Skip to content
This repository has been archived by the owner on Apr 26, 2024. It is now read-only.

Uniformize spam-checker API, part 5: expand other spam-checker callbacks to return Tuple[Codes, dict] #13044

Merged
merged 26 commits into from
Jul 11, 2022

Conversation

Yoric
Copy link
Contributor

@Yoric Yoric commented Jun 14, 2022

The objective of this push is to progressively expand the spam-checker API into letting spam-checkers provide reasons for which an event/operation has been rejected. In particular, we wish to prototype account suspension (by opposition to account removal) through spam-checkers and let users know what they need to do to have the account unsuspended.

Specifically, letting a spam-checker return an additional dict gives us a way to point users towards an individual url on which they can read up all the details about their suspension.

This is a followup to:

Pull Request Checklist

  • Pull request is based on the develop branch
  • Pull request includes a changelog file. The entry should:
    • Be a short description of your change which makes sense to users. "Fixed a bug that prevented receiving messages from other servers." instead of "Moved X method from EventStore to EventWorkerStore.".
    • Use markdown where necessary, mostly for code blocks.
    • End with either a period (.) or an exclamation mark (!).
    • Start with a capital letter.
    • Feel free to credit yourself, by adding a sentence "Contributed by @github_username." or "Contributed by [Your Name]." to the end of the entry.
  • Pull request includes a sign off
  • Code style is correct
    (run the linters)

@Yoric Yoric requested a review from a team as a code owner June 14, 2022 08:28
@Yoric
Copy link
Contributor Author

Yoric commented Jun 14, 2022

As far as I can tell, the monolith test fails because of a network issue (cannot assign IP), unrelated to the PR.

@Yoric
Copy link
Contributor Author

Yoric commented Jun 16, 2022

For some reason, poetry run ./scripts-dev/lint.sh disagree on CI and on my machine.

changelog.d/13044.misc Outdated Show resolved Hide resolved
synapse/events/spamcheck.py Outdated Show resolved Hide resolved
synapse/events/spamcheck.py Outdated Show resolved Hide resolved
tests/rest/client/test_rooms.py Outdated Show resolved Hide resolved
tests/rest/client/test_rooms.py Outdated Show resolved Hide resolved
tests/rest/media/v1/test_media_storage.py Outdated Show resolved Hide resolved
tests/rest/client/test_rooms.py Outdated Show resolved Hide resolved
tests/rest/client/test_rooms.py Show resolved Hide resolved
synapse/events/spamcheck.py Outdated Show resolved Hide resolved
@richvdh richvdh changed the title Uniformize spam-checker API, part 5: expand other spam-checker callba… Uniformize spam-checker API, part 5: expand other spam-checker callbacks to return Tuple[Codes, dict] Jun 17, 2022
@Yoric Yoric requested a review from babolivier June 30, 2022 10:47
Copy link
Contributor

@babolivier babolivier left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Sorry about the round-trip time for this one!

tests/rest/client/test_rooms.py Show resolved Hide resolved
tests/rest/client/test_rooms.py Outdated Show resolved Hide resolved
tests/rest/client/test_rooms.py Outdated Show resolved Hide resolved
@@ -1312,6 +1354,71 @@ def test_rooms_messages_sent(self) -> None:
channel = self.make_request("PUT", path, content)
self.assertEqual(200, channel.code, msg=channel.result["body"])

def test_spam_checker_check_event_for_spam(self) -> None:
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It looks to me like this test could be simplified a lot by using parameterized.expand, see

