Tally up records from httpd access logs by hour of day. Usage: python count_the_hours.py /path/to/access.log.

count_the_hours.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-

import itertools
import re
import sys

from datetime import datetime as dt

pattern = re.compile(r'.* \[([^]]*) \+[0-9]+\] .*')
date_fmt = '%d/%b/%Y:%H:%M:%S'

# hour of day counter
hours = dict(zip(range(24), itertools.repeat(0, 24)))

with open(sys.argv[1]) as f:
    for line in f.readlines():
        m = pattern.match(line)
        if not m:
            print 'no match: %s' % line
            continue

        date = dt.strptime(m.groups(1)[0], date_fmt)
        hours[date.hour] += 1

for hour, count in sorted(hours.items(), cmp=lambda x, y: cmp(x[1], y[1])):
    print '%s\t%s' % (hour, count)