Analysis of Slice-Based Metrics for Aspect-Oriented Programs

Abstract

To improve separation of concerns in software design and implementation, the technique of Aspect-Oriented Programming (AOP) was introduced. But AOP has a lot of features like aspects, advices, point-cuts, join-points etc., and because of these the usage of the existing intermediate graph representations is rendered useless. In our work we have defined a new intermediate graph representation for AOP. The construction of SDG is automated by analysing the bytecode of aspect-oriented programs that incorporates the representation of aspect-oriented features. After constructing the SDG, we propose a slicing algorithm that uses the intermediate graph and computes slices for a given AOP. Program slicing has numerous applications in software engineering activities like debugging, testing, maintenance, model checking etc. To implement our proposed slicing technique, we have developed a prototype tool that takes an AOP as input and compute its slices using our proposed slicing algorithm. To evaluate our proposed technique, we have considered some case studies by taking open source projects. The comparative study of our proposed slicing algorithm with some existing algorithms show that our approach is an efficient and scalable approach of slicing for different applications with respect to aspect-oriented programs. Software metrics are used to measure certain aspects of software. Using the slicing approach we have computed eight software metrics which quantitatively and qualitatively analyse the whole aspect project. We have compiled a metrics suite for AOP and an automated prototype tool is developed for helping the process of SDLC.