Skip to main content

Fault-driven mutation testing for Python

Resource type
Thesis type
(Project) M.Sc.
Date created
2025-07-24
Authors/Contributors
Abstract
Mutation testing is an effective method for assessing test adequacy by introducing artificial faults into programs. However, existing mutation testing techniques often overlook features unique to dynamic languages like Python. In this work, we present a set of seven mutation operators inspired by common anti-patterns in Python programs. We propose a mutation testing technique that leverages a combination of static and dynamic analyses to apply these operators, with a focus on reducing the generation of equivalent mutants. We implement our approach in a tool called Flame and evaluate its effectiveness on 15 open-source Python applications. Our results show that Flame revealed previously undetected weaknesses in the test suites for each category of the introduced mutants, while maintaining an average equivalent mutant rate of approximately 2%. This suggests Flame could be effective and complementary to prior techniques in assessing Python test suite adequacy and that it can guide improvements in test quality for Python programs.
File
Extent
55 pages.
Identifier
etd24011
Copyright statement
Copyright is held by the author(s).
Permissions
This thesis may be printed or downloaded for non-commercial research and scholarly purposes.
Academic Supervisor
Thesis advisor: Ko, Steven
Thesis advisor: Alimadadi, Saba
Language
English
Member of collection
Download file Size
etd24011.pdf 523.85 KB