Graphwalker test case selection criteria

[athar islam]

Hello

I am writing my master's thesis and I am using graphwalker and altwalker for implementation.

Can anyone please explain how test case selection is performed in graphwalker to create test cases. i mean what is the logic behind?Which criteria is followed?

Kind regards

Athar Islam

[Kristian Karl]

Hi Athar,

Cool to hear that you are using GraphWalker and altwalker in your thesis :-)

In model-based testing, the concept of test case is a bit different from traditional scripted automation.

When using GraphWalker, a test case is a combination of a:

• model
• generator
  A generator is an algorithm that choses how to select a certain path through a model.
  A generator will need a stop condition. A stop condition tells the generator when to stop.
  More on Test path generation here.
  

Using the above, a test can be designed using:

• one or more models
• where each model has one or more generators
• and each generator has one or more logical OR/AND-combined stop conditions
  

Some examples:

• random(edgecoverage(100))
  
• quickrandom(vertex_coverage(100))
  
• random ( reached_vertex(Some_vertex) or reached_edge ( Some_edge ) )
  
• random( edge_coverage(100) and time_duration(500) ) a_star(reached_vertex(v_ABC))
  

The implication of the above is that the same model(s) can be used for different kind of tests.

Below are 3 examples of designing 3 different tests, using the same model and automation code by using different generators/stop conditions.

• Smoke test example
• Functional test example
  
• Stability test example
  

Best Kristian

[Siva V]

Hi Kristian,

I have joined a team in which they use graphwalker as framework for testing. I am just starting to work on this framework. The idea of the tool is interesting.

If I have to perform data driven testing using graphwalker. I mean for example, a dropdown may accept a value from n number of values. Based on the value selected, m number of subtypes display in next page.

Usually in BDD framework, we create Scenario outline and pass combinations in a single test.

How do we go about this kind of testing in Graphwalker? I can use flows but the model would go out of proportion due to combinations of data.

If there is a way to do this, Will the report display that I have tested with these combinations?

Kindly suggest.

Thanks,

Siva