Model-driven Engineering for Dynamic Data Structures

Abstract

Model-driven engineering (MDE) has become a key technology in such diverse fields as signal processing, control engineering and software engineering. Our research has adopted the MDE paradigm for the analysis of complex software involving dynamic data structures, e.g., of device driver managers that employ custom list structures. Here, the central model studied by us is logic predicates that describe data structure shapes. This paper highlights aspects of our research on how shape predicates can support a range of activities: automated code generation for defensive programming, visualization for program comprehension and test case generation and formal verification for quality assurance. We discuss the commonalities and differences to the MDE of control-intensive systems and outline how our test case generation approach may be adapted to complex object-oriented software.