将函数作为命名参数传递以进行条件检查

Question

def basket(name, colour, smell, rotten=None):
    if rotten:
        if rotten():
            return "Can't eat!"

    return ("The fruit, {}, is {} in colour"
            "and smells like {}.".format(name, colour, smell))

In the example code above, the argument rotten , if provided, is checked before calling. It must be a function that returns a boolean (no arguments for simplicity). It consumes two lines of code and I'm not certain if this is the conventional style for this procedure. Another way I thought of doing it was to replace the current one with:

def basket(name, colour, smell, rotten=lambda: None):
    if rotten():  # Saved a line here
        return "Can't eat!"

    return ("The fruit, {}, is {} in colour"
            " and smells {}.".format(name, colour, smell))

The lambda method saves one line; apart from that I don't know which one has the advantage over the other. What is the preferable way of doing this? Is it found anywhere in the standard library code?

Answer 1

You can leave the function signature intact as in your first example and take advantage of boolean operations's short circuit behavior to turn it into one line like this:

def basket(name, colour, smell, rotten=None):
    if rotten and rotten():
        return "Can't eat!"

    return ("The fruit, {}, is {} in colour"
            "and smells like {}.".format(name, colour, smell))

If rotten is None , the and expression will fail without calling rotten() because of short circuit. If rotten is not None , it will proceed to call rotten() and will succeed the predicate is True .

Note: As mentioned in the comments, you can also use the function callable() and check if rotten can be called before actually calling it:

if callable(rotten) and rotten():

This will protect your function from rising an exception when rotten is not None but is not a callable object (passing rotten=True for example). If otherwise you prefer the function to complaint out loud when a non callable is passed, then just let it as it is.