The Design And Implementation Of Test Suit Augmentation Based On Predicate Adaptive Random Testing

With the development of computer science and software engineering, software systems have been used in many fields. In the software development life cycle, software testing plays an important part. As we know, software testing is complicate and expensive. Especially, when the software evolves, the probability of software fault will be higher. Therefore, it is necessary to find a solution to detect the bugs and the solution should be high fault detection, lowcost and feasibility.Regression testing is an important part in software development and mantainence, which aims to make sure the correctness of software when it evovles. In order to validate the software, we should test the system as long as the software is modified by some reasons. In software engineering, every software system has the original release, which we call "base line’. The base release usally has a test suit T. The management and maintennace of T is very complicate in regression testing. Regression test selection is the technique to select test cases from Tfor the modified software, and expects the selected test suit detect the faults in the new program. G.Rothermel, et al. presents the safe test case selection methods, which means selecting all the test cases that execute the modified part compared to the original software from T. Howerver, in regression testing, the original test suite may not satisfy new requirements due to the changes of software, thus new test cases should be generated.This paper presents a new test suit augmentation method based on predicate adaptive random testing(P-ART).The novel feature of our method is the combination of white box technique of software impact analysis and black box technique of adaptive random testing(ART). A new distance metric, predicate-metric, is introduced to guide ART for test suite augmentation. The experiment results show that our method could achieve higher branch coverage. It is easier for implementing automated tool for test suit augmentation compared with methods presented by others.