A brief analysis of Python's metaclass

A meta class is typically used to create classes. When executing a class definition, the interpreter must know the correct meta-class of the class. The interpreter first looks for the class attribute __metaclass__, and if this attribute exists, assigns the attribute to this class as its meta-class. If this property is not defined, it looks up __metaclass__ in the parent class. If the __metaclass__ attribute is not yet found, the interpreter checks the global variable named __metaclass__ and uses it as a meta-class if it exists. Otherwise, this class is a traditional class and used types. ClassType as a meta-class of this class.

When the class definition is executed, the correct (typically default) meta-class is checked, and the meta-class (usually) passes three parameters (to the constructor): the class name, the tuple that inherits the data from the base class, and the property Dictionary of the (Class).

When was the meta-class created?

#!/usr/bin/env python     print ' 1. Metaclass declaration '  class Meta (type):      def __init__ (CLS, name, bases, ATTRD):          super (META,CLS). __init__ (NAME,BASES,ATTRD)          print ' 3. Create class%r '% (name)     print ' 2. class Foo declaration '  class Foo (object):      __metaclass__=meta      def __init__ (self):          print ' *. Init class%r '% (self.__class__.__name__)     # What asked Hovertree.comprint ' 4. Class Foo F1 instantiation '  f1=foo ()     print ' 5. Class Foo F2 instantiation '  f2=foo ()     print ' END '  output

Results:

1. Metaclass Declaration

2. Class Foo Declaration

3. Create class ' Foo '

4. Class Foo F1 instantiation

*. Init class ' Foo '

5. Class Foo F2 instantiation

*. Init class ' Foo '

END

It can be seen that in the class declaration, the __metaclass__ method is executed, and later, when the class object is defined, only the __init__ () method of the class is called, and __init__ () in Metaclass is executed only once at the class declaration.

