A data analysis study of code smells within java repositories

Although code smells are not categorized as a bug, the results can be long-lasting and decrease both maintainability and scalability of software projects. This paper presents findings from both former and current industry individuals, aiming to detect tools that are commonly used as well as how long software developers spend on refactoring code. Based on the feedback from these individuals, a collection of smells were extracted from a small sample size of 100 Java repositories in order to validate some of the smells that are typically encountered. After analyzing these repositories, the smells typically encountered are Long Statement, Magic Number, and Unutilized Abstraction. The results of this study are applicable for developers and researchers who require insight in detecting prevalent code smells.

---

Opracowanie rekordu ze środków MEiN, umowa nr SONP/SP/546092/2022 w ramach programu "Społeczna odpowiedzialność nauki" - moduł: Popularyzacja nauki i promocja sportu (2022-2023).