web scraping - Python: import a module but avoid executing it? -

i making scrawler , want randomize request headers.
things goes that:
in configs.py have headers defined:

import random user_agents = ['1', '2', '3']  def get_random_user_agent():     return random.choice(user_agents)  headers = {'user-agent': get_random_user_agent()} 

in main.py have test code that:

from configs import headers in range(5):     print(headers['user-agent']) 

the result same one.
reckon variable 'headers' initialized when importing.
want randomized user agent.
there best practice this? thank you.

that's way python works. code parsed file accessed, code no in function executed immediately. why can have python script this:

import random print(random.randint(1,10)) 

and can executed file python random number.

just define headers in function:

def get_headers():     return {'user-agent': get_random_user_agent()}