SIMPLIFIED DESIGN OF TEST CASES BASED ON MODELS
Guntis Arnicāns, Vineta Arnicāne · 2011
Over the past decade a growing trend to use model-based testing approach is observed. Unfortunately, the industry rare exploits this approach. The main causes are complexity of this method, the need for a good specialists and the initial high consumption of resources. We offer pick up from the model-based testing approach the simplest ideas and techniques, and use them at least in the design of test cases. We offer additional recommendations that allow to improve the testing skills for the majority of testers. Our offer could be a useful in early stages of software development and testing, for instance, when the tester often uses an exploratory testing. One of the primary problems to be solved by the testers is a design of test cases. Tests can be designed at different conceptual levels. After the design of test cases, they must be specified by giving the exact input values and scenarios how to feed the system under test with these values. Then the system is tested and results are evaluated. The more test cases we produce, the more resources are consumed. Design of test cases is essential effort with aim to reduce this resource consumption. Model-based testing is based on three issues: 1. a formal model that describes the behavior of the software; 2. a test generation algorithm or criteria; 3. tools that support a test infrastructure (test cases generation, management, implementation, evaluation of results, etc.).