A cost-effective adaptive random testing algorithm for object-oriented software testing

Abstract

The rapid development of object oriented programming (OOP) technology has made it one of the mainstream programming technologies that has been widely used in the design and development of object oriented software (OOS). The inheritance, encapsulation and polymorphism properties of object-oriented language can improve the reusability, scalability and interoperability of software while increasing the difficulty of testing OOS. Researchers have proposed a variety of testing methods to test OOS among which random testing (RT) has been widely used due to its simplicity and ease of use. An OMISS-ARTsum algorithm is proposed in this paper that uses improved OMISS random test FSCS-ART with max-sum standard, which is an implementation version of fixed-sized-candidate-set ART. The OMISS-ARTsum algorithm calculates the total distance between a candidate test case and the executed test case set before the next test case is selected from the set of candidate test cases. Unlike the traditional max-sum based FSCS-ART algorithm, OMISS-ARTsum does not calculate the distance between each executed test case and the candidate case and then sum up the total distance, but uses the method of summing up all the executed test cases and the candidate cases. The information of executing test cases is saved as a whole and the distance between the executed test case set and candidate cases is calculated at the same time. Experiment shows that compared to the OMISS-ART algorithm, the proposed OMISS-ARTsum algorithm can reduce the time overhead.