Feature-Driven and Incremental Variability Generalization in Software Product Line

In the lifecycle of a software product line (SPL), incremental generalization is usually required to extend the variability of existing core assets to support the new or changed application requirements. In addition, the generalization should conform to the evolved SPL requirements which are usually represented by a feature model. In this paper, we propose a feature-driven and incremental variability generalization method based on the aspect-oriented variability implementation techniques. It addresses a set of basic scenarios where program-level JBoss-AOP based reference implementations respond to the feature-level variability generalization patterns. It also provides the corresponding guidance to compose these patterns in more complex cases. Based on the method, we present a case study and related discussions.