Maybe my question wasn’t clear enough, so let me explain it further:
Based on Karate’s example documentation, eventually, we’ll use KC (payment-service.feature) and KP (payment-service-mock.feature) for testing. I have two questions:
-
Who defines KP? Is it defined by the producer or the consumer? If KP is defined by the consumer, the scenarios in KP might be significantly fewer than those in KC, leading to test failures due to missing mock scenarios. If KP is defined by the producer, it might not include the latest scenarios defined by the consumer, and thus issues may not be exposed.
-
How to address the problem of missing scenarios in KP? Should KC be layered, with different KPs defined by different consumers running different KCs? But this approach would require KC to be aligned with the consumers as well.
i wonder if there is some method to do the merge or split action.