parse top sites from squid access log

#!/usr/bin/python

import re
import argparse

def parse_args():
    parser = argparse.ArgumentParser()
    parser.add_argument("--all", help="show for all ips", action="store_true")
    parser.add_argument("logfile", help="squid access log file", action="store_true")
    parser.add_argument("iplist", help="list of ip addresses", action="store_true")
    return parser.parse_args()

def main():
    args = parse_args()
    if args.all:
        outfile = "sites.txt"
    else:
        outfile = "sites_alarmers.txt"
    sites =[]
    log = open(args.logfile, 'r')
    ips = get_alarmers(args.iplist)
    for l in log:
        url = re.split("\s+", l)[6]
        ip = re.split("\s+", l)[2]
        if not args.all:
            if ip not in ips: continue
        if re.match("^\/",url): continue
        host = re.search("^(http[s]?):\/?\/?([^\/\s]+)",url)
        if host:
            site = host.group(2)
        else:
            site = url
        if site.find('.')==-1: continue
        found=False
        for s in sites:
            if s[0]==site:
                s[1]+=1
                found = True
                break
        if not found:
            sites.append([site,1])
            print "new site: %s" %site
    sites = sorted(sites, key=lambda x: int(x[1]))
    with open(outfile,'w') as f:
        for s in sites:
            f.write("%d : %s\n" %(s[1],s[0]))
            print s[1], s[0]

def get_alarmers(iplist):
    ips = []
    interest_list = open(iplist,'r')
    for i in interest_list: ips.append(i.strip())
    return ips

if __name__ == '__main__':
    main()