[core] Improve HTTP redirect handling (#7094)

Aligns HTTP redirect handling with what browsers commonly do and RFC standards. 

Fixes issues afac4caa7d missed.

Authored by: coletdjnz
This commit is contained in:
coletdjnz 2023-05-27 19:06:13 +12:00 committed by GitHub
parent 66468bbf49
commit 08916a49c7
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
3 changed files with 281 additions and 72 deletions

View file

@ -10,7 +10,6 @@ sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
import copy
import json
import urllib.error
from test.helper import FakeYDL, assertRegexpMatches
from yt_dlp import YoutubeDL
@ -1097,11 +1096,6 @@ class TestYoutubeDL(unittest.TestCase):
test_selection({'playlist_items': '-15::2'}, INDICES[1::2], True)
test_selection({'playlist_items': '-15::15'}, [], True)
def test_urlopen_no_file_protocol(self):
# see https://github.com/ytdl-org/youtube-dl/issues/8227
ydl = YDL()
self.assertRaises(urllib.error.URLError, ydl.urlopen, 'file:///etc/passwd')
def test_do_not_override_ie_key_in_url_transparent(self):
ydl = YDL()

View file

@ -7,40 +7,163 @@ import unittest
sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
import gzip
import http.cookiejar
import http.server
import io
import pathlib
import ssl
import tempfile
import threading
import urllib.error
import urllib.request
from test.helper import http_server_port
from yt_dlp import YoutubeDL
from yt_dlp.utils import sanitized_Request, urlencode_postdata
from .helper import FakeYDL
TEST_DIR = os.path.dirname(os.path.abspath(__file__))
class HTTPTestRequestHandler(http.server.BaseHTTPRequestHandler):
protocol_version = 'HTTP/1.1'
def log_message(self, format, *args):
pass
def _headers(self):
payload = str(self.headers).encode('utf-8')
self.send_response(200)
self.send_header('Content-Type', 'application/json')
self.send_header('Content-Length', str(len(payload)))
self.end_headers()
self.wfile.write(payload)
def _redirect(self):
self.send_response(int(self.path[len('/redirect_'):]))
self.send_header('Location', '/method')
self.send_header('Content-Length', '0')
self.end_headers()
def _method(self, method, payload=None):
self.send_response(200)
self.send_header('Content-Length', str(len(payload or '')))
self.send_header('Method', method)
self.end_headers()
if payload:
self.wfile.write(payload)
def _status(self, status):
payload = f'<html>{status} NOT FOUND</html>'.encode()
self.send_response(int(status))
self.send_header('Content-Type', 'text/html; charset=utf-8')
self.send_header('Content-Length', str(len(payload)))
self.end_headers()
self.wfile.write(payload)
def _read_data(self):
if 'Content-Length' in self.headers:
return self.rfile.read(int(self.headers['Content-Length']))
def do_POST(self):
data = self._read_data()
if self.path.startswith('/redirect_'):
self._redirect()
elif self.path.startswith('/method'):
self._method('POST', data)
elif self.path.startswith('/headers'):
self._headers()
else:
self._status(404)
def do_HEAD(self):
if self.path.startswith('/redirect_'):
self._redirect()
elif self.path.startswith('/method'):
self._method('HEAD')
else:
self._status(404)
def do_PUT(self):
data = self._read_data()
if self.path.startswith('/redirect_'):
self._redirect()
elif self.path.startswith('/method'):
self._method('PUT', data)
else:
self._status(404)
def do_GET(self):
if self.path == '/video.html':
payload = b'<html><video src="/vid.mp4" /></html>'
self.send_response(200)
self.send_header('Content-Type', 'text/html; charset=utf-8')
self.send_header('Content-Length', str(len(payload))) # required for persistent connections
self.end_headers()
self.wfile.write(b'<html><video src="/vid.mp4" /></html>')
self.wfile.write(payload)
elif self.path == '/vid.mp4':
payload = b'\x00\x00\x00\x00\x20\x66\x74[video]'
self.send_response(200)
self.send_header('Content-Type', 'video/mp4')
self.send_header('Content-Length', str(len(payload)))
self.end_headers()
self.wfile.write(b'\x00\x00\x00\x00\x20\x66\x74[video]')
self.wfile.write(payload)
elif self.path == '/%E4%B8%AD%E6%96%87.html':
payload = b'<html><video src="/vid.mp4" /></html>'
self.send_response(200)
self.send_header('Content-Type', 'text/html; charset=utf-8')
self.send_header('Content-Length', str(len(payload)))
self.end_headers()
self.wfile.write(payload)
elif self.path == '/%c7%9f':
payload = b'<html><video src="/vid.mp4" /></html>'
self.send_response(200)
self.send_header('Content-Type', 'text/html; charset=utf-8')
self.send_header('Content-Length', str(len(payload)))
