Ticket #19468: 19468-2.diff

diff --git a/django/core/handlers/wsgi.py b/django/core/handlers/wsgi.py
index 4c07105..a148751 100644
--- a/django/core/handlers/wsgi.py
+++ b/django/core/handlers/wsgi.py
@@ -7,11 +7,13 @@ from io import BytesIO
 from threading import Lock
 
 from django import http
+from django.conf import settings
 from django.core import signals
 from django.core.handlers import base
 from django.core.urlresolvers import set_script_prefix
 from django.utils import datastructures
 from django.utils.encoding import force_str, force_text, iri_to_uri
+from django.utils import six
 
 logger = logging.getLogger('django.request')
 
@@ -125,10 +127,31 @@ class LimitedStream(object):
         return line
 
 
+def fix_path_info(path):
+    """
+    On Python 3, wsgiref.WSGIRequestHandler is unconditionally decoding the
+    request path with 'iso-8859-1' encoding. We have to fix this.
+    See https://code.djangoproject.com/ticket/19468 and
+    http://bugs.python.org/issue16679
+    """
+    path_enc = path.encode('iso-8859-1')
+    try:
+        return path_enc.decode('utf-8')
+    except UnicodeDecodeError:
+        if settings.DEFAULT_CHARSET not in ('utf-8', 'iso-8859-1', 'latin-1'):
+            try:
+                return path_enc.decode(settings.DEFAULT_CHARSET)
+            except:
+                pass
+        return path
+
 class WSGIRequest(http.HttpRequest):
     def __init__(self, environ):
         script_name = base.get_script_name(environ)
-        path_info = force_text(environ.get('PATH_INFO', '/'))
+        if six.PY3:
+            path_info = fix_path_info(environ.get('PATH_INFO', '/'))
+        else:
+            path_info = force_text(environ.get('PATH_INFO', '/'))
         if not path_info or path_info == script_name:
             # Sometimes PATH_INFO exists, but is empty (e.g. accessing
             # the SCRIPT_NAME URL without a trailing slash). We really need to
diff --git a/tests/regressiontests/requests/tests.py b/tests/regressiontests/requests/tests.py
index adf824d..432c7d5 100644
--- a/tests/regressiontests/requests/tests.py
+++ b/tests/regressiontests/requests/tests.py
@@ -11,6 +11,7 @@ from django.core.handlers.wsgi import WSGIRequest, LimitedStream
 from django.http import HttpRequest, HttpResponse, parse_cookie, build_request_repr, UnreadablePostError
 from django.test.client import FakePayload
 from django.test.utils import override_settings, str_prefix
+from django.utils import six
 from django.utils import unittest
 from django.utils.http import cookie_date, urlencode
 from django.utils.timezone import utc
@@ -57,6 +58,19 @@ class RequestsTests(unittest.TestCase):
         self.assertEqual(build_request_repr(request, path_override='/otherpath/', GET_override={'a': 'b'}, POST_override={'c': 'd'}, COOKIES_override={'e': 'f'}, META_override={'g': 'h'}),
                          str_prefix("<WSGIRequest\npath:/otherpath/,\nGET:{%(_)s'a': %(_)s'b'},\nPOST:{%(_)s'c': %(_)s'd'},\nCOOKIES:{%(_)s'e': %(_)s'f'},\nMETA:{%(_)s'g': %(_)s'h'}>"))
 
+    @unittest.skipUnless(six.PY3, "PATH_INFO decoding only happens on Python 3")
+    def test_wsgirequest_path_info(self):
+        """Testing fix for #19468"""
+        request = WSGIRequest({'PATH_INFO': '/سلام
/', 'REQUEST_METHOD': 'get', 'wsgi.input': BytesIO(b'')})
+        self.assertEqual(request.path, "/سلام/")
+        # Test path encoded with iso-8859-1
+        request = WSGIRequest({'PATH_INFO': '/café/', 'REQUEST_METHOD': 'get', 'wsgi.input': BytesIO(b'')})
+        self.assertEqual(request.path, "/café/")
+        # Test when DEFAULT_CHARSET is customized
+        with override_settings(DEFAULT_CHARSET='windows-1252'):
+            request = WSGIRequest({'PATH_INFO': '/400\x80/', 'REQUEST_METHOD': 'get', 'wsgi.input': BytesIO(b'')})
+            self.assertEqual(request.path, "/400€/")
+
     def test_parse_cookie(self):
         self.assertEqual(parse_cookie('invalid@key=true'), {})