Identification of file and directory level near-miss clones for higher level cloning

Abstract

The presence of code cloning, which speaks to duplicate segments of code, has been archived to happen much of the time in programming frameworks. The principle reason for cloning is reusing the part of the code that plays out some usefulness by duplicating and rewriting it to another area in the source code. The concentration here is to detect duplication in software program, which is a noteworthy reason for poor structure in programs. The key, novel part of our duplication-detection approach is detection of near-miss clones at higher level of granularity i.e, the directory and file level. Our work is a progress over past work around there as earlier the granularity of clone detection was method level. The technique adopted incorporates a novel hybrid approach using Abstract Syntax Tree and metrics to achieve precision. The results indicate the ability to detect Type-1, 2, 3 clones at directory and folder level.