摘要
基于缺陷模式的软件测试采用静态分析技术,误报问题比较严重.通过对区间数学理论扩展,提出了区间集的运算,以及实数、布尔变量、句柄变量、数组变量等不同数据类型的区间计算方法;采用May和Must集合来描述条件语句对区间的限定;通过修改数据流方程来跟踪控制流图遍历中的区间变化.将区间运算应用于基于缺陷模式的软件测试,实验结果表明,区间运算能够较好地减少误报和提高缺陷检测精度.
Lowering the false positive rate is a challenge for software testing based on defect patterns. This paper expands the typical interval computation and proposes a method for interval set computation. The proposed interval computation of various data types include real number variable, Boolean variable, pointer variable and array variable. May and Must set are used to compute interval restricted by conditional statement. Updated data flow equation is used to compute variables＇ interval in control flow traverse. The interval computation is used in software testing based on defect patterns. Experimental results show that it can reduce false positives and provide necessary information for defect testing.
出处
《计算机辅助设计与图形学学报》
EI
CSCD
北大核心
2008年第12期1630-1635,共6页
Journal of Computer-Aided Design & Computer Graphics
基金
国家“八六三”高技术研究发展计划(2006AA01Z184,2007AA010302)
关键词
软件测试
静态分析
区间运算
缺陷模式
状态机
software testing static analysis interval computation defect patterns state machine
