Sometimes there’s an important difference between None and []. That’s by far not the most common use, but it does exist (e.g. None could mean “user didn’t supply any data” and [] could mean “user explicitly supplied empty data”).

If the distinction matters, make it explicit:

if foo is None:
    raise ValueError("foo must be defined for this operation")
if not foo:
    return None

for bar in foo:
    ...

return some_other_value

This way you’re explicit about what constitutes an error vs no data, and the caller can differentiate as well. In most cases though, you don’t need that first check, if not foo can probably just return None or use some default value or whatever, and whether it’s None or [] doesn’t matter.

if len(foo) == 0: is bad for a few reasons:

• TypeError will be raised if it’s None, which is probably unexpected
• it’s slower
• it’s longer

If you don’t care about the distinction, handle both the same way. If you do care, handle them separately.