A privacy-aware, end-to-end, CFG-based regression test selection framework for web services using only local information

Web services are composable, interoperable, and autonomous which means that a single web service interaction could involve services written in several different languages provided by several different service providers. Such interactions hamper the development of RTS techniques because RTS techniques generally require some form of implementation details which service providers in separate autonomous systems are unlikely to expose. In this work, a privacy-aware, end-to-end regression testing framework employing a RTS technique using Control-Flow Graphs (CFGs) built using only locally available information at each service and a publish/subscribe mechanism will be presented. The proposed framework is unique because it does not require service providers to expose private and sensitive implementation details in order to participate in the regression testing framework. Also, a case study will be presented to highlight the approach and be used in an empirical study to evaluate the cost-effectiveness of the approach.