How to find the sessions time from a stream of user login events in python?

Question

Given User Login Events continuously appearing in a stream processing system similar to this, where each line is 1 event (you can assume the pipeline is external to this program, this is a line-by-line function call):

1532926994 User01 LogOutSuccessful

1532926981 User02 LogInSuccessful

1532926982 User04 LogInFailed

1532926992 User01 LogInSuccessful

1532926986 User02 LogOutSuccessful

1532927003 User03 LogOutSuccessful

Implement a module with standard libraries (e.g. no Spark) that continuously processes these events and outputs user session durations as soon as a successful logout occurs in a structured format similar to:

{ "username": "User02”, "session_duration": 5 }

{ "username": "User05”, "session_duration": 10 }

I was asked this question in an interview and I could parse and extract the data but I was not able to store the user events and use it to compute the session time. Any guidance would be appreciated.

We have to write a python function which takes one line at a time and takes some action when we receive LogInSuccessful or LogOutSuccessful. I think we can ignore LogInFailed for the moment.

def get_user_session_time(stream_text):

    user_info = stream_text.split()

    if user_info[2] == 'LogInSuccessful':
       #store the data somewhere
    elif user_info[2] == 'LogOutSuccessful':
       #get the data stored in the above step 
       #compute the session time
       #print the key value pair { "username": "User02”, 
       #"session_duration": 5 }