Inferring variability from customized standard software products

Abstract

Systematic variability management is an important prerequisite for successful software reuse. However, it requires significant effort and extensive domain knowledge to document and maintain information on variability. In this paper we present a tool-supported approach which supports semi-automatically inferring variability information from customized standard software products. The approach does not only enable the identification and documentation of variability information based on existing products, it is also capable of incrementally updating this information. To guarantee quick access to reusable code artifacts (e.g. requirements, features or software components), the presented solution stores these artifacts together with related requirements and a generated variability model in an asset repository. The tool-supported approach has been applied to customizations of Microsoft Dynamics AX ERP systems. Our experiences highlight the potential and benefits of our approach compared to manually gathering information on software variability.