启发式学习自动化程序指令的可达性测试
Automatic Statement Reachability Testing Using Heuristics Algorithm
作者：金虎(四川大学 计算机学院， 四川 成都 610064)；李志蜀(四川大学 计算机学院， 四川 成都 610064)；陈良银(四川大学 计算机学院， 四川 成都 610064)
Author：(School of Computer Sci., Sichuan Univ., Chengdu 610064, China)；(School of Computer Sci., Sichuan Univ., Chengdu 610064, China)；(School of Computer Sci., Sichuan Univ., Chengdu 610064, China)
收稿日期：2006-08-22 年卷（期）页码：2007,39(4):146-150
期刊名称：工程科学与技术
Journal Name：Advanced Engineering Sciences
关键字：自动化软件测试;约束满足问题;前向检测;启发式学习
Key words：automatic software testing;constraint satisfaction problems;forward checking;heuristic study
基金项目：四川省重点科技攻关资助项目(05GG021); 国家科技型中小企业技术创新基金资助项目(06C26225101730)
中文摘要
为了测试程序指令的可达性，需要搜索可触发包含该指令的控制流路径执行的测试用例。该问题可等价为路径约束满足问题的自动求解，求解结果即为验证指令可达的测试用例。为了提高自动求解的性能，在前向检测算法基础上改进设计了R_BinSearchFC算法，算法结合了边界测试数据选取和二叉树搜索技术，可有效提高对数据空间的搜索速度，并可扩展到在实数域空间的求解搜索。实验结果表明，采用了启发式学习的R_BinSearch算法可自动生成有效的测试用例，完成对指令的可达性测试。
英文摘要
To test the reachability of the program statements, it is necessary to find out test cases that can result in corresponding control flow path execution. It is equivalent to the resolving of Path Constraint-Satisfaction-Problems, and the resolution is the test cases that can meet the reachability-testing requirement. Based on Forward Checking Algorithm, R_BinSearchFC algorithm was proposed for better searching performance. Boundary test data selecting technique and binary tree searching method were employed in the algorithm for real type data set computation and the resolution time could be efficiently reduced. Experiment results showed good effect for automatic testing statement reachability.
【关闭】
