Model-Driven Testing Approach Based on UML Sequence Diagram
This paper is devoted to model-driven testing technique of testing model generation from UML sequence diagram to UML Testing profile. The article discusses principles of model transformation and shows practical approach to specific model generation from another one. The idea of such transformation is to generate test cases from sequence diagram of the system. Described transformation rules are focused on timing aspect verification and the goal of generated test cases is to cover successful as well as unsuccessful system behaviour. Presented example of model transformation is based on a sequence diagram XMI representation, which firstly is pre-processed and moved into data base structures, and then transformation rules applied to generate testing model from data describing sequence diagram.
Several outlines on model-driven approach for testing of embedded systems
This paper is devoted to model-driven testing approaches in the context of embedded systems. The article discusses specifics of the embedded systems as well as specific testing approaches for them. As the testing objects of embedded systems next non-functional requirements were previously selected: task scheduling, time restrictions, synchronization, asynchronisms and reliability. Timing restrictions are selected for detailed analysis in model specification. For this purpose Time domain from UML profile for Schedulability, Performance and Time is presented and discussed. Testing model as destination model is specified using standardized UML Testing profile, which provides general purpose metamodel for such model definition. The main idea of the paper is to show general principles of the model-driven testing and to represent simplified example of testing of the one specific feature of the embedded system. Discussed model-driven testing approach is based on the model transformation, where the source model describes the feature of the system and the destination model is the test case model. Transformation rules in future works will provide test case generation using model of the system.