From 3d8161a1248c67db2d896a4b1355eac6f60825d2 Mon Sep 17 00:00:00 2001
From: Tim Burke <tim.burke@gmail.com>
Date: Wed, 13 Mar 2019 13:13:19 -0700
Subject: [PATCH] bpo-36274: Encode request lines as Latin-1

While this is out of spec according to RFC 7230 (which limits
expected octets to some subset of ASCII), it is often useful to
be able to mimic an out-of-spec client when testing a server or
application.

Use Latin-1 in keeping with how we handle headers and bodies.

https://bugs.python.org/issue36274
---
 Lib/http/client.py                            |  4 +--
 Lib/test/test_httplib.py                      | 34 +++++++++++++++++--
 .../2019-07-08-09-20-10.bpo-36274.8XicsH.rst  |  2 ++
 3 files changed, 36 insertions(+), 4 deletions(-)
 create mode 100644 Misc/NEWS.d/next/Library/2019-07-08-09-20-10.bpo-36274.8XicsH.rst

diff --git a/Lib/http/client.py b/Lib/http/client.py
index f61267e108a5249..f28381b4cda9424 100644
--- a/Lib/http/client.py
+++ b/Lib/http/client.py
@@ -1095,8 +1095,8 @@ def putrequest(self, method, url, skip_host=False,
                              f"(found at least {match.group()!r})")
         request = '%s %s %s' % (method, url, self._http_vsn_str)
 
-        # Non-ASCII characters should have been eliminated earlier
-        self._output(request.encode('ascii'))
+        # Encode as latin-1, like we do headers and data
+        self._output(request.encode('latin-1'))
 
         if self._http_vsn == 11:
             # Issue some standard headers for better HTTP/1.1 compliance
diff --git a/Lib/test/test_httplib.py b/Lib/test/test_httplib.py
index 9148169cc7c2e57..f399b46027da3d9 100644
--- a/Lib/test/test_httplib.py
+++ b/Lib/test/test_httplib.py
@@ -275,6 +275,37 @@ def test_ipv6host_header(self):
         conn.request('GET', '/foo')
         self.assertTrue(sock.data.startswith(expected))
 
+    def test_request_path_handling(self):
+        happy_cases = (
+            ('/', b'/'),
+            ('', b'/'),
+            ('/\xe4\xbd\xa0\xe5\xa5\xbd',
+             b'/\xe4\xbd\xa0\xe5\xa5\xbd'),
+        )
+        for caller_path, expected_path in happy_cases:
+            with self.subTest((caller_path, expected_path)):
+                conn = client.HTTPConnection('server.fqdn')
+                sock = FakeSocket('')
+                conn.sock = sock
+                conn.request('GET', caller_path)
+                expected = (b'GET ' + expected_path + b' HTTP/1.1\r\n'
+                            b'Host: server.fqdn\r\n'
+                            b'Accept-Encoding: identity\r\n\r\n')
+                self.assertEqual(sock.data, expected)
+
+        error_cases = (
+            '/\u4f60\u597d',
+            '/\udce4\udcbd\udca0\udce5\udca5\udcbd',
+        )
+        for caller_path in error_cases:
+            with self.subTest(caller_path):
+                conn = client.HTTPConnection('server.fqdn')
+                sock = FakeSocket('')
+                conn.sock = sock
+                with self.assertRaises(UnicodeEncodeError):
+                    conn.request('GET', caller_path)
+                self.assertEqual(sock.data, b'')
+
     def test_malformed_headers_coped_with(self):
         # Issue 19996
         body = "HTTP/1.1 200 OK\r\nFirst: val\r\n: nval\r\nSecond: val\r\n\r\n"
@@ -720,8 +751,7 @@ def test_send_file(self):
             sock = FakeSocket(body)
             conn.sock = sock
             conn.request('GET', '/foo', body)
-            self.assertTrue(sock.data.startswith(expected), '%r != %r' %
-                    (sock.data[:len(expected)], expected))
+            self.assertEqual(sock.data[:len(expected)], expected)
 
     def test_send(self):
         expected = b'this is a test this is only a test'
diff --git a/Misc/NEWS.d/next/Library/2019-07-08-09-20-10.bpo-36274.8XicsH.rst b/Misc/NEWS.d/next/Library/2019-07-08-09-20-10.bpo-36274.8XicsH.rst
new file mode 100644
index 000000000000000..ce7f218bb0988f2
--- /dev/null
+++ b/Misc/NEWS.d/next/Library/2019-07-08-09-20-10.bpo-36274.8XicsH.rst
@@ -0,0 +1,2 @@
+``http.client`` now accepts ISO-8859-1 request-targets. Callers are still
+encouraged to URL-quote the request-target so as to comply with the RFC.