Evolutionary generation of test data for many paths coverage

Generation of test data for path coverage is an important issue of software testing, but previous methods are only suitable for the case that a program only has a small number of paths. We focus on the problem of generating test data for many paths coverage in this paper, and present a method of evolutionary generation of test data for many paths coverage. First, target paths are divided into several groups based on their similarity, and each group forms a sub-optimization problem, which transforms a complicated optimization problem into several simpler sub-optimization problems; then a domain-based fitness is designed when genetic algorithms are employed to solve these problems; finally, these sub-optimization problems are simplified along with the process of generating test data, hence improving the efficiency of generating test data. Our method is applied in 2 benchmark programs, and compared with some previous methods. The experimental results show that our method has advantage in time-consumption and the number of uncovered target paths. Our achievement provides an efficient way for generating test data of complicated software.