A Novel Approach to Detect Duplicate Code Blocks to Reduce Maintenance Effort

It was found in many cases that a code might be a clone for one programmer but not the same for another one. This problem occurs because of inaccurate documentation. According to research, the maintainers are not aware of the original design and thus, face the difficulty of agreeing on the system’s components and their relations or understanding the work of the application. The problem also occurs because of the different team of development and maintenance resulting in more effort and time during maintenance. This paper proposes a novel approach to detect the clones at the programmer side such that if a particular code is a clone then it can be well documented. This approach will provide both the individual duplicate statements as well as the block in which they appear. The approach has been examined on seven open source systems. Keywords—Clones; Program Dependence Graph (PDG); Control Flow Graph (CFG); Abstract Syntax Tree (AST)