基于二叉树搜索空间缩减的测试数据生成  

Test data generation based on binary tree search space reduction

作　　者：邵楠[1] 周雁舟[1] 惠文涛[1] 乔辉[1] 

机构地区：[1]信息工程大学,郑州450004

出　　处：《计算机应用研究》2014年第1期188-191,共4页Application Research of Computers

基　　金：武器装备预研重点基金资助项目(9140A15060311JB5201)

摘　　要：为了减小适应度函数计算量,提高测试数据自动生成效率,提出一种基于二叉树表示的搜索空间数据缩减方法。利用二叉树编码,记录全空间中的覆盖路径和路径长度;将目标路径和测试路径长度进行对比,去除路径长度相差较大的路径;利用遗传算法生成测试数据并同已有两种方法进行比较。实验结果表明,在保证软件测试数据正确生成的情况下,该方法在进化代数和运行时间上有明显优势,生成测试数据效率高。In order to reduce the calculations of fitness function and improve the data generating efficiency, this paper proposed a method of reducing data in whole space based on binary tree. The method used binary tree code to record the overlay path and the path length of each datum; compared the program paths with the target paths, left out the paths with larger path length difference; compared the data generated by genetic algorithm with other two methods. As is shown in the experiments, this method can improve the data generating efficiency, and has a great advantage in improving algebra and the running time in ensuring the generation of correct test data.

关 键 词：软件测试 二叉树 测试数据 路径覆盖 空间缩减 

分 类 号：TP311.56[自动化与计算机技术—计算机软件与理论]