The built-in library functools contains several utility functions for working with functions.
See https://docs.python.org/2/library/functools.html
One of the most important is partial, which we will see later.
Here we look at wraps, which makes decorators work in a slightly more convenient way.
First let's try and create a decorator which might be useful. This one takes a function and turns it
into a function which can deal with None being passed in.
>>> def protect_from_none(f):
... def temp(x):
... if x is None:
... return None
... return f(x)
... return temp
We take a function which we could apply it to:
>>> def add_five(n):
... """ Adds five to things. """
... return n + 5
...
>>> add_five(10)
15
>>> add_five(None)
Traceback (most recent call last):
...
TypeError: unsupported operand type(s) for +: 'NoneType' and 'int'
Let's check some of the details of how this function is represented:
>>> add_five # doctest: +ELLIPSIS
<function add_five at 0x...>
>>> add_five.__doc__
' Adds five to things. '
Now let's define the function again, this time using the decorator.
>>> @protect_from_none
... def add_five(n):
... """ Adds five to things. """
... return n + 5
...
>>> add_five(2)
7
>>> add_five(None)
It works, but we have messed up some of the metadata of the function a bit.
>>> add_five # doctest: +ELLIPSIS
<function temp at 0x...>
>>> add_five.__doc__
To keep this metadata in the decorated function, we use wraps in the decorator.
>>> from functools import wraps
>>> def protect_from_none(f):
... @wraps(f)
... def temp(x):
... if x is None:
... return None
... return f(x)
... return temp
When we defined the function temp, we specified that it was supposed to be a replacement for f.
>>> @protect_from_none
... def add_five(n):
... """ Adds five to things. """
... return n + 5
...
>>> add_five # doctest: +ELLIPSIS
<function add_five at 0x...>
>>> add_five.__doc__
' Adds five to things. '