Design and implementation of a customizable metrics plug-in in eclipse

To analyze the quality of a software system metrics can be used that measure attributes of the software's internal structure. However, all these metrics are limited to analyzing a certain aspect of software from a certain viewpoint only. Many metric tools exist that collect and visualize a standard set of software metrics. But since the actual suitable set of metrics depends on the viewpoint and purpose of the measurement or the person measuring, a metric tool should be customizable to enable its users to concentrate on and select only the set and configuration of metrics that are needed. This Bachelor thesis presents the design and implementation of a customizable metric framework which allows such customization by enabling the user to implement own metric and then to add them to the framework. It delivers all commonalities of metrics like selection of the analyzed software elements, storing of the results, or visualization of the results. So only the specifics of every metric like the algorithm calculating the results have to be implemented by the user. The calculation of metrics can be delegated to external metric tools. The developed framework is: - Extensible: new metrics can easily be added to the framework - Flexible: many different types of metrics are supported - Adaptable: The measurement can be adapted to specific needs by specifying the analyzed elements and metric-specific options - Customizable: the visualization can be customized to the user's needs by selecting the type of visualization and the set of results visualized - Repeatable: both a calculation's configuration and results are stored persistent, so the calculation can be repeated and a single calculation's results can repeatedly be visualized from different viewpoints. The framework was built as a plug-in of the software development platform Eclipse and is integrated into SAVE, a tool for Software Architecture Visualization and Evaluation. SAVE can extract and analyze the architecture of an existing software system but did not support the measurement of metrics. The implemented framework adds a mechanism to SAVE that enables the measurement of metrics. It can calculate, store the results of and visualize all metrics registered to it. The results of a metric's calculation can be visualized within different kinds of diagrams or within an existing architecture view offered by SAVE. Finally, in order to validate the metrics plug-in a set of metrics was implemented covering all aspects of the framework's functionality. Some metrics were as well implemented by other people to validate the understandability of the framework.