The available evidence in a legacy software system, which can help in its understanding and recovery of its architecture are not always sufficient. Very often the system's documentation is poor and outdated. One may argue that the most reliable resource of information is the system's source code. Nevertheless a significant knowledge about the problem domain is required in order to facilitate the extraction of the system's useful architectural information. In this approach feature modeling is introduced as an additional step in a system's architectural recovery process. Feature modeling structures the system's functionality and supports reverse engineering by detecting the relations between source code elements and requirements. Tracing these relations may lead to a better understanding of the program's behavior and the recovery of various architectural elements. In this way, by providing a mapping between source code and features, the system's feature model supports program comprehension and architectural recovery. The approach is developed as first part of a migration methodology towards a component-based architecture of legacy systems. Recovered information about features and architecture is collected in a repository to enable a refactoring as next step. The approach is currently applied in a large project for reengineering of an industrial Image Processing System.

No abstract

In order to lower the risk, reengineering projects aim at high reuse rates. Therefore, tasks like architectural restructuring have to be performed in a way that developed new system architectures allow reuse of all valuable legacy systems' parts with minimal changes. During architectural restructuring there are two major types of modification: detection of architecture disproportions and their refactoring and detection of redundancies and their fusion. In this paper we introduce a method for applying domain knowledge for supporting these restructuring steps. The method operates on feature models. Words and terms of features and of architectural documents are analyzed by cluster analysis, information retrieval and metrics techniques. In this way, the method joins the approaches of feature analysis and of enhancing reengineering with domain knowledge by applying feature models for structuring the domain knowledge. The method results in clues and hints for the development of a new architecture. It provides an effective addition to the conventional software architecture design methods.The method was developed and applied in an industrial reengineering project within image processing domain. It has been proved to be applicable to large and complex systems even in case of heavy monolithic parts. We use examples from this project to illustrate the method.