基于依赖结构的功能测试集排序方法  

Prioritizing functional test suites based on dependency structures

作　　者：曹丽娜[1] 高建华[1] 

机构地区：[1]上海师范大学计算机科学与工程系,上海200234

出　　处：《计算机工程与设计》2015年第5期1223-1227,1238,共6页Computer Engineering and Design

基　　金：国家自然科学基金项目(61073163);上海市引进技术的吸收与创新年度计划基金项目(12CH-19);上海市企业自主创新专项基金项目(沪CXY-2013-88)

摘　　要：为解决依赖结构优先级在测试用例权值相等时存在的问题,针对流行的深度优先搜索算法进行改进。通过结合测试用例之间的功能依赖和测试用例的代码覆盖率,推导出有向无环图和算法流程图;利用推导出的有向无环图和流程图,使用权值和代码覆盖率算出最长路径作为测试集,以达到缩减测试集,同时保证代码覆盖率的目的。结合实例,将依赖结构优先级和现存的技术进行对比,验证了依赖结构优先级技术在提高错误检测率方面的可行性和实用性。To solve the problem made by dependency structures prioritization （DSP）when equal weight appears between test ca-ses,a depth-first search algorithm was improved.The directed acyclic graph and the algorithm flow diagram were achieved by combining functional dependency between test cases and code coverage of test cases.Weight and code coverage were used to cal-culate the longest path as test suites based on the directed acyclic graph and the algorithm flow diagram to reduce the number of test suites and ensure the code coverage.At last,comparing DSP with other current technologies by experiments,experimental results show DSP has a great advantage in increasing the rate of fault detection.

关 键 词：依赖结构优先级 深度优先搜索算法 有向无环图 权值 代码覆盖率 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]