lesson1.py

# We have a log file with timestamps, named events, and a start/end indicator
# We wrote some code that is intended to calculate the average duration of
# each event, but it is producing some unexpected outputs. What's happening, and
# how can it be fixed?

# Test log file
# 22000 get_foo start
# 22050 get_foo end
# 22200 get_bar start
# 22250 get_bar end
# 22400 get_foo end
# 22500 get_bar start
# 22550 get_bar end
# 22580 get_foo start
# 22600 get_bar end
# 22610 get_foo end

def main():
    # f = open("log.txt", "r")    # open log file in read mode

    with open("log.txt", "r") as f:
        lines = f.readlines()

    event_start_times = {}      # define empty dictionaries for event starts and total times
    total_times = {}            # defines total time for each event type

    for line in lines:              # Loop over each line in log file
        words = line.rstrip().split(" ")    # Define each line in file as 'word' and strip characters to the right and split by space into a list
        timestamp = int(words[0])       # first element in 'words' and convert to int
        event = words[1]                # second element in 'words' list
        event_type = words[2]                 # third element in 'words' list

        if event_type == "start":
            event_start_times[event] = timestamp    # event_start dictionary stores the [event] as key and timestamp as value if 'start'
        elif event not in total_times:              # if the event is not already stored in the total_times dict()
            start_timestamp = event_start_times[event]  # Assign a value of the first timestamp that is encountered
            total_times[event] = ((timestamp - start_timestamp), 1)     # find the total time and count of each independent event (get_foo / get_bar)
        else:                                                       # if event IS in total_times or type == 'end'
            time_count_pair = total_times[event]    # variable that returns the total time difference of each event and number of times it has occured
            try:
                start_timestamp = event_start_times.pop(event)      # Assign a value of the starting timestamp
                time = time_count_pair[0] + (timestamp - start_timestamp)  # starting time + total time difference  (THESE VARIABLES NEED TO BE SWAPPED 'star_timestamp' and 'timestamp'_
                count = time_count_pair[1] + 1
                total_times[event] = (time, count)
            except KeyError:
                print(f"Warning! Event {event} was not started. Skipping.")

        for event in total_times:
            time, count = total_times[event]
            average = time/count
            print(f"{event} took an average of {average} milliseconds.")

main()
# NamedTuple
# One of the reasons for the problems - counting in events that are END but never started (like line 5)
# exception handling