def test_blacklisted_ip_range_whitelisted_ip(self):
"""
Blacklisted but then subsequently whitelisted IP addresses can be
spidered.
"""
self.lookups["example.com"] = [(IPv4Address, "1.1.1.1")]
request, channel = self.make_request(
"GET", "url_preview?url=http://example.com", shorthand=False
)
request.render(self.preview_url)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\nContent-Type: text/html\r\n\r\n"
% (len(self.end_content),)
+ self.end_content
)
self.pump()
self.assertEqual(channel.code, 200)
self.assertEqual(
channel.json_body, {"og:title": "~matrix~", "og:description": "hi"}
)
def test_non_ascii_preview_content_type(self):
self.lookups["matrix.org"] = [(IPv4Address, "8.8.8.8")]
end_content = (
b'<html><head>'
b'<meta property="og:title" content="\xe4\xea\xe0" />'
b'<meta property="og:description" content="hi" />'
b'</head></html>'
)
request, channel = self.make_request(
"GET", "url_preview?url=http://matrix.org", shorthand=False
)
request.render(self.preview_url)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
(
b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b"Content-Type: text/html; charset=\"windows-1251\"\r\n\r\n"
)
% (len(end_content),)
+ end_content
)
self.pump()
self.assertEqual(channel.code, 200)
self.assertEqual(channel.json_body["og:title"], u"\u0434\u043a\u0430")
def _download_image(self) -> Tuple[str, str]:
"""Downloads an image into the URL cache.
Returns:
A (host, media_id) tuple representing the MXC URI of the image.
"""
self.lookups["cdn.twitter.com"] = [(IPv4Address, "10.1.2.3")]
channel = self.make_request(
"GET",
"preview_url?url=http://cdn.twitter.com/matrixdotorg",
shorthand=False,
await_result=False,
)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\nContent-Type: image/png\r\n\r\n"
% (len(SMALL_PNG), ) + SMALL_PNG)
self.pump()
self.assertEqual(channel.code, 200)
body = channel.json_body
mxc_uri = body["og:image"]
host, _port, media_id = parse_and_validate_mxc_uri(mxc_uri)
self.assertIsNone(_port)
return host, media_id
def test_nonexistent_image(self) -> None:
"""If the preview image doesn't exist, ensure some data is returned."""
self.lookups["matrix.org"] = [(IPv4Address, "10.1.2.3")]
end_content = (
b"""<html><body><img src="http://cdn.matrix.org/foo.jpg"></body></html>"""
)
channel = self.make_request(
"GET",
"preview_url?url=http://matrix.org",
shorthand=False,
await_result=False,
)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
(b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b'Content-Type: text/html; charset="utf8"\r\n\r\n') %
(len(end_content), ) + end_content)
self.pump()
self.assertEqual(channel.code, 200)
# The image should not be in the result.
self.assertNotIn("og:image", channel.json_body)
def test_non_ascii_preview_content_type(self):
self.lookups["matrix.org"] = [(IPv4Address, "10.1.2.3")]
end_content = (
b"<html><head>"
b'<meta property="og:title" content="\xe4\xea\xe0" />'
b'<meta property="og:description" content="hi" />'
b"</head></html>"
)
channel = self.make_request(
"GET",
"preview_url?url=http://matrix.org",
shorthand=False,
await_result=False,
)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
(
b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b'Content-Type: text/html; charset="windows-1251"\r\n\r\n'
)
% (len(end_content),)
+ end_content
)
self.pump()
self.assertEqual(channel.code, 200)
self.assertEqual(channel.json_body["og:title"], "\u0434\u043a\u0430")
def test_inline_data_url(self) -> None:
"""
An inline image (as a data URL) should be parsed properly.
