Slicing Of Aspect-Oriented Programs And Its Application In Test Data Generation

With the rapid development of software technology, the scale of software system is expanded incrementally that makes higher demand on modular system. As a novel software development paradigm, aspect-oriented programming (AOP) enables modular implementation of cross-cutting concerns, which can improve the maintainability and reusability of the system. It is an efficient way that uses AOP to solve code tangling and scattering in object-oriented programming. In recent years, aspect-oriented technology has become the research hotspot in the current field of software engineering.Facing the more complex programs, how to effectively analyze and maintain the program, program slicing plays an important role, which has many applications in various fields such as program debugging, testing, etc. Regards to aspect-oriented programs, its peculiar structures and mechanisms pose difficulties for slicing. To better analyze and understand aspect-oriented programs, it is much important to research on slicing of aspect-oriented programs.Targeting slicing of aspect-oriented programs and its application, we make efforts on researching. Major work and contributions of this dissertation are as follows: First, taking the impact on dependences by advice priority into account, we propose advice fluctuation graph (AFG), as well as its generation algorithm. Then we construct system dependence graph for aspect-oriented program based on advice fluctuation (AOSDG). Second, on the basis of dependence analysis for aspect-oriented programs, we propose a dynamic slicing method of aspect-oriented programs based on reduced dynamic dependence graph. We construct reduced dynamic dependence graph, which reduce the number of nodes and edges of the system dependence graph, then calculate accurate dynamic slice for aspect-oriented program. Third, we use branch function minimization strategy, combined with the dynamic slicing method of aspect-oriented programs we proposed, to help generate test data in aspect-oriented software testing, and put forward a test data generation framework for aspect-oriented programs, which can improve the efficiency of software testing. In the paper, we describe the detailed process of the above methods with case study. By using the methods, we can express the dependences in aspect-oriented programs more precisely with relatively low cost, and get accurate slice of aspect-oriented programs, which used to aspect-oriented software testing, to assure the reliability of aspect-oriented programs, then make better application and popularization of AOP in current software development process.