The Research Of Android UI Property Testing Methods

With the rapid growth of technology of Android, the number of applications that people could choose is becoming bigger. Correspondingly, the quality required is becoming higher. Besides basic functions, good User Experience(short for UE) is required too. UE includes UI(User Interface) boot time, fluency, memory etc. High quality controlling during the process of android application development is the duty of a software test engineer.At present, most of the testing jobs of android are still finished manually. These methods of verifying application manually ignore the performance problem, and also time wasting. In this paper, we focus on the research of Android UI performance testing methods and propose an automated one. Works concluded as follows:1) Analysizing factors that effect the Android UI performance, and propose different testing methods respectively. Like: Logcat View to test UI boot time, command line “dumpsys meminfo” with Heap View to test memory, GT tools to test the fluency of UI.2) Propose the optimized method of testing UI boot time and fluency in performance, attached with the analysis of the method and advantage.3) Implement the optimized method to a concrete project. At first, we cover the preliminary stage work include requirement analysis, requirement assessment, testing environment and testing case designation. Then we implement the concrete testing method. Finally, we compare and analysize the result with the same kind of application as the testing project, and then propose the points that need optimizing and improvement.4) Analysing the testing efficiency and result, test engineer achieved a better result from the optimized method on UI boot time and fluency testing.This paper proposes a serial of automated methods include using automated script to retrieve boot log for testing the UI boot time, using MonkeyRunner to recorded automated UI slide and collect the FPS value. The experiment indicates that using automated method saves half the time of test engineer and improves the accuracy.