Python script to bulk validate a drupal site

drupal-bulk-validate.py

#!/usr/bin/env python
import urllib
import time
import httplib

base = 'www.edc.org' # NO trailing slash
lim = 200 # highest nid in your {node} table

def xhmtl_is_valid(url):
    time.sleep(.75) # seconds
    conn = httplib.HTTPConnection("validator.w3.org")
    req = u"/check?" + urllib.urlencode([('uri', url), ('verbose', 1)])
    conn.request("HEAD", req)
    res = conn.getresponse()
    return res.getheader('x-w3c-validator-status') == 'Valid'


count = 1
while (count <= lim):
    path = '/node/' + str(count)
    conn = httplib.HTTPConnection(base)
    conn.request("HEAD", path)
    res = conn.getresponse()
    if res.status == 200 or res.status == 301:
        if xhmtl_is_valid('http://' + base + path):
            print path + ' is OK'
        else:
            print 'issue with ' + path
    else:
        print 'No 200 or 301 status for ' + path
    count = count +1