No abstract

This paper presents a new and very rich class of (concurrent) programming languages, based on the notion of comput.ing with parhal information, and the concommitant notions of consistency and entailment.' In this framework, computation emerges from the interaction of concurrently executing agents that communicate by placing, checking and instantiating constraints on shared variables. Such a view of computation is interesting in the context of programming languages because of the ability to represent and manipulate partial information about the domain of discourse, in the context of concurrency because of the use of constraints for communication and control, and in the context of AI because of the availability of simple yet powerful mechanisms for controlling inference, and the promise that very rich representational/programming languages, sharing the same set of abstract properties, may be possible.

Concurrent Constraint Programming introduces a new and rich class of programming languages based on the notion of computing with partial information, or constraints, that synthesize and extend work on concurrent logic programming and that offer a promising approach for treating thorny issues in the semantics of concurrent, nondeterministic programming languages. Saraswat develops an elegant and semantically tractable framework for computing with constraints, emphasizing their importance for communication and control in concurrent, programming languages. He describes the basic paradigm, illustrates its structure, discusses various augmentations, gives a simple implementation of a concrete language, and specifies its connections with other formalisms. In this framework, concurrently executing agents communicate by placing and checking constraints on shared variables in a common store. The major form of concurrency control in the system is through the operations of Atomic Tell—an agent may instantaneously place constraints only if they are consistent with constraints that have already been placed—and Blocking Ask—an agent must block when it checks a constraint that is not yet known to hold. Other operations at a finer granularity of atomicity are also presented. Saraswat introduces and develops the concurrent constraint family of programming languages based on these ideas, shows how various constraint systems can naturally realize data structures common in computer science, and presents a formal operational semantics for many languages in the concurrent constraint family. In addition, he provides a concrete realization of the paradigm on a sequential machine by presenting a compiler for the concurrent constraint language Herbrand and demonstrates a number of constraint-based concurrent programming techniques that lead to novel presentations of algorithms for many concurrent programming problems.

We develop a model for timed, reactive computation by extending the asynchronous, untimed concurrent constraint programming model in a simple and uniform way. In the spirit of process algebras, we develop some combinators expressible in this model, and reconcile their operational, logical and denotational character. We show how programs may be compiled into finite-state machines with loop-fiee computations at each state, thus guaranteeing bounded response time.

Sketching is a software synthesis approach where the programmer develops a partial implementation -a sketch -and a separate specification of the desired functionality. The synthesizer then completes the sketch to behave like the specification. The correctness of the synthesized implementation is guaranteed by the compiler, which allows, among other benefits, rapid development of highly tuned implementations without the fear of introducing bugs.We develop SKETCH, a language for finite programs with linguistic support for sketching. Finite programs include many highperformance kernels, including cryptocodes. In contrast to prior synthesizers, which had to be equipped with domain-specific rules, SKETCH completes sketches by means of a combinatorial search based on generalized boolean satisfiability. Consequently, our combinatorial synthesizer is complete for the class of finite programs: it is guaranteed to complete any sketch in theory, and in practice has scaled to realistic programming problems.Freed from domain rules, we can now write sketches as simpleto-understand partial programs, which are regular programs in which difficult code fragments are replaced with holes to be filled by the synthesizer. Holes may stand for index expressions, lookup tables, or bitmasks, but the programmer can easily define new kinds of holes using a single versatile synthesis operator.We have used SKETCH to synthesize an efficient implementation of the AES cipher standard. The synthesizer produces the most complex part of the implementation and runs in about an hour.