Names, types, and static semantic analysis

We describe a new approach for the specification and generation of the semantic analysis for typed programming languages. We specify context-sensitive syntactic properties of a language by a system of semantic rules. For various imperative programming language concepts, we discuss the required semantic rules. In particular, we show how overloading of programmer-defined identifiers can be handled. We propose an algorithm to solve these constraint systems efficiently, i.e., in time O(n^2) where n is the program size