isolate time range from apache log file(s)

badproc.py

#!/usr/bin/python
import sys
import re
import time
from datetime import timedelta, datetime
now = datetime.now()
d_format = '%d/%b/%Y:%H:%M'
try:
    back_minutes = timedelta(minutes=int(sys.argv[2])) #no more trash var
    raw_bad_process_time = now.strftime('%d/%b/%Y:') + sys.argv[1]  #gets d/m/y
    bad_process_time = datetime.strptime(raw_bad_process_time, d_format)
    back_time = bad_process_time - back_minutes
except (TypeError, IndexError):
    sys.stderr.write("Example: cat apache.log | badproc.py 00:02 7 \n")

for line in sys.stdin:
    m = re.search('[0-9]+\/\w+\/[0-9]+:[0-9]+:[0-9]+', line)
    if m:
        the_match = m.group()
        try:
            date_in_log = datetime.strptime(the_match.split()[0], d_format)
            if back_time <= date_in_log <= bad_process_time:
                sys.stdout.write(line)
        except (ValueError, IndexError):
            pass

Usage:
badproc.py
badproc.py 19:38 3

This uses python's higher level date processing to avoid the headache when the hour rolls over/month changes/etc...