self.end_headers()
self.wfile.write(payload)
elif self.path.startswith('/redirect_'):
self._redirect()
elif self.path.startswith('/method'):
self._method('GET')
elif self.path.startswith('/headers'):
self._headers()
elif self.path == '/trailing_garbage':
payload = b'<html><video src="/vid.mp4" /></html>'
self.send_response(200)
self.send_header('Content-Type', 'text/html; charset=utf-8')
self.send_header('Content-Encoding', 'gzip')
buf = io.BytesIO()
with gzip.GzipFile(fileobj=buf, mode='wb') as f:
f.write(payload)
compressed = buf.getvalue() + b'trailing garbage'
self.send_header('Content-Length', str(len(compressed)))
self.end_headers()
self.wfile.write(compressed)
elif self.path == '/302-non-ascii-redirect':
new_url = f'http://127.0.0.1:{http_server_port(self.server)}/中文.html'
self.send_response(301)
self.send_header('Location', new_url)
self.send_header('Content-Length', '0')
self.end_headers()
self.wfile.write(b'<html><video src="/vid.mp4" /></html>')
else:
assert False
self._status(404)
def send_header(self, keyword, value):
"""
Forcibly allow HTTP server to send non percent-encoded non-ASCII characters in headers.
This is against what is defined in RFC 3986, however we need to test we support this
since some sites incorrectly do this.
"""
if keyword.lower() == 'connection':
return super().send_header(keyword, value)
if not hasattr(self, '_headers_buffer'):
self._headers_buffer = []
self._headers_buffer.append(f'{keyword}: {value}\r\n'.encode())
class FakeLogger:
@ -56,36 +179,128 @@ class FakeLogger:
class TestHTTP(unittest.TestCase):
def setUp(self):
self.httpd = http.server.HTTPServer(
# HTTP server
self.http_httpd = http.server.ThreadingHTTPServer(
('127.0.0.1', 0), HTTPTestRequestHandler)
self.port = http_server_port(self.httpd)
self.server_thread = threading.Thread(target=self.httpd.serve_forever)
self.server_thread.daemon = True
self.server_thread.start()
self.http_port = http_server_port(self.http_httpd)
self.http_server_thread = threading.Thread(target=self.http_httpd.serve_forever)
# FIXME: we should probably stop the http server thread after each test
# See: https://github.com/yt-dlp/yt-dlp/pull/7094#discussion_r1199746041
self.http_server_thread.daemon = True
self.http_server_thread.start()
class TestHTTPS(unittest.TestCase):
def setUp(self):
# HTTPS server
certfn = os.path.join(TEST_DIR, 'testcert.pem')
self.httpd = http.server.HTTPServer(
self.https_httpd = http.server.ThreadingHTTPServer(
('127.0.0.1', 0), HTTPTestRequestHandler)
sslctx = ssl.SSLContext(ssl.PROTOCOL_TLS_SERVER)
sslctx.load_cert_chain(certfn, None)
self.httpd.socket = sslctx.wrap_socket(self.httpd.socket, server_side=True)
self.port = http_server_port(self.httpd)
self.server_thread = threading.Thread(target=self.httpd.serve_forever)
self.server_thread.daemon = True
self.server_thread.start()
self.https_httpd.socket = sslctx.wrap_socket(self.https_httpd.socket, server_side=True)
self.https_port = http_server_port(self.https_httpd)
self.https_server_thread = threading.Thread(target=self.https_httpd.serve_forever)
self.https_server_thread.daemon = True
self.https_server_thread.start()
def test_nocheckcertificate(self):
ydl = YoutubeDL({'logger': FakeLogger()})
self.assertRaises(
Exception,
ydl.extract_info, 'https://127.0.0.1:%d/video.html' % self.port)
with FakeYDL({'logger': FakeLogger()}) as ydl:
with self.assertRaises(urllib.error.URLError):
ydl.urlopen(sanitized_Request(f'https://127.0.0.1:{self.https_port}/headers'))
ydl = YoutubeDL({'logger': FakeLogger(), 'nocheckcertificate': True})
r = ydl.extract_info('https://127.0.0.1:%d/video.html' % self.port)
self.assertEqual(r['url'], 'https://127.0.0.1:%d/vid.mp4' % self.port)
with FakeYDL({'logger': FakeLogger(), 'nocheckcertificate': True}) as ydl:
r = ydl.urlopen(sanitized_Request(f'https://127.0.0.1:{self.https_port}/headers'))
self.assertEqual(r.status, 200)
r.close()
def test_percent_encode(self):
with FakeYDL() as ydl:
# Unicode characters should be encoded with uppercase percent-encoding
res = ydl.urlopen(sanitized_Request(f'http://127.0.0.1:{self.http_port}/中文.html'))
self.assertEqual(res.status, 200)
res.close()
# don't normalize existing percent encodings
res = ydl.urlopen(sanitized_Request(f'http://127.0.0.1:{self.http_port}/%c7%9f'))
self.assertEqual(res.status, 200)
res.close()
def test_unicode_path_redirection(self):
with FakeYDL() as ydl:
r = ydl.urlopen(sanitized_Request(f'http://127.0.0.1:{self.http_port}/302-non-ascii-redirect'))
