From 8dc73e8425f0672bb7e4734902879de53c6f7125 Mon Sep 17 00:00:00 2001
From: Russell Keith-Magee <russell@keith-magee.com>
Date: Sat, 11 Apr 2009 09:20:10 +0000
Subject: [PATCH] Fixed #10571 -- Ensured that unicode POST data is correctly
 encoded by the test client. Thanks to Rick Wagner for his help identifying
 and fixing this problem.

git-svn-id: http://code.djangoproject.com/svn/django/trunk@10513 bcc190cf-cafb-0310-a4f2-bffc1f526a37
---
 AUTHORS                                       |  1 +
 django/test/client.py                         | 11 +++++--
 .../test_client_regress/models.py             | 33 +++++++++++++++++++
 .../test_client_regress/urls.py               |  1 +
 .../test_client_regress/views.py              | 23 +++++++++++++
 5 files changed, 67 insertions(+), 2 deletions(-)

diff --git a/AUTHORS b/AUTHORS
index 1bfb3ad975..94bd91e906 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -436,6 +436,7 @@ answer newbie questions, and generally made Django that much better:
     Vlado <vlado@labath.org>
     Milton Waddams
     Chris Wagner <cw264701@ohio.edu>
+    Rick Wagner <rwagner@physics.ucsd.edu>
     wam-djangobug@wamber.net
     Wang Chun <wangchun@exoweb.net>
     Filip Wasilewski <filip.wasilewski@gmail.com>
diff --git a/django/test/client.py b/django/test/client.py
index 47232ef146..348bfa7d26 100644
--- a/django/test/client.py
+++ b/django/test/client.py
@@ -2,6 +2,7 @@ import urllib
 from urlparse import urlparse, urlunparse, urlsplit
 import sys
 import os
+import re
 try:
     from cStringIO import StringIO
 except ImportError:
@@ -25,7 +26,7 @@ from django.test.utils import ContextList
 
 BOUNDARY = 'BoUnDaRyStRiNg'
 MULTIPART_CONTENT = 'multipart/form-data; boundary=%s' % BOUNDARY
-
+CONTENT_TYPE_RE = re.compile('.*; charset=([\w\d-]+);?')
 
 class FakePayload(object):
     """
@@ -290,7 +291,13 @@ class Client(object):
         if content_type is MULTIPART_CONTENT:
             post_data = encode_multipart(BOUNDARY, data)
         else:
-            post_data = data
+            # Encode the content so that the byte representation is correct.
+            match = CONTENT_TYPE_RE.match(content_type)
+            if match:
+                charset = match.group(1)
+            else:
+                charset = settings.DEFAULT_CHARSET
+            post_data = smart_str(data, encoding=charset)
 
         parsed = urlparse(path)
         r = {
diff --git a/tests/regressiontests/test_client_regress/models.py b/tests/regressiontests/test_client_regress/models.py
index fe55ffff92..be2cb8fa37 100644
--- a/tests/regressiontests/test_client_regress/models.py
+++ b/tests/regressiontests/test_client_regress/models.py
@@ -624,3 +624,36 @@ class QueryStringTests(TestCase):
         self.assertEqual(response.context['post-bar'], 'bang')
         self.assertEqual(response.context['request-foo'], 'whiz')
         self.assertEqual(response.context['request-bar'], 'bang')
+
+class UnicodePayloadTests(TestCase):
+    def test_simple_unicode_payload(self):
+        "A simple ASCII-only unicode JSON document can be POSTed"
+        # Regression test for #10571
+        json = u'{"english": "mountain pass"}'
+        response = self.client.post("/test_client_regress/parse_unicode_json/", json,
+                                    content_type="application/json")
+        self.assertEqual(response.content, json)
+
+    def test_unicode_payload_utf8(self):
+        "A non-ASCII unicode data encoded as UTF-8 can be POSTed"
+        # Regression test for #10571
+        json = u'{"dog": "собака"}'
+        response = self.client.post("/test_client_regress/parse_unicode_json/", json,
+                                    content_type="application/json; charset=utf-8")
+        self.assertEqual(response.content, json.encode('utf-8'))
+
+    def test_unicode_payload_utf16(self):
+        "A non-ASCII unicode data encoded as UTF-16 can be POSTed"
+        # Regression test for #10571
+        json = u'{"dog": "собака"}'
+        response = self.client.post("/test_client_regress/parse_unicode_json/", json,
+                                    content_type="application/json; charset=utf-16")
+        self.assertEqual(response.content, json.encode('utf-16'))
+
+    def test_unicode_payload_non_utf(self):
+        "A non-ASCII unicode data as a non-UTF based encoding can be POSTed"
+        #Regression test for #10571
+        json = u'{"dog": "собака"}'
+        response = self.client.post("/test_client_regress/parse_unicode_json/", json,
+                                    content_type="application/json; charset=koi8-r")
+        self.assertEqual(response.content, json.encode('koi8-r'))
diff --git a/tests/regressiontests/test_client_regress/urls.py b/tests/regressiontests/test_client_regress/urls.py
index c5872bf899..99eb7b70be 100644
--- a/tests/regressiontests/test_client_regress/urls.py
+++ b/tests/regressiontests/test_client_regress/urls.py
@@ -23,4 +23,5 @@ urlpatterns = patterns('',
     (r'^check_session/$', views.check_session_view),
     (r'^request_methods/$', views.request_methods_view),
     (r'^check_unicode/$', views.return_unicode),
+    (r'^parse_unicode_json/$', views.return_json_file),
 )
diff --git a/tests/regressiontests/test_client_regress/views.py b/tests/regressiontests/test_client_regress/views.py
index 75b8d887e8..5ba7cc30a9 100644
--- a/tests/regressiontests/test_client_regress/views.py
+++ b/tests/regressiontests/test_client_regress/views.py
@@ -1,7 +1,12 @@
+from django.conf import settings
 from django.contrib.auth.decorators import login_required
 from django.http import HttpResponse, HttpResponseRedirect
 from django.core.exceptions import SuspiciousOperation
 from django.shortcuts import render_to_response
+from django.utils import simplejson
+from django.utils.encoding import smart_str
+from django.core.serializers.json import DjangoJSONEncoder
+from django.test.client import CONTENT_TYPE_RE
 
 def no_template_view(request):
     "A simple view that expects a GET request, and returns a rendered template"
@@ -63,3 +68,21 @@ def request_methods_view(request):
 
 def return_unicode(request):
     return render_to_response('unicode.html')
+
+def return_json_file(request):
+    "A view that parses and returns a JSON string as a file."
+    match = CONTENT_TYPE_RE.match(request.META['CONTENT_TYPE'])
+    if match:
+        charset = match.group(1)
+    else:
+        charset = settings.DEFAULT_CHARSET
+
+    # This just checks that the uploaded data is JSON
+    obj_dict = simplejson.loads(request.raw_post_data.decode(charset))
+    obj_json = simplejson.dumps(obj_dict, encoding=charset,
+                                cls=DjangoJSONEncoder,
+                                ensure_ascii=False)
+    response = HttpResponse(smart_str(obj_json, encoding=charset), status=200,
+                            mimetype='application/json; charset=' + charset)
+    response['Content-Disposition'] = 'attachment; filename=testfile.json'
+    return response