A Prism for Research in Software Modularization Through Aspect Mining ∗

The Prism project develops tools and techniques for discovering non-localized units of modularity in large software systems. A non-localized unit of modularity is manifest by code crosscutting concerns or aspects, such as logging, tracing, synchronization, and persistence. The objective is to be able to analyze legacy code bases containing well over thousands of classes with over millions of lines of code. Aspect discovery is based on aspect mining techniques developed in this research. The successful discovery of aspects will give rise to aspect oriented refactoring possibilities of the software system mined. Prism is developed as plug-in for the Eclipse integrated development environment complementing the AspectJ Development Tools efforts (AJDT) enhancing Eclipse with features for aspect oriented software development.

[1]  Hans-Arno Jacobsen,et al.  Refactoring middleware systems: A Case Study , 2003 .

[2]  Welf Löwe,et al.  Automatic design pattern detection , 2003, 11th IEEE International Workshop on Program Comprehension, 2003..

[3]  Hans-Arno Jacobsen,et al.  Quantifying aspects in middleware platforms , 2003, AOSD '03.

[4]  Giuseppe Visaggio,et al.  Extracting reusable functions by program slicing , 1996 .

[5]  Hans-Arno Jacobsen,et al.  Refactoring Middleware with Aspects , 2003, IEEE Trans. Parallel Distributed Syst..

[6]  Martin P. Robillard,et al.  FEAT a tool for locating, describing, and analyzing concerns in source code , 2003, 25th International Conference on Software Engineering, 2003. Proceedings..

[7]  N. Loughran,et al.  Mining Aspects , 2002 .

[8]  Leon Moonen,et al.  Java quality assurance by detecting code smells , 2002, Ninth Working Conference on Reverse Engineering, 2002. Proceedings..