On the correlation between Architectural Smells and Static Analysis Warnings

2024-06-25

arXiv.org

AI Summary

Architectural smell (AS) remediation prioritization lacks empirical guidance, hindering efficient architecture governance. Method: We conduct the first large-scale empirical study correlating static analysis warnings (SAWs) with ASs across 103 Java projects (72M LOC), using Checkstyle, FindBugs, PMD, and SonarQube. We quantify their association and identify SAWs exhibiting low co-occurrence with ASs—termed “healthy carriers.” We then propose a novel two-dimensional AS repair prioritization model integrating SAW severity and AS propensity. Contribution/Results: We find a moderate correlation between SAWs and ASs (Spearman’s ρ ≈ 0.45) and identify 33.79% of SAWs as healthy carriers. Our model achieves ranking quality comparable to AS-severity–based prioritization while improving quality assurance efficiency by ~30%. The approach provides an interpretable, data-driven foundation for precise and efficient architectural remediation.

Technology Category

• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Security and Privacy: Large-scale security measurements

Abstract

Background. Software quality assurance is essential during software development and maintenance. Static Analysis Tools (SAT) are widely used for assessing code quality. Architectural smells are becoming more daunting to address and evaluate among quality issues. Objective. We aim to understand the relationships between static analysis warnings (SAW) and architectural smells (AS) to guide developers/maintainers in focusing their efforts on SAWs more prone to co-occurring with AS. Method. We performed an empirical study on 103 Java projects totaling 72 million LOC belonging to projects from a vast set of domains, and 785 SAW detected by four SAT, Checkstyle, Findbugs, PMD, SonarQube, and 4 architectural smells detected by ARCAN tool. We analyzed how SAWs influence AS presence. Finally, we proposed an AS remediation effort prioritization based on SAW severity and SAW proneness to specific ASs. Results. Our study reveals a moderate correlation between SAWs and ASs. Different combinations of SATs and SAWs significantly affect AS occurrence, with certain SAWs more likely to co-occur with specific ASs. Conversely, 33.79% of SAWs act as"healthy carriers", not associated with any ASs. Conclusion. Practitioners can ignore about a third of SAWs and focus on those most likely to be associated with ASs. Prioritizing AS remediation based on SAW severity or SAW proneness to specific ASs results in effective rankings like those based on AS severity.

Problem

• Investigating correlation between static analysis warnings and architectural smells
• Identifying which warnings co-occur with specific architectural issues
• Prioritizing remediation efforts based on warning severity and smell proneness

Innovation

• Empirical study correlating static warnings architectural smells
• Prioritization method based severity proneness specific smells
• Identified healthy carrier warnings unrelated architectural issues