Although I've been using python for a while, and a common pattern you see is these two statements:
def main():
# Fancy code here
if __name__ == "__main__":
main()
My question is, why wouldn't you use this pattern instead?
if __name__ == "__main__":
# Fancy code here
Is this just in case you want to import main from another place? Or is there some other reason you might want to do this?
Another reason is to avoid populating the global scope with variables.
Consider this example:
def add_em(arg1, arg2):
return a + b
if __name__ == "__main__":
a = 2
b = 4
print(add_em(a, b))
Here, the add_em() function clearly has a bug: it should return arg1 + arg2, but since the a and b variables belong to the global scope, this bug will likely go undetected until add_em() is called from another context. Running the script gives no error:
$ python myscript.py
6
Using a main() function would likely detect this earlier:
def add_em(arg1, arg2):
return a + b
def main():
a = 2
b = 4
print(add_em(a, b))
if __name__ == "__main__":
main()
Now the error is detected right away:
$ python myscript.py
Traceback (most recent call last):
...
NameError: name 'a' is not defined
A function is reusable, you can call it from other code. Code under the if __name__ guard can't so be invoked.
It makes testing and encapsulation easier. All you need is
import module
module.main()
module.main(), making the code reusable.