Automated Reverse Engineering of the Technology-Induced Software System Structure.

Evolving system architectures can be complex and difficult to understand, leading to problems such as poor maintainability. Automated reverse engineering of system structure models from source code can mitigate these problems and facilitate architectural decisions. However, identifying components with their interfaces can be challenging because components are often implemented in different frameworks and interconnected in complex ways. Our approach aims to create software models using reusable concept descriptions for reverse engineering. We use structural-level mapping rules to reconstruct the static system structure from source code, assuming that the technology used can determine the components with their interfaces and deployment. We evaluate our approach on four public reference systems. The analyses show that technology-specific rules already provide good results, but the addition of project-specific rules further improves reverse engineering.