Mutation Operators for Preprocessor-Based Variability

Mutation testing has emerged as one of the most promising techniques to increase the quality of software-intensive systems. In mutation testing, random faults based on a predefined set of mutation operators are automatically injected into a program to evaluate test suites. The effectiveness of mutation testing strongly depends on the representativeness of the mutation operators. Existing operators are not sufficient to represent typical faults caused by variability. Thus, we propose a set of mutation operators for software with preprocessor-based variability. We derive the operators systematically based on a taxonomy of variability-related faults and evaluate them by investigating their applicability to real-world faults that have previously been identified in research on configurable software systems. Our goal is to leverage mutation testing to highly-variable software for its practical application and to enable empirical evaluation of testing techniques.