首页 | 官方网站   微博 | 高级检索  
     

ALL-Uses覆盖引导的回归测试用例生成方法
引用本文:于嘉维,贲可荣.ALL-Uses覆盖引导的回归测试用例生成方法[J].计算机工程与科学,2017,39(12):2278-2289.
作者姓名:于嘉维  贲可荣
作者单位:;1.海军工程大学电子工程学院
摘    要:基于扩展有限状态机(EFSM)的回归测试过程需要根据依赖关系变化对软件所做修改的影响域进行分析。为了针对软件某一功能进行修复,通常需要对多处代码进行同步修改,已有依赖分析方法在这种情况下暴露模型中触发条件和行为语句错误的效率不高。提出以ALL-Uses覆盖准则引导回归测试的方法,引入依赖关系变化因素的概念,修改待覆盖子路径的产生规则,对已有测试用例集中能有效覆盖子路径的用例进行选择、排序。针对已有测试用例无法覆盖的子路径,利用AOE活动图中求关键路径的方法将其补充为一条完整的迁移执行序列。选取三个软件进行实验,结果表明,本文方法可以在减小测试用例集规模的前提下有效提升ALL-Uses和植入错误的覆盖率,提高回归测试效率。

关 键 词:扩展有限状态机  回归测试  ALL-Uses  依赖分析
收稿时间:2017-07-13
修稿时间:2017-12-25

An ALL-Uses coverage guided regression test case generating method
YU Jia-wei,BEN Ke-rong.An ALL-Uses coverage guided regression test case generating method[J].Computer Engineering & Science,2017,39(12):2278-2289.
Authors:YU Jia-wei  BEN Ke-rong
Affiliation:(School of Electronic Engineering,Naval University of Engineering,Wuhan 430033,China)
Abstract:In the regression testing process based on extended finite state machine (EFSM), the influence domain of software modification should be analyzed according to dependence change. Because there is no consideration for multiple modifications to the software at the same time, the analysis method is used to expose the trigger conditions and behavior statement errors in the model, but its efficiency is not high. We propose an ALL-Uses coverage guided regression test case generating method. We introduce the concept of dependence factors, modify the production rules, and select and sort the test cases that can cover the sub path. Aiming at the sub paths which cannot be covered by test cases, we use the method of finding the critical path in the AOE activity diagram to supplement it into a complete migration execution sequence. Experiments on three different software show that this method can improve the coverage of ALL-Uses and implanted errors under the premise of reducing the size of the test suite, and enhance the efficiency of regression testing.
Keywords:extended finite state machine (EFSM)  regression testing  ALL-Uses  dependence analysis  
点击此处可从《计算机工程与科学》浏览原始摘要信息
点击此处可从《计算机工程与科学》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号