อิทธิพลของการทำรีแฟคทอริงต่อการตรวจทานโค้ดในโครงการซอฟต์แวร์โอเพนซอร์ส

Abstract

Recently, the use of open source software (OSS) has been growing. The OSS users claim that OSS is reliable in perspective of functional correctness because it is either developed or improved by various experienced developers. Rather than focusing on all quality aspects, the OSS developers emphasize on software functionality based on software requirement specifications. This characteristic will impose the poor code quality, which possibly introduces code smell in software. The code smells represent design situations that can affect the software quality, such as usability and maintainability. The well-known practice that the software engineering community often used to mitigate code smells is refactoring. The goal of refactoring is to improve the software design and code structure without changing the software's functionality. Typically, the OSS developers perform the code review to recover the defects in modified code. Based on the literature review, there is no study investigating the refactoring suggestions received from the code review process. To better understand the code review process in OSS projects, the author would like to collect empirical evidence that the code reviewers advised the developers to remove the code smell by refactoring. In this research, the author would like to investigate the relationship between code review's comments and refactoring based on detected code smell. Additionally, the author would like to classify the groups of code smell existing in OSS projects by adopting the algorithm, called Latent Dirichlet Allocation. The author does examine two OSS projects, including OpenStack and WikiMedia, which are contributed by many developers and code reviewers. The results of this study show that the reviewer most rarely concerned about refactoring. There were few reviewes who were attempting to encourage developers to improve the code by refactoring methods. However, the trend of comments related to code smells is increasing. This evidence indicates a good sign that the reviewers concentrate on improving the quality of software by refactoring. The author believes that this research would increase the knowledge and empirical evidence about the quality of OSS projects. Finally, the author expects that this research would also help software developers realize on employing software engineering practices, such as refactoring on both the traditional software development and OSS development.