一种利用模式信息的XPath测试集约简方法  

Test Suite Reduction for XPath Queries Using Schema Information

作　　者：郑黎晓[1] 常青玲[2,3] 

机构地区：[1]华侨大学计算机科学与技术学院,福建厦门361021 [2]中国科学院计算机网络信息中心,北京100190 [3]中国科学院大学,北京100190

出　　处：《小型微型计算机系统》2014年第12期2670-2674,共5页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61003138)资助;国家科技基础条件平台基金项目(BSDN2009-18)资助;华侨大学人才引进科研启动基金项目(12BS215)资助

摘　　要：软件测试是保证软件质量的一种重要手段,测试数据的数量和质量决定着测试的成本和效率.在基于范畴划分的XPath测试中,存在着测试数据集规模过大、测试开销过高的问题.为此,提出一种利用XM L模式信息有效约简XPath测试集的方法.该方法首先计算XML模式中的结构约束信息,例如元素允许出现的次数、元素之间的父子关系、兄弟关系等;然后利用这些信息去除范畴划分时产生的无效选择和选择的无效组合,从而减少最终测试集中的无效XML文档.理论分析和实验表明,在揭错能力较强的两种范畴划分覆盖准则下,该方法能有效减小测试集规模,从而降低测试成本、提高测试效率.Testing is an important technique to ensure the producing of high-quality software. The quantity and quality of test data significantly affect the cost and efficiency of software testing. In Category-Partition based testing for XPath queries, there exists a problem that the scale of test suite may be too large which consequently may cause high testing price. To this end, this paper proposes an effective test suite reduction method for XPath queries using XML schema information. The method first computes the structure constraints information in XML schemas, such as the allowed numbers of occurrences of elements, the parent-child or sibling relationships between elements, etc. , and then by utilizing this information eliminates invalid choices and invalid combinations of choices in category partitioning phase, which results in the reduction of invalid XML documents in the f＇mal test suite. Theoretical analysis and experimen- tal results show that for the two Category-Partition coverage criteria with high fault detection ability, the method can effectively reduce test suite, and thus can significantly decrease testing price and increase testing efficiency.

关 键 词：XPATH XML模式 范畴划分 测试集 约简 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]