Design and classification of mutation operators for OCL specification

Abstract

Mutation testing is a fault-based testing technique that validates the effectiveness of a test suite by generating simulated defects in the system under test. In mutation testing, from a program or specification p, a faulty version p′ is generated by making a single simple change to p. The mutation operators are the transition rules that define how to perform these changes. Mutation operators are the essential component to guarantee the accuracy of mutation testing. Most mutation testing techniques in the literature have been researched for traditional programming languages. Since Model-Driven Engineering (MDE) has become popular in recent years and Object Constraint Language (OCL) is a standard MDE language, there has been a rapid growth in the number of system specifications expressed in OCL. However, to the best of our knowledge, there is a lack of systematic design and classification of mutation operators which apply to OCL specification. In this paper, we propose the full set of mutation operators for the OCL standard library. These operators will support the mutation testing process for system specifications expressed in OCL and could be included in future versions of the OCL library. However, these operators have different impacts and results for mutation testing. Furthermore, different OCL participators may not need to apply all these operators based on the specific testing purposes. We therefore also classify these operators into different groups in this work. The classification of OCL mutation operators will benefit the mutation testing by allowing particular groups of operators to be involved in the testing process.