Research On Duplicated Code Detection And Automatic Refactoring

In the “bad taste” of software system,duplicate code problem stands in the breach.At present,the detection and analysis of the duplicate code could play the major role in many areas of the software system,such as upgrading,patents and maintenance.Recently,identification and auto refactoring of code smell is one of the focus in the field of software engineering.Duplicated code is a kind of common code smell in the field of software engineering,and there is wild application in software maintenace,quality and other fields.At the same time,this kind of code smell is easy to quantitatively evaluate.So there is high research and study value of identification and auto refactoring of code smell.In the recent years,there are many studies about duplicated code around the world,and many of them proposed the way to identify and refactor the duplicated code.As for the identification of duplicated code,however,there still exist many shortbacks,and the skill of refactoring is not consummate and it is hard to wildly promote.In view of the above question,to make the existing skill perfect and practical,and to have a more consummate as well as more effective refactoring skill which is combined with design pattern,is the main content of the study.This paper aims at developing an automatic detection software which can detect duplicate code and its related defects.Project work focus on major open source software platform,evaluated several typical duplicate code detection and refactoring method.Based on evaluation results,a combined implementation framework based on template method was introduced.Based on the existing skill of duplicated code identification,the study will put forward a new method which is more practical,and design an algorithm which combined with design pattern.Finally,we will make it a software tool and apply it to the actual development work.