Simple proxy server in Python with HTML text substitution.

main.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-
import lxml.html
from lxml.etree import tostring as etree_tostring
import re
import SocketServer
import SimpleHTTPServer
import shutil
import StringIO
import urllib2

HOST = '127.0.0.1'
PORT = 8232


class Proxy(SimpleHTTPServer.SimpleHTTPRequestHandler):

    def send_error(self, code, message=None):
        self.send_response(code)
        self.end_headers()

    def send_head(self):
        print self.requestline
        url = 'http://habrahabr.ru' + self.path
        opener = urllib2.build_opener()
        self.read_handle = opener.open(url)
        self.send_response(200)
        self.end_headers()

    def end_headers(self):
        self.remote_headers = {}
        if hasattr(self.read_handle.info(), "headers"):
            headers = self.read_handle.info().headers
            for header in headers:
                hparts = header.strip('\r\n').split(': ', 1)
                self.remote_headers[hparts[0]] = hparts[1]
                self.send_header(*hparts)
        if self.request_version != 'HTTP/0.9':
            self.wfile.write("\r\n")

    def process_text(self, text):
        text = re.sub(ur'(^|\b)([^\W\d_]{6})($|\b)', ur'\1\2™\3', text, flags=re.UNICODE)
        return text

    def process_html(self):
        html = lxml.html.fromstring(self.content.getvalue())
        for element in html.iter(tag='*'):
            if element.tag != 'script':
                if element.text is not None:
                    element.text = self.process_text(element.text)
                if element.tail is not None:
                    element.tail = self.process_text(element.tail)
        self.content = StringIO.StringIO(etree_tostring(
            html, method='html', encoding='utf-8', standalone=True
        ))

    def do_GET(self):
        self.send_head()
        self.content = StringIO.StringIO()
        shutil.copyfileobj(self.read_handle, self.content)
        self.read_handle.close()
        if self.remote_headers.get('Content-Type') == 'text/html; charset=UTF-8':
            self.process_html()
        self.content.seek(0)
        shutil.copyfileobj(self.content, self.wfile)


httpd = SocketServer.ForkingTCPServer((HOST, PORT), Proxy)
print "Serving at {0}:{1}".format(HOST, PORT)
httpd.serve_forever()