Code smells detection and visualization: A systematic literature review

TL;DR

This systematic literature review analyzes techniques and tools for detecting code smells, highlighting the limited use of visualization methods and emphasizing the need for more diverse, language-agnostic, and objective detection approaches.

Contribution

The paper provides a comprehensive overview of code smell detection techniques and visual support methods, identifying gaps and future directions in the field.

Findings

Search-based detection is most common (30.1%)

Most studies focus on Java and open-source projects

80% of studies do not include visualization techniques

Abstract

Context: Code smells (CS) tend to compromise software quality and also demand more effort by developers to maintain and evolve the application throughout its life-cycle. They have long been catalogued with corresponding mitigating solutions called refactoring operations. Objective: This SLR has a twofold goal: the first is to identify the main code smells detection techniques and tools discussed in the literature, and the second is to analyze to which extent visual techniques have been applied to support the former. Method: Over 83 primary studies indexed in major scientific repositories were identified by our search string in this SLR. Then, following existing best practices for secondary studies, we applied inclusion/exclusion criteria to select the most relevant works, extract their features and classify them. Results: We found that the most commonly used approaches to code smells detection are search-based (30.1%), and metric-based (24.1%). Most of the studies (83.1%) use open-source software, with the Java language occupying the first position (77.1%). In terms of code smells, God Class (51.8%), Feature Envy (33.7%), and Long Method (26.5%) are the most covered ones. Machine learning techniques are used in 35% of the studies. Around 80% of the studies only detect code smells, without providing visualization techniques. In visualization-based approaches several methods are used, such as: city metaphors, 3D visualization techniques. Conclusions: We confirm that the detection of CS is a non trivial task, and there is still a lot of work to be done in terms of: reducing the subjectivity associated with the definition and detection of CS; increasing the diversity of detected CS and of supported programming languages; constructing and sharing oracles and datasets to facilitate the replication of CS detection and visualization techniques validation experiments.