Mapping Code Clones Using Incremental Clone Detection

Abstract

Understanding the evolution of duplicated source code is necessary to investigate the phenomenon of cloning. To trace individual code clones across multiple program versions, clones of consecutive versions need to be mapped. Current methods detect clones of different versions first and then map detected clones retroactively. I present an approach to map individual clones between consecutive program versions during clone detection. The method is integrated into a token-based incremental clone detection algorithm. Clones are mapped according to the changes made to the source files of the program between versions.