Realistic context generation for mobile app testing and performance evaluation
Manoj R. Rege, Vlado Handziski, Adam Wolisz · 2017
Mobile app testing and evaluation requires exposing the app to a wide array of real world context conditions viz. location, sensor values, network conditions etc. Such comprehensive context conditions are difficult to create in a development environment on a real device, therefore, simulating them in a mobile emulator is a promising alternative. We present ContextMonkey, a framework for context generation in a mobile emulator. ContextMonkey can generate realistic context by leveraging traces in correlated and interdependent way from heterogenous sources: remote trace databases, models, and trace files. It eases the burden on developers to collect correlated traces, convert them to a common format, and feed them to the emulator in an orchestrated manner. We present examples demonstrating the simplicity and potential of ContextMonkey in app testing scenarios.