I considered the input data to be strings, with space as separator:
s = """00:03:45.08 00:06:10.07 00:04:55.46 00:04:09.78 00:04:21.08 00:03:40.40 00:05:23.87 00:04:06.54 00:03:37.22 00:04:05.82 00:06:18.77 00:04:59.04 00:02:56.44 00:04:19.76 00:04:47.39 00:03:30.67 00:04:42.27 00:04:18.71 00:04:45.48 00:03:34.84 00:04:06.15 00:04:44,54 00:04:37.37 00:05:23.74 00:06:26,34 00:04:07.06 00:04:56.44"""
This code is for you:
import datetime
list_str = s.replace(',','.').split(" ")
list_date = map(lambda x: datetime.datetime.strptime(x, '%H:%M:%S.%f') - datetime.datetime(1900, 1, 1), list_str)
str(sum(list_date, datetime.timedelta(0)))
Output:
'2:02:50.330000'
You get the same result using this one-liner:
str(sum(map(lambda x: datetime.datetime.strptime(x, '%H:%M:%S.%f') - datetime.datetime(1900, 1, 1), s.replace(',','.').split(" ")),
datetime.timedelta(0)))
Explanation
First we need to transform the input string into a list of strings:
list_str = s.replace(',','.').split(" ")
each item in the list is converted to datetime format (the format in which python represents dates) to the format '%H:%M:%S.%f' using datetime.datetime.strptime and subtracting the library start date (1900-01-01). I apply the conversion to the entire list using the map function:
list_date = map(lambda x: datetime.datetime.strptime(x, '%H:%M:%S.%f') - datetime.datetime(1900, 1, 1), list_str)
This way I get a list of datetime.timedelta elements, which I sum using the sum function. I convert the result to string format using the str() function:
str(sum(list_date, datetime.timedelta(0)))