#!/usr/bin/python
# The script saves a Webserverlisting into a file and afterwards
# compares the online content with the file
# Tested with Apache standard configuration only
# If a change is made you will be informed once and then
# the offline file will be updated
# Change theurl, username, password and fileoffline to fit your needs.
# Also change the fromaddr and the toaddrs array.
# Afterwards start the script once with initialload set to 1.
# The local file will be initialized.
# Then reset the initialload value to 0 and start the script
# periodically (e.g. crontab)
# If this is done get a good book about some Unix stuff,
# sit down on your favourite chair and wait for notification ;)

from HTMLParser import HTMLParser
import urllib2
import smtplib

initialload = 0
theurl = 'https://yourdomain.com'
username = 'username'
password = 'password'
fileoffline = '/local/path/to/file.dat'

def mail_send(msg):
   fromaddr = "serveradmin@yourdomain.com"
   toaddrs = ("firstinform@domain.com","secondinform@domain.com")
   server = smtplib.SMTP('localhost')
   server.sendmail(fromaddr, toaddrs, msg)
   server.quit()

#dont change anything beyond this line

filearrayonline = []
filearrayoffline = []
months = 'Jan','Feb','Mar','Apr','May','Jun','Jul', \
'Aug','Sep','Oct','Nov','Dec'
assozarray = []

class AnchorParser(HTMLParser):
def __init__(self, url):
HTMLParser.__init__(self)

def handle_starttag(self, tag, attrs):
if tag =='a':
for key, value in attrs:
if key == 'href':
if '?' not in value:
filearrayonline.append(value.strip())
def handle_data(self,datain):
for month in months:
if month in datain:
filearrayonline.append(datain.strip())


passman = urllib2.HTTPPasswordMgrWithDefaultRealm()
passman.add_password(None, theurl, username, password)
authhandler = urllib2.HTTPBasicAuthHandler(passman)
opener = urllib2.build_opener(authhandler)
urllib2.install_opener(opener)

parser = AnchorParser(HTMLParser)
data = urllib2.urlopen(theurl).read()
parser.feed(data)
print filearrayonline

if initialload == 1:
f1 = file(fileoffline,"w")
for fileon in filearrayonline:
f1.write(fileon + "\n")
f1.close()

f1 = file(fileoffline,"r")

for line in f1:
filearrayoffline.append(line.rstrip("\r\n"))

f1.close()

for fileon in filearrayonline:
if fileon not in filearrayoffline:
mail_send("New file in maths download")
print "New files online"
f1 = file(fileoffline,"w")
for fileon1 in filearrayonline:
f1.write(fileon1 + "\n")
f1.close()
else:
print "No new files"