├── .gitignore
├── README.md
└── proxy.py


/.gitignore:
--------------------------------------------------------------------------------
 1 | *.py[cod]
 2 | 
 3 | # C extensions
 4 | *.so
 5 | 
 6 | # Packages
 7 | *.egg
 8 | *.egg-info
 9 | dist
10 | build
11 | eggs
12 | parts
13 | bin
14 | var
15 | sdist
16 | develop-eggs
17 | .installed.cfg
18 | lib
19 | lib64
20 | 
21 | # Installer logs
22 | pip-log.txt
23 | 
24 | # Unit test / coverage reports
25 | .coverage
26 | .tox
27 | nosetests.xml
28 | 
29 | # Translations
30 | *.mo
31 | 
32 | # Mr Developer
33 | .mr.developer.cfg
34 | .project
35 | .pydevproject
36 | 


--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
 1 | flask-as-http-proxy-server
 2 | ==========================
 3 | 
 4 | Small HTTP Server used with Flask and werkzeug
 5 | 
 6 | test
 7 | =======
 8 | 
 9 | * works with python 2/3
10 | * the only needed library is flask :)
11 | 
12 | License
13 | =======
14 | BSD. use your own risk.
15 | 


--------------------------------------------------------------------------------
/proxy.py:
--------------------------------------------------------------------------------
  1 | #!/usr/bin/python
  2 | # -*- coding:utf-8 -*-
  3 | 
  4 | #############################################
  5 | # Flask & werkzeug HTTP Proxy Sample code.
  6 | # - Code by Jioh L. Jung (ziozzang@gmail.com)
  7 | #############################################
  8 | 
  9 | import sys
 10 | if sys.version_info[0] < 3:
 11 |   import httplib
 12 |   import urlparse
 13 | else:
 14 |   import http.client as httplib
 15 |   import urllib.parse as urlparse
 16 | 
 17 | import re
 18 | import urllib
 19 | import json
 20 | 
 21 | from flask import Flask, Blueprint, request, Response, url_for
 22 | from werkzeug.datastructures import Headers
 23 | from werkzeug.exceptions import NotFound
 24 | 
 25 | app = Flask(__name__)
 26 | 
 27 | # Default Configuration
 28 | DEBUG_FLAG = True
 29 | LISTEN_PORT = 7788
 30 | 
 31 | proxy = Blueprint('proxy', __name__)
 32 | 
 33 | # You can insert Authentication here.
 34 | #proxy.before_request(check_login)
 35 | 
 36 | # Filters.
 37 | HTML_REGEX = re.compile(r'((?:src|action|href)=["\'])/')
 38 | JQUERY_REGEX = re.compile(r'(\$\.(?:get|post)\(["\'])/')
 39 | JS_LOCATION_REGEX = re.compile(r'((?:window|document)\.location.*=.*["\'])/')
 40 | CSS_REGEX = re.compile(r'(url\(["\']?)/')
 41 | 
 42 | REGEXES = [HTML_REGEX, JQUERY_REGEX, JS_LOCATION_REGEX, CSS_REGEX]
 43 | 
 44 | 
 45 | def iterform(multidict):
 46 |     for key in multidict.keys():
 47 |         for value in multidict.getlist(key):
 48 |             yield (key.encode("utf8"), value.encode("utf8"))
 49 | 
 50 | def parse_host_port(h, proto):
 51 |     """Parses strings in the form host[:port]"""
 52 |     host_port = h.split(":", 1)
 53 |     if len(host_port) == 1:
 54 |         if proto.lower() == 'http': return (h, 80)
 55 |         if proto.lower() == 'https': return (h, 443)
 56 |         return (h, 443) # Default is HTTPS
 57 |     else:
 58 |         host_port[1] = int(host_port[1])
 59 |         return host_port
 60 | 
 61 | 
 62 | # For RESTful Service
 63 | @proxy.route('/proxy/<proto>/<host>/', methods=["GET", "POST", "PUT", "DELETE"])
 64 | @proxy.route('/proxy/<proto>/<host>/<path:file>', methods=["GET", "POST", "PUT", "DELETE"])
 65 | def proxy_request(proto, host, file=""):
 66 |     hostname, port = parse_host_port(host, proto)
 67 | 
 68 |     print ("H: '%s' P: %d" % (hostname, port))
 69 |     print ("F: '%s'" % (file))
 70 |     # Whitelist a few headers to pass on
 71 |     request_headers = {}
 72 |     for h in ["Cookie", "Referer", "X-Csrf-Token"]:
 73 |         if h in request.headers:
 74 |             request_headers[h] = request.headers[h]
 75 | 
 76 |     if request.query_string:
 77 |         path = "/%s?%s" % (file, request.query_string)
 78 |     else:
 79 |         path = "/" + file
 80 | 
 81 |     if request.method == "POST" or request.method == "PUT":
