基于划分的等价类测试  

Equivalence testing based on partitioning

作　　者：浦云明[1] 陈黎震[1] 

机构地区：[1]集美大学计算机工程学院,福建厦门361021

出　　处：《计算机工程与设计》2009年第19期4429-4431,共3页Computer Engineering and Design

基　　金：福建省教育厅基金项目(JA07133);厦门市科技计划基金项目(3502Z20083025)

摘　　要：等价类的两个基本因素是健壮性和单/多缺陷假设。依据这两个特性,等价类分为弱一般等价类、强一般等价类、弱健壮等价类和强健壮等价类4种形式。等价类测试的数学基础是等价关系和划分,划分的原则是每个元素仅属于某一类,不同的类之间没有公共元素。常用的输入域等价类划分是将程序的输入域划分为若干子集合,并从每个子集中选择代表性数据设计测试用例,划分保证了等价类测试的完备性和无冗余性。分析了等价类划分设计原则,设计了划分测试实例。The two factors of equivalence class are robustness and single or multiple defects. The equivalence class is divisible into four types according to these two factors, which are weak equivalence, strong equivalence, weak robust equivalence and strong robust equivalence. The equivalence relations and partition are the basis ofthe equivalence testing. The partitioning principle is that every one element is only belongs to one partition, and no public element is within different partitions. In normal input equivalence partitioning, the program inputs are partitioned a couple of parts, and the representative data is selected as test case. Partitioning guarantees the completeness and non-redundancy of equivalence testing. The principle of equivalence testing is analyzed, and the test cases of equivalence testing are designed.

关 键 词：等价关系 等价类 划分 测试用例 可靠性 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]