基于模块抽取的安卓应用变更模块测试技术的研究

安卓应用的可靠性研究是可靠、可信软件研究的重要组成部分。国内外主流的测试方法和策略主要针对整个应用的测试，存在自动化测试覆盖率不高、回归测试可扩展性不足等问题，导致新版本中功能变更模块无法得到针对性测试或者测试成本巨大，无法满足安卓应用频繁迭代更新功能的现状。针对此问题，本项目拟将开展如下研究：（1）多版本变更功能的识别技术，对安卓应用版本变更功能进行抽象并识别；（2）功能模块边界的抽取技术，定义紧耦合功能模块的边界，抽取出功能相对独立的变更功能模块，便于后续针对性测试；（3）变更功能模块的自动化测试技术，定位变更功能模块的测试入口点，对独立出来的变更功能模块进行高效的、系统化、针对性的测试。该项研究能帮助测试人员快速测试变更功能模块的可靠性和质量，减少测试时间，提高测试效率，减少用户在使用过程中发生的意外退出和异常错误，迭代式的提高安卓应用程序的质量，具有重要的研究价值和实用价值。

Research on the reliability of Android applications is an essential part of research on reliable and trustworthy software. The existing mainstream testing methods and strategies focus on testing on the whole app, with problems such as insufficient scalability of regression testing and low coverage of automated testing, which lead to failures in targeted testing of the changed function modules in the updated versions, or the testing cost is usually huge. Therefore, they cannot meet the current situation of frequent iteration and update of Android applications. To address it, this project plans to carry out studies from the following aspects: (1) Identification of changed functions in multi-versions, we abstract and identify the changed modules in the updated versions of Android apps during evolution; (2) Extraction of function module boundary, we define the boundary of function modules and extract the changed function modules with relatively independent functions for further targeted testing; (3) Automated testing of changed functional modules, we conduct efficient, systematic and targeted testing on the independent changed function modules through localizing the entry point of the changed modules. This study can help testers to quickly test the reliability and quality of the changed functional module, reduce the testing time, improve the testing efficiency, reduce exception errors, and iteratively improve the quality of Android applications, which has important research value and practical value.