CLAIRE: Combining Sets, Search, and Rules to Better Express Algorithms

This paper presents a programming language that includes paradigms that are usually associated with declarative languages, such as sets, rules and search, into an imperative (functional) language. Although these paradigms are separately well-known and available under various programming environments, the originality of the CLAIRE language comes from the tight integration, which yields interesting run-time performances, and from the richness of this combination, which yields new ways to express complex algorithmic patterns with few elegant lines. To achieve the opposite goals of a high abstraction level (conciseness and readability) and run-time performance (CLAIRE is used as a C++ pre-processor), we have developed two kinds of compiler technology. First, a pattern pre-processor handles iterations over both concrete and abstract sets (data types and program fragments), in a completely user-extensible manner. Second, an inference compiler transforms a set of logical rules into a set of functions (demons that are used through procedural attachment).