"""
self.lookups["matrix.org"] = [(IPv4Address, "10.1.2.3")]
data = base64.b64encode(SMALL_PNG)
end_content = (b"<html><head>"
b'<img src="data:image/png;base64,%s" />'
b"</head></html>") % (data, )
channel = self.make_request(
"GET",
"preview_url?url=http://matrix.org",
shorthand=False,
await_result=False,
)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
(b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b'Content-Type: text/html; charset="utf8"\r\n\r\n') %
(len(end_content), ) + end_content)
self.pump()
self.assertEqual(channel.code, 200)
self._assert_small_png(channel.json_body)
def test_audio_rejected(self) -> None:
self.lookups["matrix.org"] = [(IPv4Address, "10.1.2.3")]
end_content = b"anything"
channel = self.make_request(
"GET",
"preview_url?url=http://matrix.org",
shorthand=False,
await_result=False,
)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived((b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b"Content-Type: audio/aac\r\n\r\n") %
(len(end_content)) + end_content)
self.pump()
self.assertEqual(channel.code, 502)
self.assertEqual(
channel.json_body,
{
"errcode":
"M_UNKNOWN",
"error":
"Requested file's content type not allowed for this operation: audio/aac",
},
)
def test_ipaddr(self):
"""
IP addresses can be previewed directly.
"""
self.lookups["example.com"] = [(IPv4Address, "8.8.8.8")]
request, channel = self.make_request(
"GET", "url_preview?url=http://example.com", shorthand=False)
request.render(self.preview_url)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\nContent-Type: text/html\r\n\r\n"
% (len(self.end_content), ) + self.end_content)
self.pump()
self.assertEqual(channel.code, 200)
self.assertEqual(channel.json_body, {
"og:title": "~matrix~",
"og:description": "hi"
})
def test_blacklisted_ip_range_whitelisted_ip(self):
"""
Blacklisted but then subsequently whitelisted IP addresses can be
spidered.
"""
self.lookups["example.com"] = [(IPv4Address, "1.1.1.1")]
channel = self.make_request(
"GET",
"preview_url?url=http://example.com",
shorthand=False,
await_result=False,
)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\nContent-Type: text/html\r\n\r\n"
% (len(self.end_content),)
+ self.end_content
)
self.pump()
self.assertEqual(channel.code, 200)
self.assertEqual(
channel.json_body, {"og:title": "~matrix~", "og:description": "hi"}
)
def test_overlong_title(self):
self.lookups["matrix.org"] = [(IPv4Address, "10.1.2.3")]
end_content = (b"<html><head>"
b"<title>" + b"x" * 2000 + b"</title>"
b'<meta property="og:description" content="hi" />'
b"</head></html>")
channel = self.make_request(
"GET",
"preview_url?url=http://matrix.org",
shorthand=False,
await_result=False,
)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
(b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b'Content-Type: text/html; charset="windows-1251"\r\n\r\n') %
(len(end_content), ) + end_content)
self.pump()
self.assertEqual(channel.code, 200)
res = channel.json_body
# We should only see the `og:description` field, as `title` is too long and should be stripped out
self.assertCountEqual(["og:description"], res.keys())
def test_non_ascii_preview_httpequiv(self):
self.lookups["matrix.org"] = [(IPv4Address, "8.8.8.8")]
end_content = (
b'<html><head>'
b'<meta http-equiv="Content-Type" content="text/html; charset=windows-1251"/>'
b'<meta property="og:title" content="\xe4\xea\xe0" />'
b'<meta property="og:description" content="hi" />'
b'</head></html>')
request, channel = self.make_request(
"GET", "url_preview?url=http://matrix.org", shorthand=False)
request.render(self.preview_url)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
(b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b"Content-Type: text/html; charset=\"utf8\"\r\n\r\n") %
(len(end_content), ) + end_content)
self.pump()
self.assertEqual(channel.code, 200)
self.assertEqual(channel.json_body["og:title"], u"\u0434\u043a\u0430")
def test_cache_returns_correct_type(self):
self.lookups["matrix.org"] = [(IPv4Address, "8.8.8.8")]
request, channel = self.make_request(
"GET", "url_preview?url=http://matrix.org", shorthand=False
)
request.render(self.preview_url)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\nContent-Type: text/html\r\n\r\n"
% (len(self.end_content),)
+ self.end_content
)
self.pump()
self.assertEqual(channel.code, 200)
self.assertEqual(
