Code Clones

finished by Reto Geiger.

Final Thesis (PDF, 4 MB)

The detection of code clones is an essential problem in software maintenance and evolution. Many techniques for defining and detecting such clones have been proposed. Their potential to analyze the evolution of a software system are promising and should be combined with other techniques in software evolution analysis.

The goal of this diploma thesis is to apply Code Clone detection tools and techniques on a case study to identify a set of code clones. On the way, the effectiveness of the applied code clone techniques should be evaluated and discussed. Further, the detected code clones should then be combined with our work on software evolution analysis: based on structural and change smells provided in our Release History Database (RHDB), the detected code clones should be compared to the smell data in the RHDB to see whether the code clone detection is a good supplement of the analysis, or if it adds more value to the result when both approaches are combined. The results of the code clone analysis and the combination with smells should be integrated and represented in a prototype tool. The tool that synthesizes the different detection results shall be realized as an Eclipse plug-in.

Prof. Harald Gall, Beat Fluri