 82 |         form_data = list(iterform(request.form))
 83 |         form_data = urllib.urlencode(form_data)
 84 |         request_headers["Content-Length"] = len(form_data)
 85 |     else:
 86 |         form_data = None
 87 | 
 88 | 
 89 |     if not ('host' in request_headers.keys()):
 90 |         request_headers['host'] = hostname
 91 | 
 92 |     # if target is for HTTP, use HTTPConnection method.
 93 |     request_method = httplib.HTTPSConnection
 94 |     if proto.lower() == 'http': request_method = httplib.HTTPConnection
 95 |     conn = request_method(hostname, port)
 96 |     conn.request(request.method, path, body=form_data, headers=request_headers)
 97 |     resp = conn.getresponse()
 98 | 
 99 |     # Clean up response headers for forwarding
100 |     d = {}
101 |     response_headers = Headers()
102 |     for key, value in resp.getheaders():
103 |         print ("HEADER: '%s':'%s'" % (key, value))
104 |         d[key.lower()] = value
105 |         if key in ["content-length", "connection", "content-type"]:
106 |             continue
107 | 
108 |         if key == "set-cookie":
109 |             cookies = value.split(",")
110 |             [response_headers.add(key, c) for c in cookies]
111 |         else:
112 |             response_headers.add(key, value)
113 | 
114 |     # If this is a redirect, munge the Location URL
115 |     if "location" in response_headers:
116 |         redirect = response_headers["location"]
117 |         parsed = urlparse.urlparse(request.url)
118 |         redirect_parsed = urlparse.urlparse(redirect)
119 | 
120 |         redirect_host = redirect_parsed.netloc
121 |         if not redirect_host:
122 |             redirect_host = "%s:%d" % (hostname, port)
123 | 
124 |         redirect_path = redirect_parsed.path
125 |         if redirect_parsed.query:
126 |             redirect_path += "?" + redirect_parsed.query
127 | 
128 |         munged_path = url_for(".proxy_request",
129 |                               proto=proto,
130 |                               host=redirect_host,
131 |                               file=redirect_path[1:])
132 | 
133 |         url = "%s://%s%s" % (parsed.scheme, parsed.netloc, munged_path)
134 |         response_headers["location"] = url
135 | 
136 |     # Rewrite URLs in the content to point to our URL schemt.method == " instead.
137 |     # Ugly, but seems to mostly work.
138 |     root = url_for(".proxy_request", proto=proto, host=host)
139 |     contents = resp.read()
140 | 
141 |     # Restructing Contents.
142 |     if "content-type" in d.keys():
143 |       if d["content-type"].find("application/json") >= 0:
144 |         # JSON format conentens will be modified here.
145 |         jc = json.loads(contents)
146 |         if jc.has_key("nodes"):
147 |             del jc["nodes"]
148 |         contents = json.dumps(jc)
149 | 
150 |       else:
151 |         # Generic HTTP.
152 |         pass
153 | 
154 |         # only valid for python2 / cuz, python3's string & byte handling is diffrent from python2
155 |         #for regex in REGEXES:
156 |         #   contents = regex.sub(r'\1%s' % root, contents)
157 |     else:
158 |       # set default content-type, for error handling
159 |       d['content-type'] = 'text/html; charset=utf-8'
160 | 
161 |     # Remove transfer-encoding: chunked header. cuz proxy does not use chunk trnasfer.
162 |     if 'transfer-encoding' in d:
163 |       if d['transfer-encoding'].lower() == 'chunked':
164 |         del(d['transfer-encoding'])
165 |         d['content-length'] = len(contents)
166 | 
167 |     flask_response = Response(response=contents,
168 |                               status=resp.status,
169 |                               headers=d)
170 |     return flask_response
171 | 
172 | 
173 | app.register_blueprint(proxy)
174 | app.run(debug=DEBUG_FLAG, host='0.0.0.0', port=LISTEN_PORT)
175 | 


--------------------------------------------------------------------------------