| Code cloning complicates maintenance and hampers evolution of large software systems as it degrades their design and structure. Systematic management of software clones has the potential to translate into significant budget savings. Although various aspects of clone management have been addressed by academic research, practical application has been hampered by the lack of adequate tools and processes.This thesis research defines, implements and tests a comprehensive process for analyzing software clones in large bodies of source code. This process provides software practitioners with a necessary set of practical tools to detect, analyze, categorize and remove clones.A solution to extending a text-based clone identification technique to detect partial clones is described and integrated with an existing clone detection tool.A prototype of an interactive visual clone management tool that analyzes detected clones, clusters them and presents them to the user is introduced. This tool enables software practitioners to view, analyze, and utilize clone data to pursue their possible elimination.This thesis evaluates the described process by applying it to a commercial software system and analyzes the results. |
