Towards the Automation of Metamorphic Testing in Model Transformations
This addresses the correctness testing challenge for model transformations in Model-Driven Engineering, but appears incremental as it builds on existing metamorphic testing techniques.
The paper tackles the manual and error-prone oracle problem in testing model transformations by proposing an approach to generically define metamorphic relations, which can be automatically instantiated for any specific transformation.
Model transformations are the cornerstone of Model-Driven Engineering, and provide the essential mechanisms for manipulating and transforming models. Checking whether the output of a model transformation is correct is a manual and error-prone task, this is referred to as the oracle problem in the software testing literature. The correctness of the model transformation program is crucial for the proper generation of its output, so it should be tested. Metamorphic testing is a testing technique to alleviate the oracle problem consisting on exploiting the relations between different inputs and outputs of the program under test, so-called metamorphic relations. In this paper we give an insight into our approach to generically define metamorphic relations for model transformations, which can be automatically instantiated given any specific model transformation.