LAUSR

Software verification can ensure the software quality by inspecting the properties of program. A key issue for software verification is to check whether the software can meet user requirements especially when the requirements change frequently. To tackle this problem, we propose an approach to verify the program by inspecting the internal relations with the user requirements. In the approach, the constraints in the requirements are represented by a concern-based model defined in our previous work by Liu et al. and the internal relations of program are extracted based on static analysis methods; then, a framework of verification system is defined to inspect whether the program can satisfy the constraints for discovering the errors with their locations. The main contribution of this paper includes: 1) kinds of internal relations of program are defined and their calculation methods are given to transform the source codes to a formalized model, which is taken as the object to be verified and 2) formal description of verification system framework is given to support the automation of verification process. Since the verification tasks can be set freely based on the requirements in the system, the proposed approach can help developers to cope with the change of requirements better.