Source: 📖 Test Driven Development with Python ch19 p328
A Mock object is an object that responds to any request for an attribute or method call — it will never raise AttributeError. It can also be configured to return specific values for calls of its ubiquitous methods and attributes, and we can use the special .called and .call_args attributes to check if a method of the Mock object was called, and what arguments it was called with. It can be imported directly from unittest.mock, but it is also used implicitly in the implementation of the @patch decorator (also from unittest.mock). Consider the following shell session to illustrate the concepts:
>>> from unittest.mock imprt Mock
>>> m = Mock()
>>> m.any_attribute # random attribute call
<Mock name='mock.any_attribute' id='124750920894'>
>>> type(m.any_attribute)
<class 'unittest.mock.Mock'>
>>> m.any_method() # random method call
<Mock name='mock.any_method()' id='1249782170535'
>>> m.foo()
<Mock name='mock.foo()' id='12452710987'
>>> m.foo.called # check if foo() method has been called
True
>>> m.bar.return_value = 1 # specify return for .bar() method
>>> m.bar(42, arg='thing') # call .bar() with args
1 # returns specified value, 1
>>> m.bar.call_args # check args .bar() was called with
call(42, arg='thing')