Fraunhofer-Gesellschaft

Publica

Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

CoVeR: Component-connector view reconstruction

 
: Forster, Thomas
: Knauber, Peter; Knodel, Jens

Mannheim, 2009, IV, 90 pp.
Mannheim, Hochschule, Master Thesis, 2009
English
Master Thesis
Fraunhofer IESE ()
software architecture; view; reverse engineering; SAVE; CoVeR

Abstract
Software architecture is the means to cope with the complexity of large software systems. Typically architecture separates different concerns: at development time modules and relationships are in the focus, at deployment time binaries and physical nodes with their hardware constraints are considered. At execution time, runtime structures like communication protocols, threads, processes or sockets are of interest. Understanding the principal units of computation and analyzing certain runtime qualities of a system (e.g., availability or security and certain aspects of efficiency) requires having a Component and Connector view.
The Component and Connector view addresses the communication and distribution related concerns at architectural level. It provides abstractions like logical components and connectors, ports, roles and protocols. As a matter of fact this kind of information is frequently missing or outdated in architectural documentation, so systematic reasoning about components and connectors and hence runtime structures and their qualities, is practically impossible.
To overcome this issue, a method for reconstructing Component and Connector models from static source code is presented in this thesis. The method provides a detailed description of the reconstruction process and comes along with an Eclipse based tool support. The tool guides the user through the reconstruction process and automates error-prone, time consuming and complex reconstruction steps.
As a proof of concept for the method, the Component and Connector model of a system in the ambient assisted living domain was successfully reconstructed. The applicability and suitability of the CoVeR method was shown and the approach promises to deliver practical relevant reconstruction results in an effort-efficient manner.

: http://publica.fraunhofer.de/documents/N-131387.html