An Efficient Method for Automatic Generation of Linearly Independent Paths in White-box Testing

Xinyang Wang,Yaqiu Jiang,Wenhong Tian
2015-01-01
Abstract:Testing is the one of most significant quality assurance measures for software. It has been shown that the software testing is one of the most critical and important phases in the life cycle of software engineering. In general, software testing takes around 40-60% of the effort, time and cost. Structure-oriented test methods define test cases on the basis of the internal program structures and are widely used. Path-based test is one of the important Structure-oriented test methods during software development. However, there is still lack of automatic and highly efficient tool for generating basic paths in white-box testing. In view of this, an automatic and efficient method for generating basic paths is proposed in this paper. This method firstly transforms the source-code program into corresponding control flow graph (CFG). By modifying the original CFG to a strongly connected graph, a new algorithm (ABPC) is designed to automatically construct all basic paths. The ABPC algorithm has computational complexity linear to the number of total edges and nodes in the CFG. Through performance evaluation of many examples, it is shown that the proposed method is correct and scalable to very large test cases. The proposed method can be applied to basis path testing easily.
What problem does this paper attempt to address?