Apache log parser for IP count, path and user agent

access.py

# coding: utf-8

import os
import re
import sys

pattern = '(\d{1,3}.\d{1,3}.\d{1,3}.\d{1,3}) - - \[.*\] "GET ([^\s]+) HTTP\/1.0" \d+ \d+ ".*" "(.*)"'
ip_list = {}

for p in sys.argv:
    if p in (__file__, os.path.basename(__file__)):
        continue

    f = open(p, 'r')

    for line in f:
        result = re.match(pattern, line)
        matches = None

        if result:
            result.groups()

        if matches:
            ip = matches[0]
            path = matches[1]
            useragent = matches[2]

            if ip in ip_list:
                ip_list[ip]['count'] += 1
            else:
                ip_list[ip] = {}
                ip_list[ip]['count'] = 1
                ip_list[ip]['path'] = path
                ip_list[ip]['useragent'] = useragent

    f.close()

sorted_ip_list = sorted(ip_list.items(), key=lambda i: i[1]['count'], reverse=True)[:30]

for k, v in sorted_ip_list:
    print('%15s - %6s - [%s]' % (k, v['count'], v['path']))