Subclassing function types in python

Disclaimer: The following is based on cpython.

>>> class A(object):
...     @classmethod
...     def clsmeth(cls): pass
...     @staticmethod
...     def stemeth(): pass
...     def method(self): pass
...
>>> def func(): pass
...
>>> labd = lambda x: x

And let's define a way of finding out the corresponding type in the types module

>>> import types
>>> typenames = lambda inst: [n for n,t in types.__dict__.items() if t is type(inst)]

typenames is now a function that will evaluate to a list of names of the types of the instance we give it. Let's check that it works

>>> typenames(1)
['IntType']
>>> typenames('1')
['StringType']
>>> typenames(True)
['BooleanType']

Let's take a look at the types of different built in callable types:

>>> typenames(func)
['LambdaType', 'FunctionType']
>>> typenames(labd)
['LambdaType', 'FunctionType']
>>> # Ordinary methods
>>> typenames(A.method)
['UnboundMethodType', 'MethodType']
>>> typenames(A().method)
['UnboundMethodType', 'MethodType']
>>> # Static methods
>>> typenames(A().stemeth)
['LambdaType', 'FunctionType']
>>> typenames(A.stemeth)
>>>
['LambdaType', 'FunctionType']
>>> typenames(A.clsmeth)
['UnboundMethodType', 'MethodType']
>>> typenames(A().clsmeth)
['UnboundMethodType', 'MethodType']

So there are (at least) four built in callable types.

FunctionType
LambdaType
UnboundMethodType
MethodType

Calling them we can see the various:

>>> A.stemeth()
>>> A().stemeth()
>>> A.clsmeth()
(<class '__main__.A'>,)
>>> A().clsmeth()
(<class '__main__.A'>,)
>>> A.method()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: unbound method method() must be called with A instance as first argument (got nothing instead)
>>> A().method()
>>>

Note: types.py does pretty much the same thing to get hold of the types themselves but for educational reasons let's pretend that what we just did is not stupid.

Categories

Archives