channel.json_body, {"og:title": "~matrix~", "og:description": "hi"}
)
# Check the cache returns the correct response
request, channel = self.make_request(
"GET", "url_preview?url=http://matrix.org", shorthand=False
)
request.render(self.preview_url)
self.pump()
# Check the cache response has the same content
self.assertEqual(channel.code, 200)
self.assertEqual(
channel.json_body, {"og:title": "~matrix~", "og:description": "hi"}
)
# Clear the in-memory cache
self.assertIn("http://matrix.org", self.preview_url._cache)
self.preview_url._cache.pop("http://matrix.org")
self.assertNotIn("http://matrix.org", self.preview_url._cache)
# Check the database cache returns the correct response
request, channel = self.make_request(
"GET", "url_preview?url=http://matrix.org", shorthand=False
)
request.render(self.preview_url)
self.pump()
# Check the cache response has the same content
self.assertEqual(channel.code, 200)
self.assertEqual(
channel.json_body, {"og:title": "~matrix~", "og:description": "hi"}
)
def test_cache_returns_correct_type(self):
self.lookups["matrix.org"] = [(IPv4Address, "8.8.8.8")]
request, channel = self.make_request(
"GET", "url_preview?url=http://matrix.org", shorthand=False
)
request.render(self.preview_url)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\nContent-Type: text/html\r\n\r\n"
% (len(self.end_content),)
+ self.end_content
)
self.pump()
self.assertEqual(channel.code, 200)
self.assertEqual(
channel.json_body, {"og:title": "~matrix~", "og:description": "hi"}
)
# Check the cache returns the correct response
request, channel = self.make_request(
"GET", "url_preview?url=http://matrix.org", shorthand=False
)
request.render(self.preview_url)
self.pump()
# Check the cache response has the same content
self.assertEqual(channel.code, 200)
self.assertEqual(
channel.json_body, {"og:title": "~matrix~", "og:description": "hi"}
)
# Clear the in-memory cache
self.assertIn("http://matrix.org", self.preview_url._cache)
self.preview_url._cache.pop("http://matrix.org")
self.assertNotIn("http://matrix.org", self.preview_url._cache)
# Check the database cache returns the correct response
request, channel = self.make_request(
"GET", "url_preview?url=http://matrix.org", shorthand=False
)
request.render(self.preview_url)
self.pump()
# Check the cache response has the same content
self.assertEqual(channel.code, 200)
self.assertEqual(
channel.json_body, {"og:title": "~matrix~", "og:description": "hi"}
)
def connect_logging_client(reactor, client_id):
# This is essentially tests.server.connect_client, but disabling autoflush on
# the client transport. This is necessary to avoid an infinite loop due to
# sending of data via the logging transport causing additional logs to be
# written.
factory = reactor.tcpClients.pop(client_id)[2]
client = factory.buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, reactor))
client.makeConnection(FakeTransport(server, reactor, autoflush=False))
return client, server
def test_accept_language_config_option(self):
"""
Accept-Language header is sent to the remote server
"""
self.lookups["example.com"] = [(IPv4Address, "10.1.2.3")]
# Build and make a request to the server
channel = self.make_request(
"GET",
"preview_url?url=http://example.com",
shorthand=False,
await_result=False,
)
self.pump()
# Extract Synapse's tcp client
client = self.reactor.tcpClients[0][2].buildProtocol(None)
# Build a fake remote server to reply with
server = AccumulatingProtocol()
# Connect the two together
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
# Tell Synapse that it has received some data from the remote server
client.dataReceived(
b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\nContent-Type: text/html\r\n\r\n"
% (len(self.end_content),)
+ self.end_content
)
# Move the reactor along until we get a response on our original channel
self.pump()
self.assertEqual(channel.code, 200)
self.assertEqual(
channel.json_body, {"og:title": "~matrix~", "og:description": "hi"}
)
# Check that the server received the Accept-Language header as part
# of the request from Synapse
self.assertIn(
(
b"Accept-Language: en-UK\r\n"
b"Accept-Language: en-US;q=0.9\r\n"
b"Accept-Language: fr;q=0.8\r\n"
b"Accept-Language: *;q=0.7"
),
server.data,
)
def test_oembed_photo(self) -> None:
"""Test an oEmbed endpoint which returns a 'photo' type which redirects the preview to a new URL."""
