10 likes
·
512 reads
4 comments
Great Article! Allow me light degression - if the context is about testing different users I would prefer fixtures though. It would increase the number of tests in the project, but will keep the config file clean with simply one project.
But, maybe if you are lazy and just want to maintain only one test, the concept itself has the right to exist :)
Eugene, thank you for your feedback! Your points are indeed valid, in this case we simply can use fixtures and keep one project. However, parameterization with different users by project might be reasonable when you'd like to run slightly different set of test cases against each user for example. Then it might be easier to configure them in a project scope. Also, this approach allows us to see report (built-in or Allure report) divided by project (by target user), if that's what we'd like to see at the end. The way we handle parameterization depends on what we would like to solve and possible constraints present.
Good point. I would agree, that the approach or high-level decision of a framework is context-dependent.
My approach to differentiate test cases in reports using different test data would be to add test group annotations in the test description. Thus all tests would be grouped by the assigned group.
Test groups could be stored in the enum.Eleonora Belova
Good post