log_parser.py

import json
import datetime

def humanize_time(time, units):
  INTERVALS = [1, 60, 3600, 86400, 604800, 2419200, 29030400]
  NAMES = [('second', 'seconds'),
           ('minute', 'minutes'),
           ('hour', 'hours'),
           ('day', 'days'),
           ('week', 'weeks'),
           ('month', 'months'),
           ('year', 'years')]
  res = []

  unit = list(map(lambda a: a[1], NAMES)).index(units)
  # Convert to seconds
  seconds = time * INTERVALS[unit]

  for i in range(len(NAMES)-1, -1, -1):
    a = seconds // INTERVALS[i]
    if a > 0:
       res.append("%s %s" % (a, NAMES[i][1 % a]))
       seconds -= a * INTERVALS[i]

  time_str = ", ".join(res)
  return time_str

def data_from_str(dstr):
  return datetime.datetime.strptime(dstr, "%d %b %Y - %H:%M:%S")

def time_diff(log_line):
  if "date_start_time" in log_line and "date_end_time" in log_line:
    start = data_from_str(log_line["date_start_time"])
    end = data_from_str(log_line["date_end_time"])
    return (end - start).seconds
  else:
    return 0

# white_list is case insensitive
def get_stats(log_json, white_list=[]):
  stats = dict()
  for line in log_json:
    skip_stat = len(white_list) !=0 and "ActiveText" in line and all([el.lower() not in line.get("ActiveText", "").lower() for el in white_list])

    if "ActiveText" in line and not skip_stat:
      key = line.get("ActiveText", "").split("- ")[-1]

      stats[key] = stats.get(key, 0) + time_diff(line)
  return stats

def humanize_stats(stats, units= "seconds"):
  humanize_stats = stats
  for k in stats:
    humanize_stats[k] = humanize_time(humanize_stats[k], units)
  return humanize_stats

###
### How to use
###

# path to log
log_file = "activity.log"

# open log file to json
with open(log_file, encoding="utf-8") as f:
  log_json = ([json.loads(line, encoding="utf-8") for line in f])

#get stats from json
stats = get_stats(log_json)
whitelisted_stats = get_stats(log_json, white_list=['Opera','sublime'])

# print stats with time in seconds by each process
print("raw stats")
print(stats)

# pring stats for apps in whitelist
print("\nonly allowed stats")
print(whitelisted_stats)

# print stats with time in humanize_time by each process
print("\nhumanize_stats")
print(humanize_stats(stats))