Zope接口是如何实现的？
我试图了解 Zope 界面是如何工作的。我知道 Interface 只是 InterfaceClass 的一个实例，而 InterfaceClass 只是一个普通的类。但是如果Interface只是一个类实例，为什么它可以作为基类来继承呢？
例如 IFoo类（接口）： pass
你能给我一些见解吗？谢谢。
I tried to understand how Zope interface work. I know Interface is just an instance of InterfaceClass which is just an ordinary Class. But if Interface is just a class instance, why it can be used as a base class to be inherited from?
e.g.
Class IFoo(Interface):
pass
Could you give me some insights? Thank you.
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（1）
Python 本质上是灵活的，任何对象都可以是基类，只要它看起来像基类即可。与 Python 的情况一样，这意味着实现一些预期在 Python 类上找到的属性。
Interface 类（或其基础规范和元素）设置了多个。查找以双下划线 (
__) 开头的任何变量集以获取理解：__module__：包含 python 路径模块的字符串。__name__：定义类的名称。__bases__：该类的基类。__doc__：（可选）类的文档字符串。另外，作为基类使用时，会调用InterfaceClass
__init__方法； Python 基本上将基类视为元类，每当我们在类定义中使用它时，都会创建基类的类（元类）的新实例。这意味着__init__方法将传递新的__name__和__bases__值，以及所有新的类属性作为关键字参数（包括__module__和可选的__doc__）。这全部记录在 Python 数据模型的 标准类型层次结构部分中文档（查找有关特殊属性的“类”段落），以及 同一个文档，在自定义类创建部分（具有
__class__属性的基类被视为类型）。因此，任何至少定义
__module__、__name__和__bases__属性的Python实例，以及合适的__init__方法将作为其他类的基类。 Python 会完成剩下的工作。Python is inherently flexible, and any object can be a base class as long as it looks like a base class. As is always the case with Python, that means implementing some attributes that are expected to be found on a Python classes.
The Interface class (or it's bases Specification and Element) sets several. Look for any variables set starting with a double underscore (
__) to gain an understanding:__module__: A string containing the python path module.__name__: The name under which the class was defined.__bases__: The base classes of this class.__doc__: (optional) The docstring of the class.In addition, the InterfaceClass
__init__method will be called when used as a base class; Python basically treats base classes as metaclasses, and a new instance of the base class's class (metaclass) will be created whenever we use it in a class definition. This means that the__init__method will be passed the new__name__and__bases__values, as well as all the new class attributes as keyword arguments (including__module__and an optional__doc__).This is all documented in the Standard type hierarchy section of the Python Data Model document (look for the 'classes' paragraph on special attributes), and in the same document, in the Customizing class creation section (base classes with a
__class__attribute are deemed a type).So, any python instance that defines at least
__module__,__name__and__bases__attributes, and a suitable__init__method will work as a base class for other classes. Python does the rest.