Functions as First-Class Objects in Python

One thing that might be helpful to point out to readers here is that if your application passes or uses an function as an object, make sure not to actually call the function (i.e. include the ()). Something like:

# Might not work:
def call_func(some_func()):
    return some_func("123")

# Probably what is needed:
def call_func(some_func):
    return some_func("123")

This tripped me up in a QT application and was not fun to debug…

What does return f do?

Hi @bhaskarj4. The f at the beginning of the quotes is how you denote an f-string in Python. This is one way in which you can format a string. To learn more about f-strings you can use this course: realpython.com/courses/formatting-python-strings/

What is a first-class object in Python? Is there a tutorial about it? I have watched OOP in Python and it wasn’t covered.

@shuhuali1010 The phrase first-class object or first-class citizen is informal jargon, which isn’t specific to Python. Essentially, it’s a fancy way of saying that functions behave the same as regular values like strings or numbers.

For example, you can assign a function to a variable and use that variable to call it:

>>> def add(x, y):
...     return x + y
... 
>>> operation = add
>>> operation(2, 3)
5

You can also pass a function to another function as a parameter:

>>> veggies = ["eggplant", "corn", "tomato"]
>>> sorted(veggies, key=len)
['corn', 'tomato', 'eggplant']

Finally, you can return a function from a so-called factory function:

>>> def make_adder(amount):
...     def add(x):
...         return x + amount
...     return add
... 
>>> add3 = make_adder(3)
>>> add3(2)
5

It’s a common way of defining decorators in Python.

Amazing! Now I can understand functions!!! Thank you Chris! :)

As @Paul M, mentioned. What’s the rule of thumb when not to include the () in a function. When a function is assigned to a variable, or used as an argument in another function or used in an iterable? Or more simply only use the () when called?

@sndselecta Your last sentence sums it up, i.e., you’d only append the parens when calling the function. Otherwise, you can use the function name as a reference to call later from a different place in your code.

This issue, when to and when not to use parens with function names, was discussed extensively in the Dec 20th office hours. See discussion in Slack under the office hours channel.

I’m a bit confused. Where is the association between the greeter_func() as well as its argument and the other two functions? I would have expected an undefined error, since greeter_func() has never been defined globally or locally.

@sndselecta There’s no formal association becuase it’s all happening at runtime. greeter_func is an arbitrarily chosen name for the greet_bob() function’s only argument. It allows you to access a callable object that was passed into the greet_bob() function when someone called that function. In this case, greet_bob() proxies the call to whatever function gets passed as an argument and immediately returns with the value to the caller.

Perhaps this code snippet will help you see the whole picture better:

>>> def greet_bob(greeter_func):
...     return greeter_func("Bob")
...

>>> def say_hello(name):
...     return f"Hello, {name}"
...

>>> def say_bye(name):
...     return f"Bye, {name}"
...

>>> greet_bob(say_hello)
'Hello, Bob'

>>> greet_bob(say_bye)
'Bye, Bob'

You start by defining a higher-level function, greet_bob(), which accepts another function as an argument. Then, you proceed to define two other functions that return a relevant message. Finally, you call your higher-level function with the two functions to print different messages to Bob.

Oh I see it now, association → the greeter_func doesn’t have to be defined as it is defined as an argument.