Composing programming languages by combining Action-semantics modules

Abstract

This article demonstrates a method for composing a programming language by combining action-semantics modules. Each module is defined separately, and then a programming-language module is defined by combining existing modules. This method enables the language designer to gradually develop a language by defining, selecting and combining suitable modules. The resulting modular structure is substantially different from that previously employed in action-semantic descriptions. It also discusses how to resolve the conflicts that may arise when combining modules, and indicates some advantages that action semantics has over other approaches in this respect.