Mocking objects in unit tests is pretty straightforward as every object used in a test case is usually created in a scope of one test class. In functional tests it's a bit harder since we either don't have full control over objects being created or it's simply too laborious to mock half the framework. [...] We're getting the service from a container [in the example] and calling a method which should send a lead. The problem is we don't want to actually call an API while executing Behat scenarios.
Rather than hitting up the API for each test, he opts to create mock objects and results with the tools Mockery has to offer. He gives code for a "is API available" method that either returns a valid container or a mocked object, depending on how it was called.