Python 模拟文件中导入的类的初始化

我有一个文件副本_x_to _y.py，如下所示：

from abcd import F

def function_to_be_tested():
     F()

from xyz import XY

class F():
   def __init__(self, arg1):
       self.xy = XY(arg1)

在abcd.py文件中，我有如下内容：

from abcd import F

def function_to_be_tested():
     F()

from xyz import XY

class F():
   def __init__(self, arg1):
       self.xy = XY(arg1)

我想在我的测试用例中模拟XY的init。 我试着用以下方法模拟F的init：

def mock_func(*args, **kwargs):
    pass

@patch('path/to/copy_x_to_y.F.__init__', mock_func)
def test():
    assert function_to_be_tested() is None

但它总是调用XY的init，导致初始化调用时出错

---

使用arg1连接S3。如何测试这种结构？

为什么要模拟XY的初始？是否要返回XY的特定对象，是否要检查XY.\uuuu init\uuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuu

问题的一个可能解决方案是模拟整个类，但让它返回一个普通对象。下面是一个例子：

>>> from unittest.mock import patch
>>> class MyClass:
...   def __init__(self, val):
...     self._val = val
...   def foo(self):
...     print(self._val)
... 
>>> a = MyClass(1)
>>> a.foo()
1
>>> patcher = patch('__main__.MyClass', return_value=a)
>>> mock_class = patcher.start()
>>> b = MyClass(2)  # This will return a.
>>> b.foo()
1
>>> mock_class.call_args_list
[call(2)]
>>> patcher.stop()

在你的情况下，这将是：

from xyz import XY
from path/to/copy_x_to_y import function_to_be_tested
def test():
  arg1 = ...
  a = XY(arg1)  # Has to be called before the patch to get a "normal" object.
  with patch('xyz.XY', return_value=a) as mock_xy:
    # Run funcion to be tested here and check results.
    function_to_be_tested()
    assert ...

一些旁注： 不过，如果您真的需要这样做的话，可以直接模拟init。 如果使用补丁装饰器，则必须为装饰函数提供一个额外的参数，该参数是类或对象的模拟参数。 另外，补丁通常是独占的？用于修补类，而patch.object用于修补类或模块内的成员。

---

想要模拟XY的初始值的原因是什么？是否要返回XY的特定对象，是否要检查XY.\uuuu init\uuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuu

问题的一个可能解决方案是模拟整个类，但让它返回一个普通对象。下面是一个例子：

>>> from unittest.mock import patch
>>> class MyClass:
...   def __init__(self, val):
...     self._val = val
...   def foo(self):
...     print(self._val)
... 
>>> a = MyClass(1)
>>> a.foo()
1
>>> patcher = patch('__main__.MyClass', return_value=a)
>>> mock_class = patcher.start()
>>> b = MyClass(2)  # This will return a.
>>> b.foo()
1
>>> mock_class.call_args_list
[call(2)]
>>> patcher.stop()

在你的情况下，这将是：

from xyz import XY
from path/to/copy_x_to_y import function_to_be_tested
def test():
  arg1 = ...
  a = XY(arg1)  # Has to be called before the patch to get a "normal" object.
  with patch('xyz.XY', return_value=a) as mock_xy:
    # Run funcion to be tested here and check results.
    function_to_be_tested()
    assert ...

一些旁注： 不过，如果您真的需要这样做的话，可以直接模拟init。 如果使用补丁装饰器，则必须为装饰函数提供一个额外的参数，该参数是类或对象的模拟参数。 另外，补丁通常是独占的？用于修补类，而patch.object用于修补类或模块内的成员。

---

我来这里是想找到一个答案来检查XY.\uuuu init\uuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuu；我想旁注1会启用这个功能，用mock_init.assert_调用，_with…，我来这里是想找到一个答案来检查XY.u init_uu是否用特定参数调用；我猜侧注1会启用它，使用mock_init.assert_调用_with。。。。