代码之家 › 专栏 › 技术社区 › Lone Learner

mock上的相同调用链是否总是返回完全相同的mock对象?

monkeypatching mocking unit-testing python

Lone Learner · 技术社区 · 5 年前

代码:

from unittest.mock import Mock

mock = Mock()

print('mock.f():', id(mock.f()))
print('mock.f().g().h():', id(mock.f().g().h()))
print('mock():', id(mock()))
print('mock().f():', id(mock().f()))
print()

print('mock.f():', id(mock.f()))
print('mock.f().g().h():', id(mock.f().g().h()))
print('mock():', id(mock()))
print('mock().f():', id(mock().f()))
print()

print('mock.f(1):', id(mock.f(1)))
print('mock.f(2).g(3).h(4):', id(mock.f(2).g(3).h(4)))
print('mock(5):', id(mock(5)))
print('mock(6).f(7):', id(mock(6).f(7)))
print()

mock.f(): 4483288208
mock.f().g().h(): 4483354192
mock(): 4483368976
mock().f(): 4483708880

mock.f(): 4483288208
mock.f().g().h(): 4483354192
mock(): 4483368976
mock().f(): 4483708880

mock.f(1): 4483288208
mock.f(2).g(3).h(4): 4483354192
mock(5): 4483368976
mock(6).f(7): 4483708880

观察:

输出显示,对mock的指定链式函数调用总是在程序的生存期内返回相同的对象,而不管我们调用了多少次。

例如,第一个调用 mock.f().g().h() ,第二个呼叫 模拟.f().g().h() ,甚至第三个调用也有不同的参数 mock.f(2).g(3).h(4)

问题:

我们能相信这种行为吗?在一个程序的生命周期内, 会返回完全相同的模拟对象吗?
是否保证即使同一个调用链具有不同的参数,例如。, 模拟f(2.g(3.h(4) 模拟.f().g().h() ?

背景:

from urllib import request
from unittest.mock import Mock, patch

with patch('urllib.request.urlopen') as mock_urlopen:
    mock_urlopen.return_value = Mock()
    mock_urlopen().getcode.return_value = 200
    assert request.urlopen('').getcode() == 200

我可以写这样的代码:

from urllib import request
from unittest.mock import Mock, patch

with patch('urllib.request.urlopen') as mock_urlopen:
    mock_urlopen().getcode.return_value = 200
    assert request.urlopen('').getcode() == 200

上面的例子太简单了,只是为了演示。我想保留一些独立的例子。但是如果我可以依赖这个特性,当函数调用链很长时,它将变得非常方便。这就是为什么我要寻找某种参考资料或文件,表明我可以依赖这种行为。

0 回复 | 直到 5 年前

Tarun Lalwani 5 年前

如果你看一下文件 lib/python3.7/unittest/mock.py

def __getattr__(self, name):
    if name in {'_mock_methods', '_mock_unsafe'}:
        raise AttributeError(name)
    elif self._mock_methods is not None:
        if name not in self._mock_methods or name in _all_magics:
            raise AttributeError("Mock object has no attribute %r" % name)
    elif _is_magic(name):
        raise AttributeError(name)
    if not self._mock_unsafe:
        if name.startswith(('assert', 'assret')):
            raise AttributeError(name)

    result = self._mock_children.get(name)
    if result is _deleted:
        raise AttributeError(name)
    elif result is None:
        wraps = None
        if self._mock_wraps is not None:
            # XXXX should we get the attribute without triggering code
            # execution?
            wraps = getattr(self._mock_wraps, name)

        result = self._get_child_mock(
            parent=self, name=name, wraps=wraps, _new_name=name,
            _new_parent=self
        )
        self._mock_children[name]  = result

    elif isinstance(result, _SpecState):
        result = create_autospec(
            result.spec, result.spec_set, result.instance,
            result.parent, result.name
        )
        self._mock_children[name]  = result

    return result

如您所见,对象被缓存在 _mock_children 因此每个调用都会返回对象。但数据会更新。通过运行下面的代码可以看到

from unittest.mock import Mock

mock = Mock()

mock.a(10)
mock.a.assert_called_with(10)
mock.a(2)
mock.a.assert_called_with(10)

结果呢

Traceback (most recent call last):
  File ".../workbench.py", line 8, in <module>
    mock.a.assert_called_with(10)
  File ....lib/python3.7/unittest/mock.py", line 834, in assert_called_with
    raise AssertionError(_error_message()) from cause
AssertionError: Expected call: a(10)
Actual call: a(2)