ASE 2017

2017 32nd IEEE/ACM International Conference on Automated Software Engineering (ASE 2017), October 30 – November 3, 2017, Urbana-Champaign, IL, USA

Desktop Layout

Mobile Development
Technical Research
SimplyDroid: Efficient Event Sequence Simplification for Android Application
Bo Jiang, Yuxuan Wu, Teng Li, and W. K. Chan
(Beihang University, China; City University of Hong Kong, China)
Abstract: To ensure the quality of Android applications, many automatic test case generation techniques have been proposed. Among them, the Monkey fuzz testing tool and its variants are simple, effective and widely applicable. However, one major drawback of the family of Monkey tools is that they often generate a large number of events in a failure-inducing input trace, which makes the follow-up debugging activities hard to apply. It is desirable to simplify or reduce the input event sequence while triggering the same failure. In this paper, we propose an efficient event trace representation and the SimplyDroid tool with three hierarchical delta-debugging algorithms each operating on this trace representation to simplify crash traces. We have evaluated SimplyDroid on a suite of real-life Android applications with 92 crash traces. The empirical result shows that our new algorithms in SimplyDroid are both efficient and effective in reducing these event traces.


Time stamp: 2019-06-19T15:42:02+02:00