Test Data Generation for Model Transformations Combining Partition and Constraint Analysis

摘要: Model-Driven Engineering (MDE) is a software engineering paradigm where models play key role. In MDE-based development process, are successively transformed into other and eventually the final source code by means of chain model transformations. Since writing transformations an error-prone task, mechanisms to ensure their reliability greatly needed. One way achieving this testing. A challenging aspect when testing generation adequate input test data. Most existing approaches generate data following black-box approach based on some sort partition analysis that exploits structural features metamodel transformation. However, these analyses pay no attention OCL invariants or do it very superficially. paper, we propose mechanism systematically analyzes constraints in order fine-tune therefore, Our can be used isolation, combined with white-box approaches.