Unveiling Inefficiencies in Open-Source Code Using Multistage Analysis with Software Metrics

Software development is challenging due to its technical complexity and time-consuming nature. To overcome these difficulties, various technical solutions have been introduced. In commercial software development, code repositories serve as valuable resources, reducing the time and cost involved in the process. The utilization of pre-developed open code repositories has proven to reduce development time. However, ample amount of work has notdetermined whether these repositories are testable, maintainable, free of dead code, and have a concise implementation of equivalent algorithms. The objective of this article is to address this gap by thoroughly analyzing the complexity and maintainability of code repositories, determining the impact of removing dead code on size, complexity, and maintainability. For this study, a total of 200 Python open-source code were analyzed using RADON, a widely-used metric tool for assessing cyclomatic complexity, size, volume, and maintainability. The identification of dead code within the repositories was accomplished using Vulture, supplemented by expert evaluation. It has been revealed that the majority of the examined code included dead code, and the removal of this code led to a significant reduction in cyclomatic complexity, volume, and size, whileimproving code maintainability, as observed by the Mann Whitney U test. The study concludes that the blind use of open-source code is not safe. It strongly recommends the community to thoroughly explore and examine such code from different perspectives before actual implementation. The novelty of this study lies in the use of multiple software metrics in a multi-stage analysis to examine the impact of removing dead code on program complexity, size, and maintainability.