北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2013, Vol. 36 ›› Issue (6): 18-22.doi: 10.13190/j.jbupt.2013.06.004

• 论文 • 上一篇    下一篇

面向路径测试的演化测试数据生成方法

王晓军, 闵丽娟   

  1. 南京邮电大学 信息网络技术研究所, 南京 210003
  • 收稿日期:2012-10-26 出版日期:2013-12-31 发布日期:2013-10-08
  • 作者简介:王晓军(1968—),女,副研究员,E-mail:xjwang@njupt.edu.cn.
  • 基金资助:

    国家自然科学基金项目(61003237)

Genetic Test Data Generation for Path-Oriented Testing

WANG Xiao-jun, MIN Li-juan   

  1. Institute of Information and Network Technology, Nanjing University of Posts and Telecommunications, Nanjing 210003, China
  • Received:2012-10-26 Online:2013-12-31 Published:2013-10-08

摘要:

针对现有方法主要用于顺序程序的路径覆盖的问题,提出了一种可应用于并发程序路径覆盖的适应值函数构造方法及实现算法,测试者可以运用该方法生成覆盖指定并发路径的测试用例.该方法的核心是通过计算测试用例的执行轨迹和目标路径之间的相似度来评估测试用例的适应值.实验结果表明,该方法为搜索可覆盖小概率目标路径的测试用例提供较强的导向性,并可以完全自动化地为每条目标路径构造出其适应值函数.

关键词: 软件测试, 演化测试, 路径覆盖测试, 遗传算法, 测试数据生成

Abstract:

Nowadays many researchers have proposed different methods to generate test data using genetic algorithm for a specified path.Existing methods primarily focus on the path coverage for sequential programs testing.An approach to construct the fitness function for test case generation in concurrent programs testing is provoked.This approach evaluates the fitness values of the test case by assessing the similarity between the path exercised by a generated input data and the target concurrent path.Algorithm for the fitness function is proposed.Studies indicate that this approach provides guidance for seeking the test case covering small probability target path, and generates fitness function for each target path automatically.

Key words: software testing, evolutionary testing, path-oriented testing, genetic algorithms, test data generator

中图分类号: