diff options
author | Patrick Cloke <clokep@users.noreply.github.com> | 2021-10-08 14:14:42 -0400 |
---|---|---|
committer | GitHub <noreply@github.com> | 2021-10-08 14:14:42 -0400 |
commit | 1b112840d2c6dafa131eba4f0285409bb7345661 (patch) | |
tree | cead0015dbaceb5a188b3b95e8a6cfe2d82960dc /tests/rest/media/v1 | |
parent | Revert accidental push to develop. (diff) | |
download | synapse-1b112840d2c6dafa131eba4f0285409bb7345661.tar.xz |
Autodiscover oEmbed endpoint from returned HTML (#10822)
Searches the returned HTML for an oEmbed endpoint using the autodiscovery mechanism (`<link rel=...>`), and will request it to generate the preview.
Diffstat (limited to 'tests/rest/media/v1')
-rw-r--r-- | tests/rest/media/v1/test_url_preview.py | 100 |
1 files changed, 99 insertions, 1 deletions
diff --git a/tests/rest/media/v1/test_url_preview.py b/tests/rest/media/v1/test_url_preview.py index ce43de780b..8698135a76 100644 --- a/tests/rest/media/v1/test_url_preview.py +++ b/tests/rest/media/v1/test_url_preview.py @@ -725,9 +725,107 @@ class URLPreviewTests(unittest.HomeserverTestCase): }, ) + def test_oembed_autodiscovery(self): + """ + 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.assertTrue(body["og:image"].startswith("mxc://")) + self.assertEqual(body["og:image:height"], 1) + self.assertEqual(body["og:image:width"], 1) + self.assertEqual(body["og:image:type"], "image/png") + def _download_image(self): """Downloads an image into the URL cache. - Returns: A (host, media_id) tuple representing the MXC URI of the image. """ |