self.lookups["publish.twitter.com"] = [(IPv4Address, "10.1.2.3")]
self.lookups["cdn.twitter.com"] = [(IPv4Address, "10.1.2.3")]
result = {
"version": "1.0",
"type": "photo",
"url": "http://cdn.twitter.com/matrixdotorg",
}
oembed_content = json.dumps(result).encode("utf-8")
channel = self.make_request(
"GET",
"preview_url?url=http://twitter.com/matrixdotorg/status/12345",
shorthand=False,
await_result=False,
)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
(b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b'Content-Type: application/json; charset="utf8"\r\n\r\n') %
(len(oembed_content), ) + oembed_content)
self.pump()
# Ensure a second request is made to the photo URL.
client = self.reactor.tcpClients[1][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived((b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b"Content-Type: image/png\r\n\r\n") %
(len(SMALL_PNG), ) + SMALL_PNG)
self.pump()
# Ensure the URL is what was requested.
self.assertIn(b"/matrixdotorg", server.data)
self.assertEqual(channel.code, 200)
body = channel.json_body
self.assertEqual(body["og:url"],
"http://twitter.com/matrixdotorg/status/12345")
self._assert_small_png(body)
def test_oembed_rich(self):
"""Test an oEmbed endpoint which returns HTML content via the 'rich' type."""
# Route the HTTP version to an HTTP endpoint so that the tests work.
with patch.dict(
"synapse.rest.media.v1.preview_url_resource._oembed_patterns",
{
re.compile(
r"http://twitter\.com/.+/status/.+"
): "http://publish.twitter.com/oembed",
},
clear=True,
):
self.lookups["publish.twitter.com"] = [(IPv4Address, "10.1.2.3")]
result = {
"version": "1.0",
"type": "rich",
"html": "<div>Content Preview</div>",
}
end_content = json.dumps(result).encode("utf-8")
channel = self.make_request(
"GET",
"preview_url?url=http://twitter.com/matrixdotorg/status/12345",
shorthand=False,
await_result=False,
)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
(
b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b'Content-Type: application/json; charset="utf8"\r\n\r\n'
)
% (len(end_content),)
+ end_content
)
self.pump()
self.assertEqual(channel.code, 200)
self.assertEqual(
channel.json_body,
{"og:title": None, "og:description": "Content Preview"},
)
def connect_client(reactor: IReactorTCP, client_id: int) -> AccumulatingProtocol:
"""
Connect a client to a fake TCP transport.
Args:
reactor
factory: The connecting factory to build.
"""
factory = reactor.tcpClients.pop(client_id)[2]
client = factory.buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, reactor))
client.makeConnection(FakeTransport(server, reactor))
return client, server
def test_oembed_format(self):
"""Test an oEmbed endpoint which requires the format in the URL."""
self.lookups["www.hulu.com"] = [(IPv4Address, "10.1.2.3")]
result = {
"version": "1.0",
"type": "rich",
"html": "<div>Content Preview</div>",
}
end_content = json.dumps(result).encode("utf-8")
channel = self.make_request(
"GET",
"preview_url?url=http://www.hulu.com/watch/12345",
shorthand=False,
await_result=False,
)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
(
b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b'Content-Type: application/json; charset="utf8"\r\n\r\n'
)
% (len(end_content),)
+ end_content
)
self.pump()
# The {format} should have been turned into json.
self.assertIn(b"/api/oembed.json", server.data)
# A URL parameter of format=json should be provided.
self.assertIn(b"format=json", server.data)
self.assertEqual(channel.code, 200)
body = channel.json_body
self.assertEqual(
body,
{
"og:url": "http://www.hulu.com/watch/12345",
"og:description": "Content Preview",
},
)
def test_lose_connection(self):
"""
We log the URI correctly redacted when we lose the connection.
"""
class HangingResource(Resource):
"""
A Resource that strategically hangs, as if it were processing an
answer.
