errors-by-date.py

logerrors.py

import os
import datetime
from collections import Counter
import json
import argparse


parser = argparse.ArgumentParser(
    description='Count Tomcat errors and group by date.'
)
parser.add_argument(
    'path',
    metavar='path',
    type=str,
    help='the path to the log files'
)
args = parser.parse_args()
dates = []

for root,dirs,files in os.walk(args.path):
    print len(files), "files in", args.path
    for file in files:
       f=open(os.path.join(root,file), 'r')
       for line in f.readlines():
            if "ERROR" in line:
                if "[" not in line.split(',')[0]:
                    date = line.split(',')[0]
                else:
                    date = line.split(',')[0].split('[')[1]
                day = datetime.datetime.strptime(date, '%Y-%m-%d %H:%M:%S').date()
                dates.append(day.strftime('%Y-%m-%d'))
       f.close()

date_counter = Counter(dates)
print json.dumps(date_counter, indent=2)
print reduce(lambda x, y: x + y, date_counter.values(), 0) , "errors"