Aspect-oriented model-driven code generation approach for improving code reusability and maintainability

Abstract

Software development teams always need methods that can help in producing high-quality software with reduced development effort and delivery time. Model-Driven Engineering (MDE) as well as Aspect-Oriented Software Development (AOSD) techniques help in reducing the delivery time, and also positively contribute to quality of the produced software. Through the use of AOSD techniques in combination with MDE, an integration of excellent abstraction mechanisms of MDE and capabilities of AOSD with regards to modularity and composition of concerns can be perceived, which is expected to enhance the positive effects of both techniques. To this end, different integration approaches have appeared in literature, but aspect-oriented code generation has advantages over the other approaches. Consequently, a number of aspect-oriented code generation approaches have been offered, but all such approaches lack several features mandatory to materialize a workable integration of aspect technologies in the context of MDE. To address these issues, this research was conducted to present an approach for aspect-oriented model-driven code generation, which focuses on elaborating the conceptual relationship between design models and the implementation code, and exploits the same to obtain aspect-oriented code that is more reusable and maintainable. The key outcomes of this research are the elaboration of the conceptual mappings between elements of visual design and constructs of the code, mapping of the visual models to implementation-specific text-based models, and a technique for generation of aspect-oriented code. The applicability of the proposed approach is shown by the use of case studies, whereas the quality of the approach is measured using reusability and maintainability metrics. A comparison of the proposed approach with existing approaches substantiates its efficacy in terms of reusability and maintainability of code, showing an outperformance of other approaches by the proposed approach against 78% of the employed quality metrics.