From Function to Method
Bruno Desthuilliers / John Posner
Python newcomers often have hard time understanding the "magic" behind Python's methods. And the truth is that Python's object model can be a bit peculiar when compared to most mainstream (or not-so-mainstream) object-oriented programming languages. (There are quite a few threads on c.l.py with either direct or indirect questions about what makes a Python method.)
Here's a brief -- but hopefully helpful -- overview of what exactly is a Python method, showing how Python magically inserts self or cls into the argument list of a method call.
From Function ...
The def statement always yields a function object. Always. If you don't believe it, try the following snippet:
class Foo(object):
def bar(self):
return "baaz"
print type(Foo.bar) # <type 'instancemethod'>
print type(Foo.__dict__['bar']) # <type 'function'>So, why is it that type(Foo.bar) is not the same as type(Foo.__dict__['bar'])? The answer is: attribute lookup rules and the descriptor protocol.
... To Method, via the Descriptor Protocol
The descriptor protocol specifies that during an attribute lookup, if a name resolves to a class attribute and this attribute has a __get__ method, then this __get__ method is called. The argument list to this call includes either:
- the instance and the class itself, or
None and the class itself
The return value of this call becomes the result of the attribute lookup. This mechanism is what provides support for computed attributes.
The function type implements this descriptor protocol. So when a function is an attribute of a class object and you access it as an attribute of the class itself, its __get__ method is called with None and the class as arguments. When you access it as an attribute of an instance of the class, its __get__ method is called with the instance and the class as arguments.
With the instance object (if any) and class object available, it's easy to create a method object that wraps the function object. This is itself a callable object; calling it mostly injects the instance as the first item in the argument list and returns the result of calling the wrapped function object.
A (naive) implementation of the whole thing might look like this:
class method(object):
def __init__(self, func, instance, cls):
self.im_func = func
self.im_self = instance
self.im_class = cls
def __call__(self, *args, **kw):
# XXX : all sanity checks removed for readability
if self.im_self:
args = (self.im_self,) + args
return self.im_func(*args, **kw)
class function(object):
def __get__(self, instance, cls):
return method(self, instance, cls)So, what turns a function into a method is not that the function is defined in a class statement's body (well, not directly at least). Rather, it's that the function is an attribute of the class. For what it's worth, the following code is perfectly legal:
class Foo(object):
pass
def func(obj):
print "obj is %s " % obj
Foo.method = func
f = Foo()
f.method() # all three of these
Foo.method(f) # invocations produce
func(f) # the same result
References
In the official Python documentation, the basics are at:
For an in-depth treatment, including a discussion of static methods and class methods, see:
Descriptor HowTo Guide (Python 2) - Descriptor HowTo Guide (Python 3).