I was introduced to a 5-layer-model for GUI test automation frameworks, that was considered common knowledge. But I would be interested in a source describing this model. The layers are the following:
Has anyone seen this kind of architecture before and can tell me a paper or book describing it? "Scientific" sources are preferred.
The best place to start is over at http://www.thebraidytester.com/ where Michael Hunter describes the stack he built for testing Microsoft Expression in number of articles, papers and blog posts.
I have used this approach as the inspiration for my own watin based stack which is available at http://testingstax.codeplex.com
Essentially the framework breaks down into the following components.
The key design principle is that all the concerns have been separated. The test, from the business model, the physical implementation from the test data and verification engine.
Michael Hunters automation stack http://www.thebraidytester.com/stack.html is a good reference.
I've also blogged about a similar approach here in a series of blogpost. http://elusivebug.blogspot.com/2009/05/test-automation-approach-object.html
In a layered architecture, the test automation code is divided into three layers.
1.Test cases-Focusing on the test logic of the application.
2.The domain layer-Modeling the system under test in domain terms, encapsulating HTTP requests, browser control, result parsing logic, and providing an interface for the test cases layer.
3.The system under test-Which layer 2 will operate directly on.
