GUI自动化测试用例生成策略的研究  

Study on generation strategies of GUI automation testing cases

作　　者：陆永忠[1] 汪春[1] 聂松林[2] 

机构地区：[1]华中科技大学软件学院,湖北武汉430074 [2]华中科技大学机械科学与工程学院,湖北武汉430074

出　　处：《系统工程与电子技术》2009年第1期174-177,共4页Systems Engineering and Electronics

基　　金：国家自然科学基金(50675074;50775081);华中科技大学自然科学基金(2007Q006B)资助课题

摘　　要：针对目前在图形用户界面(graphic user interface,GUI)自动化测试方法中存在的手工依赖性和测试缺乏准确性等问题,提出了一种改进的GUI自动化测试算法。该算法包括两种基于事件流图的GUI自动化测试用例生成策略:基于蚁群算法的日常冒烟测试用例生成策略和基于宽度优先搜索生成树的深度回归测试用例生成策略。将这两种策略应用于没有考虑分层的GUI事件流图模型中,得到标准GUI的测试用例,然后再进行GUI测试。结合Microsoft UI Automation框架和Visual Studio 2005开发平台,对该算法进行了实验研究。研究表明:该算法可以提高GUI测试的自动化程度和准确性。Aiming at the present problems in the course of GUI automation test such as large dependence on manual work, lack of testing accuracy and so forth, an improved automated GUI testing algorithm is proposed. It includes two strategies based on its event-flow graph： a daily smoke testing strategy based on the ant colony optimization algorithm （ACO） and a deep regression testing strategy based on the spanning tree algorithm. Two strategies are applied to the non-hierarchical GUI event-flow graph, and the standard testing cases are obtained. Then GUI test is done. With the help of Microsoft UI Automation and Visual Studio 2005, a testing study on the proposed algorithm is done. The studying results show that it can improve the automation and accuracy of GUI test.

关 键 词：软件测试自动化 冒烟测试 回归测试 蚁群算法 图形用户界面 事件流图 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]