Analyzing variability: Capturing semantic ripple effects

Abstract

This paper shows how to incrementally analyze how variability described in the Common Variability Language (CVL) affects the semantics of a model in a domain-specific language (DSL). CVL is a generic language for modeling variability. Using Alloy for definition of semantics we perform analysis to capture the elements in the model, which are semantically affected by applying the variabilities specified by the CVL model. An extension to the CVL editor is provided to automate the analysis. To illustrate the approach, we combine CVL with the Train Control Language (TCL) to capture how the semantics of TCL models are affected when applying CVL to them. We show how the analysis can be applied e.g., for testing. © 2011 Springer-Verlag.