Fixed #26158 -- Rewrote http.parse_cookie() to better match browsers.

collinanderson · timgraham · commit 93a135d111c2 · 2016-03-15T12:24:06.000-04:00
diff --git a/django/http/cookie.py b/django/http/cookie.py
@@ -57,18 +57,21 @@ def _BaseCookie__set(self, key, real_value, coded_value):
 
 
 def parse_cookie(cookie):
-    if cookie == '':
-        return {}
-    if not isinstance(cookie, http_cookies.BaseCookie):
-        try:
-            c = SimpleCookie()
-            c.load(cookie)
-        except http_cookies.CookieError:
-            # Invalid cookie
-            return {}
-    else:
-        c = cookie
+    """
+    Return a dictionary parsed from a `Cookie:` header string.
+    """
     cookiedict = {}
-    for key in c.keys():
-        cookiedict[key] = c.get(key).value
+    if six.PY2:
+        cookie = force_str(cookie)
+    for chunk in cookie.split(str(';')):
+        if str('=') in chunk:
+            key, val = chunk.split(str('='), 1)
+        else:
+            # Assume an empty name per
+            # https://bugzilla.mozilla.org/show_bug.cgi?id=169091
+            key, val = str(''), chunk
+        key, val = key.strip(), val.strip()
+        if key or val:
+            # unquote using Python's algorithm.
+            cookiedict[key] = http_cookies._unquote(val)
     return cookiedict
diff --git a/docs/releases/1.10.txt b/docs/releases/1.10.txt
@@ -354,6 +354,10 @@ Requests and Responses
   :attr:`~django.http.HttpRequest.content_params` attributes which are
   parsed from the ``CONTENT_TYPE`` header.
 
+* The parser for ``request.COOKIES`` is simplified to better match the behavior
+  of browsers. ``request.COOKIES`` may now contain cookies that are invalid
+  according to :rfc:`6265` but are possible to set via ``document.cookie``.
+
 Serialization
 ~~~~~~~~~~~~~
 
diff --git a/tests/httpwrappers/tests.py b/tests/httpwrappers/tests.py
@@ -676,6 +676,8 @@ def test_decode(self):
         c2 = SimpleCookie()
         c2.load(c.output()[12:])
         self.assertEqual(c['test'].value, c2['test'].value)
+        c3 = parse_cookie(c.output()[12:])
+        self.assertEqual(c['test'].value, c3['test'])
 
     def test_decode_2(self):
         """
@@ -686,6 +688,8 @@ def test_decode_2(self):
         c2 = SimpleCookie()
         c2.load(c.output()[12:])
         self.assertEqual(c['test'].value, c2['test'].value)
+        c3 = parse_cookie(c.output()[12:])
+        self.assertEqual(c['test'].value, c3['test'])
 
     def test_nonstandard_keys(self):
         """
@@ -699,6 +703,52 @@ def test_repeated_nonstandard_keys(self):
         """
         self.assertIn('good_cookie', parse_cookie('a:=b; a:=c; good_cookie=yes').keys())
 
+    def test_python_cookies(self):
+        """
+        Test cases copied from Python's Lib/test/test_http_cookies.py
+        """
+        self.assertEqual(parse_cookie('chips=ahoy; vienna=finger'), {'chips': 'ahoy', 'vienna': 'finger'})
+        # Here parse_cookie() differs from Python's cookie parsing in that it
+        # treats all semicolons as delimiters, even within quotes.
+        self.assertEqual(
+            parse_cookie('keebler="E=mc2; L=\\"Loves\\"; fudge=\\012;"'),
+            {'keebler': '"E=mc2', 'L': '\\"Loves\\"', 'fudge': '\\012', '': '"'}
+        )
+        # Illegal cookies that have an '=' char in an unquoted value.
+        self.assertEqual(parse_cookie('keebler=E=mc2'), {'keebler': 'E=mc2'})
+        # Cookies with ':' character in their name.
+        self.assertEqual(parse_cookie('key:term=value:term'), {'key:term': 'value:term'})
+        # Cookies with '[' and ']'.
+        self.assertEqual(parse_cookie('a=b; c=[; d=r; f=h'), {'a': 'b', 'c': '[', 'd': 'r', 'f': 'h'})
+
+    def test_cookie_edgecases(self):
+        # Cookies that RFC6265 allows.
+        self.assertEqual(parse_cookie('a=b; Domain=example.com'), {'a': 'b', 'Domain': 'example.com'})
+        # parse_cookie() has historically kept only the last cookie with the
+        # same name.
+        self.assertEqual(parse_cookie('a=b; h=i; a=c'), {'a': 'c', 'h': 'i'})
+
+    def test_invalid_cookies(self):
+        """
+        Cookie strings that go against RFC6265 but browsers will send if set
+        via document.cookie.
+        """
+        # Chunks without an equals sign appear as unnamed values per
+        # https://bugzilla.mozilla.org/show_bug.cgi?id=169091
+        self.assertIn('django_language', parse_cookie('abc=def; unnamed; django_language=en').keys())
+        # Even a double quote may be an unamed value.
+        self.assertEqual(parse_cookie('a=b; "; c=d'), {'a': 'b', '': '"', 'c': 'd'})
+        # Spaces in names and values, and an equals sign in values.
+        self.assertEqual(parse_cookie('a b c=d e = f; gh=i'), {'a b c': 'd e = f', 'gh': 'i'})
+        # More characters the spec forbids.
+        self.assertEqual(parse_cookie('a   b,c<>@:/[]?{}=d  "  =e,f g'), {'a   b,c<>@:/[]?{}': 'd  "  =e,f g'})
+        # Unicode characters. The spec only allows ASCII.
+        self.assertEqual(parse_cookie('saint=André Bessette'), {'saint': force_str('André Bessette')})
+        # Browsers don't send extra whitespace or semicolons in Cookie headers,
+        # but parse_cookie() should parse whitespace the same way
+        # document.cookie parses whitespace.
+        self.assertEqual(parse_cookie('  =  b  ;  ;  =  ;   c  =  ;  '), {'': 'b', 'c': ''})
+
     def test_httponly_after_load(self):
         """
         Test that we can use httponly attribute on cookies that we load
diff --git a/tests/requests/tests.py b/tests/requests/tests.py
@@ -10,7 +10,6 @@
 from django.core.handlers.wsgi import LimitedStream, WSGIRequest
 from django.http import (
     HttpRequest, HttpResponse, RawPostDataException, UnreadablePostError,
-    parse_cookie,
 )
 from django.test import RequestFactory, SimpleTestCase, override_settings
 from django.test.client import FakePayload
@@ -183,9 +182,6 @@ def wsgi_str(path_info):
         request = WSGIRequest({'PATH_INFO': wsgi_str("/سلام/"), 'REQUEST_METHOD': 'get', 'wsgi.input': BytesIO(b'')})
         self.assertEqual(request.path, "/سلام/")
 
-    def test_parse_cookie(self):
-        self.assertEqual(parse_cookie('invalid@key=true'), {})
-
     def test_httprequest_location(self):
         request = HttpRequest()
         self.assertEqual(request.build_absolute_uri(location="https://www.example.com/asdf"),
