Specification-based Test Case Generation with Constrained Genetic Programming

Since current specification-based testing (SBT) faces some challenges in regression test case generation, we have already proposed a new method for test case generation that combines formal specification and genetic algorithms (GA). This method mainly reconfigures formal specifications though GA to generate inputs data that can kill as many as possible mutants of the target program under test. In this paper, we propose ideas to improve the operability and the accuracy of solution search of this method. Specifically, we propose a specification-level constrained operation using genetic programming and discuss effectiveness from the viewpoint of clarity of chromosome notation and ability to search for solutions.