| Summary: | 碩士 === 國立臺北科技大學 === 資訊工程系研究所 === 103 === Android devices are getting more prevalent recently and the number of android applications (apps) has increased dramatically. The operations of Android apps usually involve varied gestures and rich events from touchscreens and sensors. Therefore, testing tools of Android apps can be different from those of traditional software applications. Currently, most of existing Android capture and replay testing tools can only capture user interactions with UI components and are unable to record and replay complicated gestures. This thesis presents a capture and replay method for testing Android apps. Particularly, the method recognizes user operations via analyzing Android low-level event logs. The events sequences will be analyzed and user operations including gestures, key press and sensor events are recognized and identified. The method then converts the Android low-level events into corresponding keywords and generates keyword-driven testing scripts which can be easily modified to add assertions. When executing the testing scripts, Android low-level events which represent original user operations will be reproduced for the target device and injected to Android input subsystem. Thus, the complicated gestures, operations of non-GUI components and sensor events can be replayed across different devices. Moreover, a tool is implemented to automate the proposed method. A case study is presented to illustrate the usefulness of the method. By using the supporting tool, the testing time and cost can be largely reduced.
|
|---|
