Design And Implementation Of Code Clone Visualization System

Duplication of code is a common phenomenon in the developmentand maintenance of large software systems. The detection and removal of duplicated code has become a standard activity during the refactoring phases of a software life-cycle. However, with the continuous expansion of software systems, the number of clones code detected by clone detection tools is enormous. Analysts do not know how to start the work when facing a large of irregular clone code informations, which has brought great challenges to the refactoring and maintenance work.Visualization technology solve the problem how to display the data which has opened a new chapter to the analysts. The visualization process is based on qualitative or quantitative raw data, in the form of a graphical display out of view through FIG available for exploration, testing and interaction.Each of the present clone code visualization methods and models has its own advantages and disadvantages, and how to make better use of these visualization methods and models to help analysis look into the detection result, clone analysis system cloning distribution relationship still requires a combination of practical problems. At the same time, due to the different realization’s principle of clone detection tools, the form of the detected results are not the same, which make analysts review and visualization difficult.Therefore, the detection result files of some mainstream clone detection tools will be unified format in this system, provides the basis data flows for visualization. At the same time, the paper researches and implements multiple clone methods of visualizaiton in different perspective, and interactively show system clone detection results. From the macro perspective, the visualization methods include the Sunburst View which show the system’s clone distribution, and the Edge Bundle View which show the system’s clone relation. Frome the detail perspective, the visualization methods are under:one is the Heat Map View which show the subsystem’s clone relation, the second is the Clone Graph View which show relation of clone classes distrubuted in different files, and the last one is the Tree Map view which show the distribution of clone fragments in one clone class. The system also combine the visualization methods with the source code view.At last, a prototype of the visual system of clone code is designed and tested.It is suitable for the open source software projects of different languages, and supports friendly interactive function. The system provide an additional work for the refactoring and maintenance of code clone analysts and improve the efficiency of the analysis to some extent, which have important theoretical and application value.