Ripple is a dedicated scripting language for the Semantic Web. Ripple programs not only query the Semantic Web, but also reside within it as RDF data structures, forming a global network of interlinked programs. Ripple is best classified as a relational stack language, closely related to functional stack languages such as Joy, Factor and Cat. As a Semantic Web interface, Ripple is a fast, text-based linked data crawler and browser with all of the flexibility of a Turing-complete programming language.
A Paper on Ripple
Abstract. The idea of linked programs, or procedural RDF metadata, has not been deeply explored. This paper introduces a dedicated scripting language for linked data, called Ripple, whose programs both operate upon and reside in RDF graphs. Ripple is a variation on the concatenative theme of functional, stack-oriented languages such as Joy and Factor, and takes a multivalued, pipeline approach to query composition. The Java implementation includes a query API, an extensible library of primitive functions, and an interactive command-line interpreter.
Shinavier, J., Functional Programs as Linked Data (PDF/HTML). In 3rd Workshop on Scripting for the Semantic Web, Innsbruck, Austria, June 2007.