Check https on sites

README.md

top-1m.csv can be get it from alexa's website:

https://support.alexa.com/hc/articles/200449834

check-https.py

#!/usr/bin/python
import socket, ssl, threading, fnmatch

# Output log file
f_out = open('site_log.txt', 'w')

# Input, alexa top 1m sites
f_in  = open('top-1m.csv', 'r')

reader = threading.Lock()
writer = threading.Lock()
printer = threading.Lock()

class SitesChecker(threading.Thread):
  def __init__(self, threadNum):
    self.threadNum = threadNum
    threading.Thread.__init__(self)
  
  def log(self, msg):
    with printer:
      print('[Thread %d] %s' % (self.threadNum, msg))

  def write(self, host, passed):
    with writer:
      f_out.write('%s: %s\n' % (host, 'OK' if passed else 'FAIL'))

  # Check a site in a different thread
  def run(self):
    while 1:
      host = self.next()
      if not host:
        self.log('Complete.');
        break
      self.check(host)
  
  # Get next url
  def next(self):
    with reader:
      line = f_in.readline().rstrip('\n')
      if line == '':
        return False
      return line.split(',')[1]
  
  def check(self, host):
    self.log('Check cert for %s..' % host)
    passed = False
    try:
      ip = socket.getaddrinfo(host, 443)[0][4][0]
      sock = socket.socket()
      sock.settimeout(1.0)
      sock.connect((ip, 443))
      sock = ssl.wrap_socket(sock, cert_reqs=ssl.CERT_REQUIRED, ca_certs='cacerts.txt')
      cert = sock.getpeercert()
      for field in cert['subjectAltName']:
        if field[0] == 'DNS':
          if fnmatch.fnmatch(host, field[1]):
            passed = True
            break

      if not passed:
        for field in cert['subject']:
          if field[0][0] == 'commonName':
            certhost = field[0][1]
            if fnmatch.fnmatch(host, certhost):
              passed = True
              break
    except:
      self.log('Failed to connect, next ..')

    self.write(host, passed)

if __name__ == "__main__":
  lstThreads = []
  for i in range(1, 20):
    sc = SitesChecker(i)
    sc.start()
    lstThreads.append(sc)
  
  for t in lstThreads:
    t.join()

  f_out.close()