Extending a Taxonomy of Bad Code Smells with Metrics

Bad Smells define in an informal way code flaws, in order to suggest refactorings, their aim is to improve the design of the code. Current taxonomies group code smells, making use of similarity or correlation criteria between them, and leading to a manual revision of the code. By other side, it is suggested the assistance of using metrics in the detection of bad smells. Metrics can be collected automatically helping to suggest the presence of flaws. Nevertheless, current taxonomies do not link these concepts. This work tries to establish additional criteria when we want to classify bad smells. These criteria are also related to metric features. Following the current classifications, we propose a method to evaluate the suitability of the tools assisting bad code smell detection, as well as selection and implementation of metrics linked with bad code smells.