time.py

import dateutil.parser
import re

date_re = r'(-?(?:[1-9][0-9]*)?[0-9]{4})-(1[0-2]|0[1-9])-(3[01]|0[1-9]|[12][0-9])T(2[0-3]|[01][0-9]):([0-5][0-9]):([0-5][0-9])(\.[0-9]+)?(Z|[+-](?:2[0-3]|[01][0-9]):[0-5][0-9])?'
full_re = f'({date_re}) (\[(\d*)/\d*\]) (.*)'
match = re.compile(full_re).match

def process_log(lines):
    last_date = None
    total_time = 0
    times = {}

    i = 1
    for line in lines:
        # remove extra time precision, that causes ISO date parsing to fail
        line = line[:26] + line[27:]

        m = match(line)
        if not m:
            continue

        date = m.group(1)
        date = dateutil.parser.isoparse(date)
        step = int(m.group(11))
        msg = m.group(12)

        assert(step == i)
        i = i + 1

        if last_date is not None:
            time = (date - last_date).total_seconds()
            times[msg] = time
            total_time = total_time + time
        last_date = date
    return times


f = open('gcc.txt', 'r')
lines = f.readlines()
times_gcc = process_log(lines)

f = open('clang.txt', 'r')
lines = f.readlines()
times_clang = process_log(lines)

for msg, t_clang in times_clang.items():
    t_gcc = times_gcc[msg]
    t_ratio = t_clang / t_gcc
    #print(f'{t_clang} {t_gcc} {msg}')
    print(f'{t_ratio} {t_gcc} {t_clang} {msg}')

#print(total_time)