Advancing Code Smell Detection: Live Global Data Detection and Data Clumps Testcases

In the realm of software development, code smells pose considerable challenges during system maintenance and evolution. Although several tools identify code smells, few provide real-time detection and refactoring guidance, which are integral for efficient coding practices. This paper presents an advanced tool for live detection and semi-automatic resolution of data clumps and global data, distinct types of code smell in Java. Our tool’s prowess lies in its analysis of project structures, abstract syntax trees, and variable types to identify and address data clumps promptly. We successfully integrated our live detection technique into an IntelliJ plugin, achieving remarkable real-time detection with a median time of less than 1 s in 4 project case studies. Our method was extended to live detect Global Data. We are defining 9 base test cases for data clumps. The study contributes to the ongoing discourse on effective and efficient methods to remove code smells, enhancing software quality and developer productivity.