GUPRO is an integrated workbench to support program understanding of heterogenous software systems on arbitrary levels of granularity. GUPRO can be adapted to specific needs by an appropriate conceptual model of the target software. GUPRO is based on graph-technology. It heavily relies on graph querying and graph algorithms. Source code is extracted into a graph repository which can be viewed by an integrated querying and browsing facility. For C-like languages GUPRO browsing includes a

more »

... treatment of preprocessor facilities. This paper summarizes the work done on GUPRO during the last seven years. ½ ÖØ ß ÃÙÐÐ ß Ê Ö ß Ï ÒØ Ö (ii) comprehending software systems, e. g. learning what software components do, how they operate, and how they interact, (iii) evolving software systems, e. g. correcting, changing, adapting, and extending software systems. Activities in reverse engineering and in program comprehension, during isolated maintenance phases as well as throughout incremental software development processes, follow an Extract-Abstract-View-Metaphor [27]. Data about source code artifacts are extracted into a software repository. Using various analysis techniques, these data are abstracted to provide a deeper understanding of the software system. Abstractions of the system are then viewed by appropriate visualization means. Source code Visualization extract view modify abstract Repository