基于FSM的测试用例生成算法

研究了基于FSM模型的测试用例自动生成算法来验证实现的FSM与规格FSM的一致性。基于状态机分解的思想提出了新的用例生成方法,DC-方法。相对其他通用方法而言,它不必依据测试树中所有的子路径集合,从而使得用例集减少,同时减少每次重置到初始态的次数,测试序列的长度也有所减少,节省了测试时间,提高了效率,并且具有良好的扩展性,可以方便应用于包含数据流的测试,只需在搜索到所有的DU-Path后直接经过运算就能得出测试用例。并给出了生成用例方法的正确性证明。

A Test Cases Generation Algorithm based on Finite State Models

Based on the Finite State Machine （FSM） ,this paper investigates the automatic generation of test cases which is an important problem for conformance testing. On the basis of spiting the machine, the paper presents a new method for generating test cases, DC - method, which allows for test cases gen- eration even in the absence of a transition cover set for the test tree in contrast to other general methods. In this case, it reduces the size of the test suit and the test sequence＇ s length as well as the frequency of the reset back to the initial state. The test time is also saved and the efficiency is improved. Moreover, it has a good expansibility. When testing models containing data flow, test cases could be directly generated by some operations just after finding all DU - Paths. The work also presents proofs of correctness for this generalization.