Sara: self-replay augmented record and replay for Android in industrial cases

Record-and-replay tools are indispensable for quality assurance of mobile applications. Due to its importance, an increasing number of tools are being developed to record and replay user interactions for Android. However, by conducting an empirical study of various existing tools in industrial settings, researchers have revealed a gap between the characteristics requested from industry and the performance of publicly available record-and-replay tools. The study concludes that no existing tools under evaluation are sufficient for industrial applications. In this paper, we present a record-and-replay tool called SARA towards bridging the gap and targeting a wide adoption. Specifically, a dynamic instrumentation technique is used to accommodate rich sources of inputs in the application layer satisfying various constraints requested from industry. A self-replay mechanism is proposed to record more information of user inputs for accurate replaying without degrading user experience. In addition, an adaptive replay method is designed to enable replaying events on different devices with diverse screen sizes and OS versions. Through an evaluation on 53 highly popular industrial Android applications and 265 common usage scenarios, we demonstrate the effectiveness of SARA in recording and replaying rich sources of inputs on the same or different devices.