Changing module variables after import

问题

SO pyton gurus! I've just found an astonishing phenomenon that I don't understand. The problem can be best shown as code:

#== kid.py ==#
import dad

def spam ():
    dad.spam()


#== dad.py ==#
import kid

x = 1
print "body", x

x = 2
def spam ():
    print "spam", x

if __name__ == '__main__':
    x = 3
    spam()
    kid.spam()
    print "main", x

I'm using Python 2.7.3. Can you guess the output of python dad.py? The answer is (I wish SO had a spoiler shading tag) body 1 body 1 spam 3 spam 2 main 3. So could you explain

1. Why is body 1 printed twice?
2. How can dad.x != kid.dad.x be?
3. If I really need to make the two modules import each other, how can I modify it to get kid.dad.x properly updated?

回答1:

1. Because loading dad.py as the __main__ module is independent of importing dad.py as the dad module.
2. See my answer to 1.
3. Import __main__ instead if you must. But in general, don't try this. Find another way to accomplish your tasks (e.g. classes).

Printing __name__ at the top of dad.py will illustrate this.

来源：https://stackoverflow.com/questions/12242417/changing-module-variables-after-import