A systematic literature review on Android-specific smells

Abstract

Context:Code smells are well-known concepts in Object-Oriented (OO) programs as symptoms that negatively impact software quality and cause long-term issues. However, the domain-specific smells in Android have not yet been investigated well. Android smells often refer to the misuse of mobile SDK and causes of performance, accessibility, and efficiency issues that end-users can perceive. Objective:This study aims to provide a clear overview of state-of-the-art techniques for addressing Android-specific code smells to understand existing methods and open challenges, which help the community understand the significance of Android smells and the current status of research. Methods:We conducted a Systematic Literature Review of 4,820 distinct papers published until 2021, following a consolidated methodology applied in software engineering. 35 primary studies were selected. Results:The known Android smells cannot be treated equally in the proposed approaches, as they mainly focus on detecting performance-related smells. The proposed approaches capture various features to detect smell instances using different analysis techniques in Android applications. In addition, the Android community continuously identifies new types of smells to improve apps’ quality. Conclusion:The research community still encounters several challenges. Thus, this paper outlines various directions for the necessary investigation as future work.