changeset:   101031:1f2cfcd5a83f
branch:      3.5
parent:      101027:386712b16c74
user:        Martin Panter <vadmium+py@gmail.com>
date:        Sun Apr 17 02:17:03 2016 +0000
files:       Lib/test/test_wsgiref.py Lib/wsgiref/simple_server.py Misc/ACKS Misc/NEWS
description:
Issue #26717: Stop encoding Latin-1-ized WSGI paths with UTF-8

Patch by Anthony Sottile.


diff -r 386712b16c74 -r 1f2cfcd5a83f Lib/test/test_wsgiref.py
--- a/Lib/test/test_wsgiref.py	Sat Apr 16 23:38:25 2016 +0000
+++ b/Lib/test/test_wsgiref.py	Sun Apr 17 02:17:03 2016 +0000
@@ -1,3 +1,4 @@
+from unittest import mock
 from unittest import TestCase
 from wsgiref.util import setup_testing_defaults
 from wsgiref.headers import Headers
@@ -221,6 +222,29 @@
                 b"data",
                 out)
 
+    def test_cp1252_url(self):
+        def app(e, s):
+            s("200 OK", [
+                ("Content-Type", "text/plain"),
+                ("Date", "Wed, 24 Dec 2008 13:29:32 GMT"),
+                ])
+            # PEP3333 says environ variables are decoded as latin1.
+            # Encode as latin1 to get original bytes
+            return [e["PATH_INFO"].encode("latin1")]
+
+        out, err = run_amock(
+            validator(app), data=b"GET /\x80%80 HTTP/1.0")
+        self.assertEqual(
+            [
+                b"HTTP/1.0 200 OK",
+                mock.ANY,
+                b"Content-Type: text/plain",
+                b"Date: Wed, 24 Dec 2008 13:29:32 GMT",
+                b"",
+                b"/\x80\x80",
+            ],
+            out.splitlines())
+
 
 class UtilityTests(TestCase):
 
diff -r 386712b16c74 -r 1f2cfcd5a83f Lib/wsgiref/simple_server.py
--- a/Lib/wsgiref/simple_server.py	Sat Apr 16 23:38:25 2016 +0000
+++ b/Lib/wsgiref/simple_server.py	Sun Apr 17 02:17:03 2016 +0000
@@ -82,7 +82,7 @@
         else:
             path,query = self.path,''
 
-        env['PATH_INFO'] = urllib.parse.unquote_to_bytes(path).decode('iso-8859-1')
+        env['PATH_INFO'] = urllib.parse.unquote(path, 'iso-8859-1')
         env['QUERY_STRING'] = query
 
         host = self.address_string()
diff -r 386712b16c74 -r 1f2cfcd5a83f Misc/ACKS
--- a/Misc/ACKS	Sat Apr 16 23:38:25 2016 +0000
+++ b/Misc/ACKS	Sun Apr 17 02:17:03 2016 +0000
@@ -1376,6 +1376,7 @@
 Paul Sokolovsky
 Evgeny Sologubov
 Cody Somerville
+Anthony Sottile
 Edoardo Spadolini
 Geoffrey Spear
 Clay Spence
diff -r 386712b16c74 -r 1f2cfcd5a83f Misc/NEWS
--- a/Misc/NEWS	Sat Apr 16 23:38:25 2016 +0000
+++ b/Misc/NEWS	Sun Apr 17 02:17:03 2016 +0000
@@ -107,6 +107,9 @@
 Library
 -------
 
+- Issue #26717: Stop encoding Latin-1-ized WSGI paths with UTF-8.  Patch by
+  Anthony Sottile.
+
 - Issue #26735: Fix :func:`os.urandom` on Solaris 11.3 and newer when reading
   more than 1,024 bytes: call ``getrandom()`` multiple times with a limit of
   1024 bytes per call.