Hadoopmutator: A cloud-based mutation testing framework

Abstract

Mutation testing is a software engineering methodology where code mutation is used to assess the quality of a testing technique. Mutation testing is carried out by injecting errors in the code and measuring the ability of a testing tool to detect these errors. However, it is a time-consuming process, as tests need to be run on many variants of the code, called mutants. Each mutant represents a version of the code under test, with an injected error. In this paper, we propose HadoopMutator; a cloud-based mutation testing framework that reuses the MapReduce programming model in order to speed up the generation and testing of mutants. We show, through experimentation, that we can significantly enhance the performance of automated mutation testing and provide a scalable solution that is applicable for large-scale software projects. Based on two use cases, we show that the performance can be enhanced 10 folds, on average, using our proposed framework. By treating source code as data, our work paves the way for new reuse opportunities of the novel datacentric frameworks.