KClone: A Proposed Approach to Fast Precise Code Clone Detection
KClone: A Proposed Approach to Fast Precise Code Clone Detection
复制标题
DOI：
--
复制
发表时间：
2009
期刊：
影响因子：
--
通讯作者：
Yue Jia;D. Binkley;M. Harman;J. Krinke;M. Matsushita;King’s College London
中科院分区：
文献类型：
--
作者：
Yue Jia;D. Binkley;M. Harman;J. Krinke;M. Matsushita;King’s College London
在克隆检测的所有应用中，拥有精确且高效的克隆识别算法非常重要。本文提出并概述了一种用于克隆检测的新算法 KClone，该算法结合了词法和局部依赖分析的新颖组合，以在保持速度的同时实现精度。本文还报告了使用我们一直在试验的 KClone 实现进行案例研究的初步结果。结果表明，与基于令牌和基于 PDG 的技术相比，KClone 能够找到类型 1、2 和 3 克隆。该论文还报告了 KClone 与 CCFinderX 性能对比的初步实证研究结果。
In all applications of clone detection it is important to have precise and efficient clone identification algorithms. This paper proposes and outlines a new algorithm, KClone for clone detection that incorporates a novel combination of lexical and local dependence analysis to achieve precision, while retaining speed. The paper also reports on the initial results of a case study using an implementation of KClone with which we have been experimenting. The results indi- cate the ability of KClone to find types-1,2, and 3 clones compared to token-based and PDG-based techniques. The paper also reports results of an initial empirical study of the performance of KClone compared to CCFinderX.