StrictMock solves the problem of having mocks that behave like the real thing. To really accomplish that we need a way of defining what a mocked method call will return. We also need a way of putting the mock in place of real objects. These are problems solved with patching.

TestSlide provides patching tools specialized in different problems. They are not only useful to configure StrictMock, but any Python object, including “real” ones, like a database connection. You can configure canned responses for specific calls, simulate network timeouts or anything you may need for your test.

Here’s a summary of the patching tools available either via testslide.TestCase or TestSlide’s DSL. Also check the comprehensive Cheat Sheet.


Changes the value of an attribute. Eg:

self.patch_attribute(math, "pi", 3)
math.pi  # => 3
mock_callable() / mock_async_callable()

Similar to patch_attribute() but designed to work with sync/async functions/methods. It creates and patches mocks for callables, which implements call arguments constraints, different call behaviors (return value, raise exception etc) and optional call assertions. Eg:

self.mock_callable("os.path", "exists")\
os.path.exists("/bin")  # => False

Allows classes to return mocks when new instances are created instead of real instances. It has the same fluid interface as mock_callable()/mock_async_callable(). Eg:

popen_mock = StrictMock(template=subprocess.Popen)
self.mock_constructor(subprocess, "Popen")\
subprocess.Popen(["/bin/true"])  # => popen_mock