I am trying to write tests with loopback-testing.
I am a bit confused, and there's barely any documentation at all.
I'd like to test a model for which only a user with role "admin" has WRITE rights.
Now, if I do:
lt.describe.whenCalledByUserWithRole(test_config.adminUserCredentials, test_config.adminRole, ....)
loopback
will actually create a user with test_config.adminUserCredentials
prior to login! Why is it doing that?
The correct behavior, IMHO, should be:
- Create the user in test setup
- Create the role in test setup
- Associate the role to the user
- When running the test, only check that the user has rights on the requested operation
But it looks to be quite tricky to do with loopback-testing. If I create a user in setup, the test will crash because whenCalledByUserWithRole
will in the process try to create the user again, which loopback will deny saying the user already exists. If I don't create a user and call whenCalledByUserWithRole
, this user won't be associated to the "admin" role for some reason (even if the name suggests so), and the test fails.
How do I do this correctly?
Aucun commentaire:
Enregistrer un commentaire