomatic categorical structure that the capability space model validates, and use these axioms to give a categorical semantics for our comonadic type system. We then give an equational theory (substitution and the call-by-value β and η laws) for the imperative lambda calculus, and show its soundness relative to this semantics. Finally, we give a translation of the pure simply-typed lambda calculus into our comonadic imperative calculus, and show that any two terms which are βη-equal in the STLC are equal in the equational theory of the comonadic calculus, establishing that pure programs can be mapped in an equation-preserving way into our imperative calculus.

The Size-Change Termination Principle for Constructor Based Languages

This paper describes an automatic termination checker for a generic first-order call-by-value language in ML style. We use the fact that value are built from variants and tuples to keep some information about how arguments of recursive call evolve during evaluation. The result is a criterion for termination extending the size-change termination principle of Lee, Jones and Ben-Amram that can detect size changes inside subvalues of arguments. Moreover the corresponding algorithm is easy to implement, making it a good candidate for experimentation.

An intuitionistic λ-calculus with exceptions

We introduce a typed λ-calculus which allows the use of exceptions in the ML style. It is an extension of the system $AF_2$ of Krivine & Leivant (Krivine, 1990; Leivant, 1983). We show its main properties: confluence, strong normalization and weak subject reduction. The system satisfies the “the proof as program” paradigm as in $AF_2$. Moreover, the underlined logic of our system is intuitionistic logic.

Dependently Typed Records in Type Theory

Abstract. The language Pebble of Burstall and Lampson proposed dependent types as the underlying principle in a unified framework to explain facilities for programming in the large, such as modules and signatures , as well as for programming in the small. This proposal soon extended to large scale formal proof development as well. In fact, the functional approach to modularity has turned out to be a hard problem, which is still far from a fully satisfactory solution. This paper discusses aspects of this approach, including representations of records, informative signatures, sharing, and subtyping. My main contribution in this paper is to show that structures with dependent types and manifest fields (roughly ML style modules) are internally definable in a type theoretic framework extended with inductive-recursive definition. This shows that powerful modules follow from general principles without module-specific assumptions.

Type dispatch for named hierarchical types

Type dispatch constructs are an important feature of many programming languages. Scheme has predicates for testing the runtime type of a value. Java has a class cast expression and a try statement for switching on an exception's class. Crucial to these mechanisms, in typed languages, is type refinement: The static type system will use type dispatch to refine types in successful branches. Considerable previous work has addressed type case constructs for structural type systems without subtyping, but these do not extend to named type systems with subtyping, as is common in object oriented languages. Previous work on type dispatch in named type systems with subtyping has not addressed its implementation formally.This paper describes a number of type dispatch constructs that share a common theme: class cast and class case constructs in object oriented languages, ML style exceptions, hierarchical extensible sums, and multimethods. I describe a unifying mechanism, tagging , that abstracts the operation of these constructs, and I formalise a small tagging language. After discussing how to implement the tagging language, I present a typed language without type dispatch primitives, and a give a formal translation from the tagging language.

Parallel functional programming on recursively defined data via data-parallel recursion

This article proposes a new language mechanism for data-parallel processing of dynamically allocated recursively defined data. Different from the conventional array-based data- parallelism, it allows parallel processing of general recursively defined data such as lists or trees in a functional way. This is achieved by representing a recursively defined datum as a system of equations, and defining new language constructs for parallel transformation of a system of equations. By integrating them with a higher-order functional language, we obtain a functional programming language suitable for describing data-parallel algorithms on recursively defined data in a declarative way. The language has an ML style polymorphic type system and a type sound operational semantics that uniformly integrates the parallel evaluation mechanism with the semantics of a typed functional language. We also show the intended parallel execution model behind the formal semantics, assuming an idealized distributed memory multicomputer.

Dynamic typing in polymorphic languages

AbstractThere are situations in programming where some dynamic typing is needed, even in the presence of advanced static type systems. We investigate the interplay of dynamic types with other advanced type constructions, discussing their integration into languages with explicit polymorphism (in the style of system F), implicit polymorphism (in the style of ML), abstract data types, and subtyping.

Deciding type isomorphisms in a type-assignment framework

AbstractThis paper provides a formal treatment of isomorphic types for languages equipped with an ML style polymorphic type inference mechanism. The results obtained make less justified the commonplace feeling that (the core of) ML is a subset of second order λ-calculus: we can provide an isomorphism of types that holds in the core ML language, but not in second order λ-calculus. This new isomorphism allows to provide a complete (and decidable) axiomatization of all the types isomorphic in ML style languages, a relevant issue for thetype as specificationsparadigm in library searches. This work is a very extended version of Di Cosmo (1992): we provide both a thorough theoretical treatment of the topic and describe a practical implementation of a library search system so that the paper can be used as a reference both by those interested in the formal theory of ML style languages, and by those simply concerned with implementation issues. The new isomorphism can also be used to extend the usual ML type-inference algorithm, as suggested by Di Cosmo (1992). Building on that proposal, we introduce a better type-inference algorithm that behaves well in the presence of non-functional primitives like references and exceptions. The algorithm described here has been implemented easily as a variation to the Caml-Light 0.4 system.

Lambda Calculus Characterizations of Poly-Time

We consider typed λ-calculi with pairing over the algebra W of words over {0, 1}, with a destructor and discriminator function. We show that the poly-time functions are precisely the functions (1) λ-representable using simple types, with abstract input (represented by Church-like terms) and concrete output (represented by algebra terms); (2) λ-representable using simple types, with abstract input and output, but with the input and output representations differing slightly; (3) λ-representable using polymorphic typing with type quantification ranging over multiplicative types only; (4) λ-representable using simple and list types (akin to ML style) with abstract input and output; and (5) λ-representable over the algebra of flat lists (in place of W), using simple types, with abstract input and output.

Modularising the specification of a small database system in extended ML

Abstract A case study in the modular specification and refinement of a small database system is presented in Extended ML. Two similar requirements specifications are given and a program development step from each these is presented. The structure resulting from the first program development step is similar to that given in [FiJ90] and is presented as an answer to the challenge problem given in that paper, while the second development step is presented as a possible alternative which is more suited to the Extended ML style of program development. In the context of these two development steps the module facilities of Extended ML, their role in specification and program development and their ability to meet the challenge of [FiJ90] are examined.

Static type inference for parametric classes

Central features of object-oriented programming are method inheritance and data abstraction attained through hierarchical organization of classes. Recent studies show that method inheritance can be nicely supported by ML style type inference when extended to labeled records. This is based on the fact that a function that selects a field ƒ of a record can be given a polymorphic type that enables it to be applied to any record which contains a field ƒ. Several type systems also provide data abstraction through abstract type declarations. However, these two features have not yet been properly integrated in a statically checked polymorphic type system. This paper proposes a static type system that achieves this integration in an ML-like polymorphic language by adding a class construct that allows the programmer to build a hierarchy of classes connected by multiple inheritance declarations. Moreover, classes can be parameterized by types allowing “generic” definitions. The type correctness of class declarations is statically checked by the type system. The type system also infers a principal scheme for any type correct program containing methods and objects defined in classes.