Code Smells: A Comprehensive Online Catalog and Taxonomy

Abstract

Context: Code Smells—a concept not fully understood among programmers, crucial to the code quality, and yet unstandardized in the scientific literature. Objective: Goal (#1): To provide a widely accessible Catalog that can perform useful functions both for researchers as a unified data system, allowing immediate information extraction, and for programmers as a knowledge base. Goal (#2): To identify all possible concepts characterized as Code Smells and possible controversies. Goal (#3): To characterize the Code Smells by assigning them appropriate characteristics. Method: We performed a combined search of formally published literature and grey material strictly on Code Smell and related concepts where it might never have been mentioned, along with the term “Code Smell” as a keyword. The results were analyzed and interpreted using the knowledge gathered, classified, and verified for internal consistency. Results: We identified 56 Code Smells, of which 15 are original propositions, along with an online catalog. Each smell was classified according to taxonomy, synonyms, type of problem it causes, relations, etc. In addition, we have found and listed 22 different types of Bad Smells called hierarchies and drew attention to the vague distinction between the Bad Smell concepts and Antipatterns. Conclusion: This work has the potential to raise awareness of how widespread and valuable the concept of Code Smells within the industry is and fill the gaps in the existing scientific literature. It will allow further research to be carried out consciously because access to the accumulated information resource is no longer hidden or difficult. Unified data will allow for better reproducibility of the research, and the subsequent results obtained may be more definitive