"""
def render(self, request):
return NOT_DONE_YET
# Set up a logging handler that we can inspect afterwards
output = StringIO()
handler = logging.StreamHandler(output)
logger.addHandler(handler)
old_level = logger.level
logger.setLevel(10)
self.addCleanup(logger.setLevel, old_level)
self.addCleanup(logger.removeHandler, handler)
# Make a resource and a Site, the resource will hang and allow us to
# time out the request while it's 'processing'
base_resource = Resource()
base_resource.putChild(b'', HangingResource())
site = SynapseSite("test", "site_tag", {}, base_resource, "1.0")
server = site.buildProtocol(None)
client = AccumulatingProtocol()
client.makeConnection(FakeTransport(server, self.reactor))
server.makeConnection(FakeTransport(client, self.reactor))
# Send a request with an access token that will get redacted
server.dataReceived(b"GET /?access_token=bar HTTP/1.0\r\n\r\n")
self.pump()
# Lose the connection
e = Failure(Exception("Failed123"))
server.connectionLost(e)
handler.flush()
# Our access token is redacted and the failure reason is logged.
self.assertIn("/?access_token=<redacted>", output.getvalue())
self.assertIn("Failed123", output.getvalue())
def test_oembed_rich(self):
"""Test an oEmbed endpoint which returns HTML content via the 'rich' type."""
self.lookups["publish.twitter.com"] = [(IPv4Address, "10.1.2.3")]
result = {
"version": "1.0",
"type": "rich",
# Note that this provides the author, not the title.
"author_name": "Alice",
"html": "<div>Content Preview</div>",
}
end_content = json.dumps(result).encode("utf-8")
channel = self.make_request(
"GET",
"preview_url?url=http://twitter.com/matrixdotorg/status/12345",
shorthand=False,
await_result=False,
)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
(
b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b'Content-Type: application/json; charset="utf8"\r\n\r\n'
)
% (len(end_content),)
+ end_content
)
self.pump()
self.assertEqual(channel.code, 200)
body = channel.json_body
self.assertEqual(
body,
{
"og:url": "http://twitter.com/matrixdotorg/status/12345",
"og:title": "Alice",
"og:description": "Content Preview",
},
)
def test_reactor(self):
"""Apply the blacklisting reactor and ensure it properly blocks connections to particular domains and IPs."""
agent = Agent(
BlacklistingReactorWrapper(
self.reactor,
ip_whitelist=self.ip_whitelist,
ip_blacklist=self.ip_blacklist,
),
)
# The unsafe domains and IPs should be rejected.
for domain in (self.unsafe_domain, self.unsafe_ip):
self.failureResultOf(
agent.request(b"GET", b"http://" + domain), DNSLookupError
)
# The safe domains IPs should be accepted.
for domain in (
self.safe_domain,
self.allowed_domain,
self.safe_ip,
self.allowed_ip,
):
d = agent.request(b"GET", b"http://" + domain)
# Grab the latest TCP connection.
(
host,
port,
client_factory,
_timeout,
_bindAddress,
) = self.reactor.tcpClients[-1]
# Make the connection and pump data through it.
client = client_factory.buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
b"HTTP/1.0 200 OK\r\nContent-Length: 0\r\nContent-Type: text/html\r\n\r\n"
)
response = self.successResultOf(d)
self.assertEqual(response.code, 200)
def test_blacklist_port(self) -> None:
"""Tests that blacklisting URLs with a port makes previewing such URLs
fail with a 403 error and doesn't impact other previews.
"""
self.lookups["matrix.org"] = [(IPv4Address, "10.1.2.3")]
bad_url = quote("http://matrix.org:8888/foo")
good_url = quote("http://matrix.org/foo")
channel = self.make_request(
"GET",
"preview_url?url=" + bad_url,
shorthand=False,
await_result=False,
)
self.pump()
self.assertEqual(channel.code, 403, channel.result)
channel = self.make_request(
"GET",
"preview_url?url=" + good_url,
shorthand=False,
await_result=False,
)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\nContent-Type: text/html\r\n\r\n"
% (len(self.end_content), ) + self.end_content)
self.pump()
self.assertEqual(channel.code, 200)
def test_oembed_photo(self):
"""Test an oEmbed endpoint which returns a 'photo' type which redirects the preview to a new URL."""
