Parallel Clone Code Detector in MapReduce

摘要

Programmers often copy code to improveefficiency, and different developers may write the same codeindependently, these behaviors bring clone code to theproject. Clone code makes the project hard to maintain andweakens the robustness, and the bugs in these code segmentswould undermine the whole project. The state-of-the-artclone code detectors are either not able to find code withsame semantics, or computationally expensive. And if clonecode detector is to be performed on plenty number of code,the main memory of one machine may not able to hold allthe information.In this paper we focus on the parallel of the clone codedetector, we utilize the Program Dependence Graph (PDG)-based clone code detection method, which can not onlycheck the code in contiguous syntax, but also the code withthe same semantics. We present an approach to parallel theisomorphism matching in the PDG. By using MapReduceparadigm, we dramatically enhance the speed of thismethod.