IDE-based real-time focused search for near-miss clones

  • Authors:

  • Minhaz F. Zibran;Chanchal K. Roy

  • Affiliations:

  • University of Saskatchewan, Saskatoon, SK, Canada;University of Saskatchewan, Saskatoon, SK, Canada

  • Venue:
  • Proceedings of the 27th Annual ACM Symposium on Applied Computing
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Code clone is a well-known code smell that needs to be detected and managed during the software development process. However, the existing clone detectors have one or more of the three shortcomings: (a) limitation in detecting Type-3 clones, (b) they come as stand-alone tools separate from IDE and thus cannot support clone-aware development, (c) they overwhelm the developer with all clones from the entire code-base, instead of a focused search for clones of a selected code segment of the developer's interest. This paper presents our IDE-integrated clone search tool, that addresses all the above issues. For clone detection, we adapt a suffix-tree-based hybrid algorithm. Through an asymptotic analysis, we show that our approach for clone detection is both time and memory efficient. Moreover, using three separate empirical studies, we demonstrate that our tool is flexibly usable for searching exact (Type-1) and near-miss (Type-2 and Type-3) clones with high precision and recall.