Code Smells Detection and Visualization: A Systematic Literature Review

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