Flexible Static Semantic Checking Using First-Order Logic

Introduction Many errors in programs can be detected statically with the right tools. Compilers usually report some basic errors, and a variety of analysis tools can reason about programs based on static information such as data flow graphs. For many interpreted or experimental languages, however, these tools are lacking. Using traditional techniques for developing analysis tools, the cost of developing good tools may not be deemed worthwhile for languages still under development or languages with a small user base. This poster presents a static semantic checker implemented using xlinkit [2], a COTS constraint checker, and a custom set of semantic rules expressed in first-order logic. This approach has made it easy to turn assertions about program structure into the parts of a working checker. Our checker is for Little-JIL, a maturing graphical process programming language. The checker features visual error reporting and can e#ectively detect and report over twenty conditions includ