# Route the HTTP version to an HTTP endpoint so that the tests work.
with patch.dict(
"synapse.rest.media.v1.preview_url_resource._oembed_patterns",
{
re.compile(
r"http://twitter\.com/.+/status/.+"
): "http://publish.twitter.com/oembed",
},
clear=True,
):
self.lookups["publish.twitter.com"] = [(IPv4Address, "10.1.2.3")]
self.lookups["cdn.twitter.com"] = [(IPv4Address, "10.1.2.3")]
result = {
"version": "1.0",
"type": "photo",
"url": "http://cdn.twitter.com/matrixdotorg",
}
oembed_content = json.dumps(result).encode("utf-8")
end_content = (
b"<html><head>"
b"<title>Some Title</title>"
b'<meta property="og:description" content="hi" />'
b"</head></html>"
)
channel = self.make_request(
"GET",
"preview_url?url=http://twitter.com/matrixdotorg/status/12345",
shorthand=False,
await_result=False,
)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
(
b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b'Content-Type: application/json; charset="utf8"\r\n\r\n'
)
% (len(oembed_content),)
+ oembed_content
)
self.pump()
client = self.reactor.tcpClients[1][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
(
b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b'Content-Type: text/html; charset="utf8"\r\n\r\n'
)
% (len(end_content),)
+ end_content
)
self.pump()
self.assertEqual(channel.code, 200)
self.assertEqual(
channel.json_body, {"og:title": "Some Title", "og:description": "hi"}
)
def test_oembed_autodiscovery(self) -> None:
"""
Autodiscovery works by finding the link in the HTML response and then requesting an oEmbed URL.
1. Request a preview of a URL which is not known to the oEmbed code.
2. It returns HTML including a link to an oEmbed preview.
3. The oEmbed preview is requested and returns a URL for an image.
4. The image is requested for thumbnailing.
"""
# This is a little cheesy in that we use the www subdomain (which isn't the
# list of oEmbed patterns) to get "raw" HTML response.
self.lookups["www.twitter.com"] = [(IPv4Address, "10.1.2.3")]
self.lookups["publish.twitter.com"] = [(IPv4Address, "10.1.2.3")]
self.lookups["cdn.twitter.com"] = [(IPv4Address, "10.1.2.3")]
result = b"""
<link rel="alternate" type="application/json+oembed"
href="http://publish.twitter.com/oembed?url=http%3A%2F%2Fcdn.twitter.com%2Fmatrixdotorg%2Fstatus%2F12345&format=json"
title="matrixdotorg" />
"""
channel = self.make_request(
"GET",
"preview_url?url=http://www.twitter.com/matrixdotorg/status/12345",
shorthand=False,
await_result=False,
)
self.pump()
client = self.reactor.tcpClients[0][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
(b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b'Content-Type: text/html; charset="utf8"\r\n\r\n') %
(len(result), ) + result)
self.pump()
# The oEmbed response.
result2 = {
"version": "1.0",
"type": "photo",
"url": "http://cdn.twitter.com/matrixdotorg",
}
oembed_content = json.dumps(result2).encode("utf-8")
# Ensure a second request is made to the oEmbed URL.
client = self.reactor.tcpClients[1][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived(
(b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b'Content-Type: application/json; charset="utf8"\r\n\r\n') %
(len(oembed_content), ) + oembed_content)
self.pump()
# Ensure the URL is what was requested.
self.assertIn(b"/oembed?", server.data)
# Ensure a third request is made to the photo URL.
client = self.reactor.tcpClients[2][2].buildProtocol(None)
server = AccumulatingProtocol()
server.makeConnection(FakeTransport(client, self.reactor))
client.makeConnection(FakeTransport(server, self.reactor))
client.dataReceived((b"HTTP/1.0 200 OK\r\nContent-Length: %d\r\n"
b"Content-Type: image/png\r\n\r\n") %
(len(SMALL_PNG), ) + SMALL_PNG)
self.pump()
# Ensure the URL is what was requested.
self.assertIn(b"/matrixdotorg", server.data)
self.assertEqual(channel.code, 200)
body = channel.json_body
self.assertEqual(body["og:url"],
"http://www.twitter.com/matrixdotorg/status/12345")
self._assert_small_png(body)
