A New Approach to Quantifying Software Code Complexity

Abstract

This research presents a novel approach to quantifying software code complexity through the development of a new metric and a web-based tool using Python. Traditional complexity metrics often fall short in addressing modern software challenges such as concurrency, dynamic memory management, and exception handling. The goal is to create a paradigm-independent metric that combines conventional factors with new dimensions for a more comprehensive assessment of code complexity. The tool enables practical application of this metric with an intuitive interface for indepth code analysis. It includes features like a secure login system, an interactive dashboard for complexity evaluation, and review sections offering detailed feedback on Python code. Data is collected from industry evaluations and Python repositories on platforms like GitHub and Kaggle, ensuring relevance and robustness. The effectiveness of the new metric and tool is assessed by comparing them with traditional complexity metrics, supported by insights from professional testers. Results demonstrate improved accuracy, adaptability, and flexibility in capturing complex software behaviors often overlooked by traditional metrics. The tool consistently outperforms metrics like Cyclomatic Complexity and Weighted Composite Metric in evaluating Python code complexity, providing more nuanced insights into modern challenges such as concurrency and exception handling. The anticipated outcomes highlight the enhanced accuracy and practical relevance of the new metric for assessing complex software systems. The tool serves as a valuable resource for developers, offering deeper insights into Python code complexity and supporting informed decisions in software design, optimization, and maintenance.