#!/usr/bin/env python
import re, socket, SocketServer, urllib2, urlparse
from config import *
class ThreadingTCPServer(SocketServer.ThreadingMixIn, SocketServer.TCPServer):
"""TCPServer with ThreadingMixIn added."""
pass
class Handler(SocketServer.BaseRequestHandler):
"""Main request handler."""
def handle(self):
"""Handle a request."""
# readline is pretty convenient
f = self.request.makefile()
# read request line
reqline = line = f.readline()
split = line.rstrip('\r\n').split(' ')
http_version = len(split) > 2 and split[2] or 'HTTP/0.9'
if split[0] != 'GET':
# only GET is implemented
return self.error_page(http_version, 501, 'Not Implemented')
# parse the URL
request_url = split[1]
parsed = urlparse.urlparse(request_url)
# make a path
path = parsed.path
if parsed.query != '': path += '?' + parsed.query
if path == '': path == '/'
# get the hostname for later
host = parsed.netloc.split(':')
hostname = host[0]
# read out the headers
while line.rstrip('\r\n') != '':
line = f.readline()
try:
if hostname == 'web.archive.org':
if path[:5] != '/web/':
# launch settings
return self.handle_settings(parsed.query)
else:
# pass-through requests to web.archive.org
# required for QUICK_IMAGES
print '[>] [QI] {0}'.format('/'.join(request_url.split('/')[5:]))
conn = urllib2.urlopen(request_url)
elif GEOCITIES_FIX and hostname == 'www.geocities.com':
# apply GEOCITIES_FIX and pass it through
split = request_url.split('/')
hostname = split[2] = 'www.oocities.org'
request_url = '/'.join(split)
print '[>] {0}'.format(request_url)
conn = urllib2.urlopen(request_url)
else:
# get from Wayback
print '[>] {0}'.format(request_url)
conn = urllib2.urlopen('http://web.archive.org/web/{0}/{1}'.format(DATE, request_url))
except urllib2.HTTPError as e:
# an error has been found
print '[!] {0} {1}'.format(e.code, e.reason)
return self.error_page(http_version, e.code, e.reason)
# get content type
content_type = conn.info().getheader('Content-Type')
if not CONTENT_TYPE_ENCODING and content_type.find(';') > -1: content_type = content_type[:content_type.find(';')]
# send headers
self.request.sendall('{0} 200 OK\r\nContent-Type: {1}\r\n\r\n'.format(http_version, content_type))
# set the mode: [0]wayback [1]oocities
mode = 0
if GEOCITIES_FIX and hostname in ['www.oocities.org', 'www.oocities.com']: mode = 1
if content_type[:9] == 'text/html' in content_type: # HTML
toolbar = mode == 1 # oocities header starts without warning
after_header = False
redirect_page = False
for line in conn:
line = line.rstrip('\r\n')
if mode == 0:
if toolbar:
if line == '':
# toolbar is done - resume relaying on the next line
toolbar = False
after_header = True
continue
elif redirect_page:
# this is a really bad way to deal with Wayback's 302
# pages, but necessary with the way this proxy works
match = re.search('Impatient\\?
', line)
if match:
line = 'WaybackProxy RedirectIf you are not redirected, click here.