东北大学学报（自然科学版）

东北大学学报:自然科学版 ›› 2015, Vol. 36 ›› Issue (12): 1696-1700.DOI: 10.12068/j.issn.1005-3026.2015.12.006

• 信息与控制 • 上一篇    下一篇

基于复杂网络的类间集成测试序列生成方法

赵玉丽， 王莹， 于海， 朱志良   

  1. (东北大学 软件学院， 辽宁 沈阳110819)
  • 收稿日期:2014-11-01 修回日期:2014-11-01 出版日期:2015-12-15 发布日期:2015-12-07
  • 通讯作者: 赵玉丽
  • 作者简介:赵玉丽(1985-)，女，内蒙古赤峰人，东北大学讲师，博士； 朱志良(1962-)，男，辽宁沈阳人，东北大学教授，博士生导师.
  • 基金资助:
    国家自然科学基金资助项目(61202085，61374178，61402092)； 中央高校基本科研业务费专项资金资助项目(N130317001，N130417004).

An Inter-Class Integration Test Order Generation Method Based on Complex Networks

ZHAO Yu-li， WANG Ying， YU Hai， ZHU Zhi-liang   

  1. School of Software， Northeastern University， Shenyang 110819， China.
  • Received:2014-11-01 Revised:2014-11-01 Online:2015-12-15 Published:2015-12-07
  • Contact: YU Hai
  • About author:-
  • Supported by:
    -

RichHTML

0

PDF (PC)

740

摘要： 类间集成测试是面向对象软件测试的一个重要部分.将类抽象成节点，类间调用关系抽象成边，建立面向对象软件的复杂网络模型，根据软件网络的结构特征，分析类节点的影响力和复杂性，给出一种软件重要节点的度量方法.在此基础上，将错误被尽早发现的思想应用在类间集成测试排序中，在确保构造的测试桩复杂度较小的前提下，保证重要节点优先测试，得到了一种基于复杂网络的类间集成测试序列生成方法.最后，在开源软件DNS 1.2.0上进行分析和验证，证明了该方法的有效性.

关键词: 集成测试, 复杂网络, 节点重要性, 测试序列, 桩复杂度

Abstract: Inter-class integration test is a critical part of the object-oriented software testing. Representing a class by a node in a complex network， two classes are connected only if there exist an invoke relationship between them. Then， a complex network model representing object-oriented software could be constructed. According to the characteristics of the complex network structure， the influence and complexity of each class node was analyzed. Further， an approach for measure the significant nodes in the software network were provided. Moreover， the idea that error should be found as soon as possible was applied to the inter-class integration test order， an inter-class integration test order generation method was proposed， in which the test priority of the significant nodes and the lower test stub complexity were ensured. Simulation result on the open source software DNS 1.2.0 indicated the effectiveness of the proposed method.

Key words: integration test, complex network, node significance, test order, stub complexity

中图分类号: