DNose: Dart Test Smell Detector

Test smells, anti-patterns in test code, have emerged as key indicators of test quality, attracting significant attention in software engineering research. Automated detection of test smells is crucial for maintaining effective and readable test suites, yet existing tools often lack comprehensive language support. At the same time, sentiment analysis is gaining prominence as a technique for understanding developers’ emotional states during coding activities, providing deeper insights into software quality. Motivated by these research gaps, we present DNose, the first dedicated tool for detecting test smells in Dart-based projects. DNose identifies 14 types of test smells, correlates their introduction with developer sentiment derived from commit messages, and provides integrated insights into test code quality and developer emotions. Our approach contributes to enhancing test maintainability while illuminating emotional aspects of software development practices. Link to the video: https://youtu.be/fCYdDZA8Hkg.