Detection Of Duplicate Code

Abstract

Task of managing duplicated or “cloned” code has occupied the minds of programmers for the past 50 years. During this time, researchers and practitioners have developed a variety of techniques for removing or avoiding it by employing functions, macros and other programming abstractions. Functional abstraction was designed into early programming languages, such as Fortran and Lisp. Object-oriented programming, originating with Simula-67, has provided further mechanisms for parameterized reuse to avoid duplication. Aspect-oriented programming has allowed cross-cutting duplication to be abstracted. Engineering practices like Refactoring and Extreme Programming have promoted specific methodologies of abstracting duplicated code. In the last decade, a multitude of tools have been developed (both in research and in industry) that help programmers semi-automatically find and refactor existing duplication into functions, macros and methods. Given this long-term commitment to programming abstractions as a solution use “duplicated code” and “cloned code” synonymously to mean two or more multi-line code fragments that are either identical or similar, particularly in their structure. Duplicated code, it stands to reason that there should be little duplication left in practice. Keywords: Software maintenance, code duplication detection, code visualization