self.assertEqual(r.url, f'http://127.0.0.1:{self.http_port}/%E4%B8%AD%E6%96%87.html')
r.close()
def test_redirect(self):
with FakeYDL() as ydl:
def do_req(redirect_status, method):
data = b'testdata' if method in ('POST', 'PUT') else None
res = ydl.urlopen(sanitized_Request(
f'http://127.0.0.1:{self.http_port}/redirect_{redirect_status}', method=method, data=data))
return res.read().decode('utf-8'), res.headers.get('method', '')
# A 303 must either use GET or HEAD for subsequent request
self.assertEqual(do_req(303, 'POST'), ('', 'GET'))
self.assertEqual(do_req(303, 'HEAD'), ('', 'HEAD'))
self.assertEqual(do_req(303, 'PUT'), ('', 'GET'))
# 301 and 302 turn POST only into a GET
self.assertEqual(do_req(301, 'POST'), ('', 'GET'))
self.assertEqual(do_req(301, 'HEAD'), ('', 'HEAD'))
self.assertEqual(do_req(302, 'POST'), ('', 'GET'))
self.assertEqual(do_req(302, 'HEAD'), ('', 'HEAD'))
self.assertEqual(do_req(301, 'PUT'), ('testdata', 'PUT'))
self.assertEqual(do_req(302, 'PUT'), ('testdata', 'PUT'))
# 307 and 308 should not change method
for m in ('POST', 'PUT'):
self.assertEqual(do_req(307, m), ('testdata', m))
self.assertEqual(do_req(308, m), ('testdata', m))
self.assertEqual(do_req(307, 'HEAD'), ('', 'HEAD'))
self.assertEqual(do_req(308, 'HEAD'), ('', 'HEAD'))
# These should not redirect and instead raise an HTTPError
for code in (300, 304, 305, 306):
with self.assertRaises(urllib.error.HTTPError):
do_req(code, 'GET')
def test_content_type(self):
# https://github.com/yt-dlp/yt-dlp/commit/379a4f161d4ad3e40932dcf5aca6e6fb9715ab28
with FakeYDL({'nocheckcertificate': True}) as ydl:
# method should be auto-detected as POST
r = sanitized_Request(f'https://localhost:{self.https_port}/headers', data=urlencode_postdata({'test': 'test'}))
headers = ydl.urlopen(r).read().decode('utf-8')
self.assertIn('Content-Type: application/x-www-form-urlencoded', headers)
# test http
r = sanitized_Request(f'http://localhost:{self.http_port}/headers', data=urlencode_postdata({'test': 'test'}))
headers = ydl.urlopen(r).read().decode('utf-8')
self.assertIn('Content-Type: application/x-www-form-urlencoded', headers)
def test_cookiejar(self):
with FakeYDL() as ydl:
ydl.cookiejar.set_cookie(http.cookiejar.Cookie(
0, 'test', 'ytdlp', None, False, '127.0.0.1', True,
False, '/headers', True, False, None, False, None, None, {}))
data = ydl.urlopen(sanitized_Request(f'http://127.0.0.1:{self.http_port}/headers')).read()
self.assertIn(b'Cookie: test=ytdlp', data)
def test_no_compression_compat_header(self):
with FakeYDL() as ydl:
data = ydl.urlopen(
sanitized_Request(
f'http://127.0.0.1:{self.http_port}/headers',
headers={'Youtubedl-no-compression': True})).read()
self.assertIn(b'Accept-Encoding: identity', data)
self.assertNotIn(b'youtubedl-no-compression', data.lower())
def test_gzip_trailing_garbage(self):
# https://github.com/ytdl-org/youtube-dl/commit/aa3e950764337ef9800c936f4de89b31c00dfcf5
# https://github.com/ytdl-org/youtube-dl/commit/6f2ec15cee79d35dba065677cad9da7491ec6e6f
with FakeYDL() as ydl:
data = ydl.urlopen(sanitized_Request(f'http://localhost:{self.http_port}/trailing_garbage')).read().decode('utf-8')
self.assertEqual(data, '<html><video src="/vid.mp4" /></html>')
class TestClientCert(unittest.TestCase):
@ -112,8 +327,8 @@ class TestClientCert(unittest.TestCase):
'nocheckcertificate': True,
**params,
})
r = ydl.extract_info('https://127.0.0.1:%d/video.html' % self.port)
self.assertEqual(r['url'], 'https://127.0.0.1:%d/vid.mp4' % self.port)
r = ydl.extract_info(f'https://127.0.0.1:{self.port}/video.html')
self.assertEqual(r['url'], f'https://127.0.0.1:{self.port}/vid.mp4')
def test_certificate_combined_nopass(self):
self._run_test(client_certificate=os.path.join(self.certdir, 'clientwithkey.crt'))
@ -188,5 +403,22 @@ class TestProxy(unittest.TestCase):
self.assertEqual(response, 'normal: http://xn--fiq228c.tw/')
class TestFileURL(unittest.TestCase):
# See https://github.com/ytdl-org/youtube-dl/issues/8227
def test_file_urls(self):
tf = tempfile.NamedTemporaryFile(delete=False)
tf.write(b'foobar')
tf.close()
url = pathlib.Path(tf.name).as_uri()
with FakeYDL() as ydl:
self.assertRaisesRegex(
urllib.error.URLError, 'file:// URLs are explicitly disabled in yt-dlp for security reasons', ydl.urlopen, url)
with FakeYDL({'enable_file_urls': True}) as ydl:
res = ydl.urlopen(url)
self.assertEqual(res.read(), b'foobar')
res.close()
os.unlink(tf.name)
if __name__ == '__main__':
unittest.main()

