Use Forwarded, X-Forwarded-Scheme and X-Forwarded-Host for better scheme and host resolution

CHANGES.rst

@@ -41,6 +41,8 @@ Changes
 
 - Do not unquote `+` in match_info values #1816
 
+- Use Forwarded, X-Forwarded-Scheme and X-Forwarded-Host for better scheme and host resolution. #1134
+
 - Fix sub-application middlewares resolution order #1853
 
 

CONTRIBUTORS.txt

@@ -62,6 +62,7 @@ Enrique Saez
 Erich Healy
 Eugene Chernyshov
 Eugene Naydenov
+Evert Lammerts
 Frederik Gladhorn
 Frederik Peter Aalund
 Gabriel Tremblay

aiohttp/hdrs.py

@@ -50,6 +50,7 @@
 ETAG = istr('ETAG')
 EXPECT = istr('EXPECT')
 EXPIRES = istr('EXPIRES')
+FORWARDED = istr('FORWARDED')
 FROM = istr('FROM')
 HOST = istr('HOST')
 IF_MATCH = istr('IF-MATCH')
@@ -89,3 +90,5 @@
 WANT_DIGEST = istr('WANT-DIGEST')
 WARNING = istr('WARNING')
 WWW_AUTHENTICATE = istr('WWW-AUTHENTICATE')
+X_FORWARDED_HOST = istr('X-FORWARDED-HOST')
+X_FORWARDED_PROTO = istr('X-FORWARDED-PROTO')

aiohttp/web_request.py

@@ -150,16 +150,31 @@ def secure(self):
         """
         return self.url.scheme == 'https'
 
+    @reify
+    def _forwarded(self):
+        forwarded = self._message.headers.get(hdrs.FORWARDED)
+        if forwarded is not None:
+            parsed = re.findall(
+                r'^by=([^;]*); +for=([^;]*); +host=([^;]*); +proto=(https?)$',
+                forwarded)
+            if parsed:
+                return parsed[0]
+        return None
+
     @reify
     def _scheme(self):
+        proto = 'http'
         if self._transport.get_extra_info('sslcontext'):
-            return 'https'
-        secure_proxy_ssl_header = self._secure_proxy_ssl_header
-        if secure_proxy_ssl_header is not None:
-            header, value = secure_proxy_ssl_header
+            proto = 'https'
+        elif self._secure_proxy_ssl_header is not None:
+            header, value = self._secure_proxy_ssl_header
             if self.headers.get(header) == value:
-                return 'https'
-        return 'http'
+                proto = 'https'
+        elif self._forwarded:
+            _, _, _, proto = self._forwarded
+        elif hdrs.X_FORWARDED_PROTO in self._message.headers:
+            proto = self._message.headers[hdrs.X_FORWARDED_PROTO]
+        return proto
 
     @property
     def method(self):
@@ -179,16 +194,29 @@ def version(self):
 
     @reify
     def host(self):
-        """Read only property for getting *HOST* header of request.
+        """ Hostname of the request.
+
+        Hostname is resolved through the following headers, in this order:
+
+        - Forwarded
+        - X-Forwarded-Host
+        - Host
 
-        Returns str or None if HTTP request has no HOST header.
+        Returns str, or None if no hostname is found in the headers.
         """
-        return self._message.headers.get(hdrs.HOST)
+        host = None
+        if self._forwarded:
+            _, _, host, _ = self._forwarded
+        elif hdrs.X_FORWARDED_HOST in self._message.headers:
+            host = self._message.headers[hdrs.X_FORWARDED_HOST]
+        elif hdrs.HOST in self._message.headers:
+            host = self._message.headers[hdrs.HOST]
+        return host
 
     @reify
     def url(self):
         return URL('{}://{}{}'.format(self._scheme,
-                                      self._message.headers.get(hdrs.HOST),
+                                      self.host,
                                       str(self._rel_url)))
 
     @property

tests/test_web_request.py

@@ -248,6 +248,62 @@ def test_https_scheme_by_secure_proxy_ssl_header_false_test(make_request):
     assert req.secure is False
 
 
+def test_https_scheme_by_forwarded_header(make_request):
+    req = make_request('GET', '/',
+                       headers=CIMultiDict(
+                           {'Forwarded': 'by=; for=; host=; proto=https'}))
+    assert "https" == req.scheme
+    assert req.secure is True
+
+
+def test_https_scheme_by_malformed_forwarded_header(make_request):
+    req = make_request('GET', '/',
+                       headers=CIMultiDict({'Forwarded': 'malformed value'}))
+    assert "http" == req.scheme
+    assert req.secure is False
+
+
+def test_https_scheme_by_x_forwarded_proto_header(make_request):
+    req = make_request('GET', '/',
+                       headers=CIMultiDict({'X-Forwarded-Proto': 'https'}))
+    assert "https" == req.scheme
+    assert req.secure is True
+
+
+def test_https_scheme_by_x_forwarded_proto_header_no_tls(make_request):
+    req = make_request('GET', '/',
+                       headers=CIMultiDict({'X-Forwarded-Proto': 'http'}))
+    assert "http" == req.scheme
+    assert req.secure is False
+
+
+def test_host_by_forwarded_header(make_request):
+    req = make_request('GET', '/',
+                       headers=CIMultiDict(
+                           {'Forwarded': 'by=; for=; host'
+                                         '=example.com; proto=https'}))
+    assert req.host == 'example.com'
+
+
+def test_host_by_forwarded_header_malformed(make_request):
+    req = make_request('GET', '/',
+                       headers=CIMultiDict({'Forwarded': 'malformed value'}))
+    assert req.host is None
+
+
+def test_host_by_x_forwarded_host_header(make_request):
+    req = make_request('GET', '/',
+                       headers=CIMultiDict(
+                           {'X-Forwarded-Host': 'example.com'}))
+    assert req.host == 'example.com'
+
+
+def test_host_by_host_header(make_request):
+    req = make_request('GET', '/',
+                       headers=CIMultiDict({'Host': 'example.com'}))
+    assert req.host == 'example.com'
+
+
 def test_raw_headers(make_request):
     req = make_request('GET', '/',
                        headers=CIMultiDict({'X-HEADER': 'aaa'}))