Automated analysis of variability models: The SeVaTax process

Abstract

Variability management includes a set of techniques and methods for defining, modeling, implementing and testing variabilities within the development of a Software Product Line (SPL). Within the testing activity, several approaches have proposed novel techniques for automatic analysis of variability models. However, in spite of the research community has reached some consensus about the base scenarios that should be evaluated, the large number of modeling approaches makes that the way of evaluating those scenarios is still extensively researched. In this work we propose the SeVaTax process which takes variability models based on orthogonal variability model (OVM) primitives as inputs, and generates a formal model representation. Then, it uses a SAT-based solver for analyzing a wide set of validation scenarios and provides a different level of responses, even proposing some specific actions for correcting the models. Finally, we compare our proposal to others in the literature, based on the supported validations.