View file

@ -1664,61 +1664,44 @@ class YoutubeDLRedirectHandler(urllib.request.HTTPRedirectHandler):
The code is based on HTTPRedirectHandler implementation from CPython [1].
This redirect handler solves two issues:
- ensures redirect URL is always unicode under python 2
- introduces support for experimental HTTP response status code
308 Permanent Redirect [2] used by some sites [3]
This redirect handler fixes and improves the logic to better align with RFC7261
and what browsers tend to do [2][3]
1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
3. https://github.com/ytdl-org/youtube-dl/issues/28768
2. https://datatracker.ietf.org/doc/html/rfc7231
3. https://github.com/python/cpython/issues/91306
"""
http_error_301 = http_error_303 = http_error_307 = http_error_308 = urllib.request.HTTPRedirectHandler.http_error_302
def redirect_request(self, req, fp, code, msg, headers, newurl):
"""Return a Request or None in response to a redirect.
This is called by the http_error_30x methods when a
redirection response is received. If a redirection should
take place, return a new Request to allow http_error_30x to
perform the redirect. Otherwise, raise HTTPError if no-one
else should try to handle this url. Return None if you can't
but another Handler might.
"""
m = req.get_method()
if (not (code in (301, 302, 303, 307, 308) and m in ("GET", "HEAD")
or code in (301, 302, 303) and m == "POST")):
if code not in (301, 302, 303, 307, 308):
raise urllib.error.HTTPError(req.full_url, code, msg, headers, fp)
# Strictly (according to RFC 2616), 301 or 302 in response to
# a POST MUST NOT cause a redirection without confirmation
# from the user (of urllib.request, in this case). In practice,
# essentially all clients do redirect in this case, so we do
# the same.
# Be conciliant with URIs containing a space. This is mainly
# redundant with the more complete encoding done in http_error_302(),
# but it is kept for compatibility with other callers.
newurl = newurl.replace(' ', '%20')
CONTENT_HEADERS = ("content-length", "content-type")
# NB: don't use dict comprehension for python 2.6 compatibility
newheaders = {k: v for k, v in req.headers.items() if k.lower() not in CONTENT_HEADERS}
new_method = req.get_method()
new_data = req.data
remove_headers = []
# A 303 must either use GET or HEAD for subsequent request
# https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.4
if code == 303 and m != 'HEAD':
m = 'GET'
if code == 303 and req.get_method() != 'HEAD':
new_method = 'GET'
# 301 and 302 redirects are commonly turned into a GET from a POST
# for subsequent requests by browsers, so we'll do the same.
# https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.2
# https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.3
if code in (301, 302) and m == 'POST':
m = 'GET'
elif code in (301, 302) and req.get_method() == 'POST':
new_method = 'GET'
# only remove payload if method changed (e.g. POST to GET)
if new_method != req.get_method():
new_data = None
remove_headers.extend(['Content-Length', 'Content-Type'])
new_headers = {k: v for k, v in req.headers.items() if k.lower() not in remove_headers}
return urllib.request.Request(
newurl, headers=newheaders, origin_req_host=req.origin_req_host,
unverifiable=True, method=m)
newurl, headers=new_headers, origin_req_host=req.origin_req_host,
unverifiable=True, method=new_method, data=new_data)
def extract_timezone(date_str):