I have a function call that returns an object:
r = Foo(x,y)
where r
has a rich set of nested properties. For example, I can access r.prop_a.prop_b.prop_c
. I would like to mock Foo
, such that a specific leaf property of r
is modified, i.e. such that r.prop_a.prop_b.prop_c
returns a value under my control:
>> r = Foo(x,y)
>> r.prop_a.prop_b.prop_c
'fish'
>> # some mock magic patching of Foo is taking place here
>> r = Foo(x,y)
>> r.prop_a.prop_b.prop_c
'my_fish'
I do not care about intermediate properties much.
Is there an elegant way to mock nested properties with mock?
Replace the mock object's attribute call as you would expect:
>> r1 = r_original(x, y)
>> r1.prop_a.prop_b.prop_c
'fish'
>> returner = mock.MagicMock()
>> returner.prop_a.prop_b.prop_c = 'fish'
>> r_mocked = mock.MagicMock(spec_set=r_original, return_value=returner)
>> r2 = r_mocked(x, y)
>> r2.prop_a.prop_b
MagicMock name='returner.prop_a.prop_b' id='87412560'>
>> r2.prop_a.prop_b.prop_c
'fish'
This allows you the full power of mocking while defining a specific value.
If you want to expose the original properties elsewhere, you can define a wrapper class:
class OverrideAttributePath(object):
"""A proxy class where we override a specific attribute path with the
value given. For any other attribute path, we just return
attributes on the wrapped object.
"""
def __init__(self, thing, path, value):
self._thing = thing
self._path = path
self._value = value
def __dir__(self):
return dir(self._thing)
def __len__(self):
return len(self._thing)
def __getitem__(self, index):
if self._path == [index]:
return self._value
elif self._path[0] == index:
return OverrideAttributePath(
self._thing[index], self._path[1:], self._value)
else:
return self._thing[index]
def __getattr__(self, key):
if self._path == [key]:
return self._value
elif self._path[0] == key:
return OverrideAttributePath(
getattr(self._thing, key), self._path[1:], self._value)
else:
return getattr(self._thing, key)
Usage is then as follows:
>>> r = Foo(x,y)
>>> r2 = OverrideAttributePath(r, ['prop_a', 'prop_b', 'prop_c'], 'my_fish')
>>> r2.prop_a.prop_b.prop_c
'my_fish'
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With