Challenges of Processing Data Clumps within Plugin Architectures of Integrated Development Environment
Nils Baumgartner, Elke Pulvermüller
TL;DR
The paper addresses the challenge of detecting and refactoring data clumps, a code-smell that can span files and IDEs. It proposes a centralized, AST-based detection approach exposed through a CLI/NPM package and Docker workflows to decouple detection from source access, enabling cross-environment applicability and reproducible automation. The approach emphasizes modular, reusable components and visualization tools to support developers and CI/CD pipelines while preserving data locality and security. This centralized model aims to improve software quality across languages and IDEs, reducing maintenance costs and enabling scalable, environment-agnostic refactoring workflows.
Abstract
In this study, we explore advanced strategies for enhancing software quality by detecting and refactoring data clumps, special types of code smells. Our approach transcends the capabilities of integrated development environments, utilizing a novel method that separates the detection of data clumps from the source access. This method facilitates data clump processing. We introduce a command-line interface plugin to support this novel method of processing data clumps. This research highlights the efficacy of modularized algorithms and advocates their integration into continuous workflows, promising enhanced code quality and efficient project management across various programming and integrated development environments.