@parameterized.expand(
[
# host
[b"localhost", b"http", b"localhost", 1080, None],
[b"localhost:9988", b"http", b"localhost", 9988, None],
# host+scheme
[b"https://localhost", b"https", b"localhost", 1080, None],
[b"https://localhost:1234", b"https", b"localhost", 1234, None],
# ipv4
[b"1.2.3.4", b"http", b"1.2.3.4", 1080, None],
[b"1.2.3.4:9988", b"http", b"1.2.3.4", 9988, None],
# ipv4+scheme
[b"https://1.2.3.4", b"https", b"1.2.3.4", 1080, None],
[b"https://1.2.3.4:9988", b"https", b"1.2.3.4", 9988, None],
# ipv6 - without brackets is broken
# [
# b"2001:0db8:85a3:0000:0000:8a2e:0370:effe",
# b"http",
# b"2001:0db8:85a3:0000:0000:8a2e:0370:effe",
# 1080,
# None,
# ],
# [
# b"2001:0db8:85a3:0000:0000:8a2e:0370:1234",
# b"http",
# b"2001:0db8:85a3:0000:0000:8a2e:0370:1234",
# 1080,
# None,
# ],
# [b"::1", b"http", b"::1", 1080, None],
# [b"::ffff:0.0.0.0", b"http", b"::ffff:0.0.0.0", 1080, None],
# ipv6 - with brackets
[
b"[2001:0db8:85a3:0000:0000:8a2e:0370:effe]",
b"http",
b"2001:0db8:85a3:0000:0000:8a2e:0370:effe",
1080,
None,
],
[
b"[2001:0db8:85a3:0000:0000:8a2e:0370:1234]",
b"http",
b"2001:0db8:85a3:0000:0000:8a2e:0370:1234",
1080,
None,
],
[b"[::1]", b"http", b"::1", 1080, None],
[b"[::ffff:0.0.0.0]", b"http", b"::ffff:0.0.0.0", 1080, None],
# ipv6+port
[
b"[2001:0db8:85a3:0000:0000:8a2e:0370:effe]:9988",
b"http",
b"2001:0db8:85a3:0000:0000:8a2e:0370:effe",
9988,
None,
],
[
b"[2001:0db8:85a3:0000:0000:8a2e:0370:1234]:9988",
b"http",
b"2001:0db8:85a3:0000:0000:8a2e:0370:1234",
9988,
None,
],
[b"[::1]:9988", b"http", b"::1", 9988, None],
[b"[::ffff:0.0.0.0]:9988", b"http", b"::ffff:0.0.0.0", 9988, None],
# ipv6+scheme
[
b"https://[2001:0db8:85a3:0000:0000:8a2e:0370:effe]",
b"https",
b"2001:0db8:85a3:0000:0000:8a2e:0370:effe",
1080,
None,
],
[
b"https://[2001:0db8:85a3:0000:0000:8a2e:0370:1234]",
b"https",
b"2001:0db8:85a3:0000:0000:8a2e:0370:1234",
1080,
None,
],
[b"https://[::1]", b"https", b"::1", 1080, None],
[b"https://[::ffff:0.0.0.0]", b"https", b"::ffff:0.0.0.0", 1080, None],
# ipv6+scheme+port
[
b"https://[2001:0db8:85a3:0000:0000:8a2e:0370:effe]:9988",
b"https",
b"2001:0db8:85a3:0000:0000:8a2e:0370:effe",
9988,
None,
],
[
b"https://[2001:0db8:85a3:0000:0000:8a2e:0370:1234]:9988",
b"https",
b"2001:0db8:85a3:0000:0000:8a2e:0370:1234",
9988,
None,
],
[b"https://[::1]:9988", b"https", b"::1", 9988, None],
# with credentials
[
b"https://user:[email protected]:9988",
b"https",
b"1.2.3.4",
9988,
b"user:pass",
],
[b"user:[email protected]:9988", b"http", b"1.2.3.4", 9988, b"user:pass"],
[
b"https://user:[email protected]:9988",
b"https",
b"proxy.local",
9988,
b"user:pass",
],
[
b"user:[email protected]:9988",
b"http",
b"proxy.local",
9988,
b"user:pass",
],
]
)
def test_parse_proxy(
self,
proxy_string: bytes,
expected_scheme: bytes,
expected_hostname: bytes,
expected_port: int,
expected_credentials: Optional[bytes],
):
"""
Tests that a given proxy URL will be broken into the components.
Args:
proxy_string: The proxy connection string.
expected_scheme: Expected value of proxy scheme.
expected_hostname: Expected value of proxy hostname.
expected_port: Expected value of proxy port.
expected_credentials: Expected value of credentials.
Must be in form '<username>:<password>' or None
"""
proxy_cred = None
if expected_credentials:
proxy_cred = ProxyCredentials(expected_credentials)
self.assertEqual(
(
expected_scheme,
expected_hostname,
expected_port,
proxy_cred,
),
parse_proxy(proxy_string),
)

or

@parameterized.expand(
[
(["federation"], "auth_fail"),
([], "no_resource"),
(["openid", "federation"], "auth_fail"),
(["openid"], "auth_fail"),
]
)
def test_openid_listener(self, names, expectation):
"""
Test different openid listener configurations.
401 is success here since it means we hit the handler and auth failed.
"""
config = {
"port": 8080,
"type": "http",
"bind_addresses": ["0.0.0.0"],
"resources": [{"names": names}],
}
# Listen with the config
self.hs._listen_http(parse_listener_def(config))
# Grab the resource from the site that was told to listen
site = self.reactor.tcpServers[0][1]
try:
site.resource.children[b"_matrix"].children[b"federation"]
except KeyError:
if expectation == "no_resource":
return
raise
channel = make_request(
self.reactor, site, "GET", "/_matrix/federation/v1/openid/userinfo"
)
self.assertEqual(channel.code, 401)

for examples.

Copy link
Contributor Author

@Yoric Yoric Jul 1, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

So, introduce a new framework to this test along with magic code generation to get rid of one loop?
I can do that if you insist, but my first impression is that this would actually increase cognitive load of reading that test.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What do you mean by introducing a new framework? We already use it quite a bit in Synapse's tests suite. In my opinion it makes tests much easier to read as it reduces the complexity of the actual test code.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I mean that reading this test doesn't require understanding that framework yet. If you wish me to introduce it, I can do that, of course.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Well, your wish is my command :)

tests/rest/media/v1/test_media_storage.py Outdated Show resolved Hide resolved
@Yoric Yoric requested a review from babolivier July 1, 2022 08:26
Copy link
Contributor

@babolivier babolivier left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This plus I really think we should be using parameterized in test_spam_checker_check_event_for_spam.

tests/rest/client/test_rooms.py Outdated Show resolved Hide resolved
tests/rest/client/test_rooms.py Outdated Show resolved Hide resolved
@Yoric Yoric requested a review from babolivier July 5, 2022 13:22
@babolivier babolivier enabled auto-merge (squash) July 11, 2022 15:33
@babolivier babolivier merged commit 11f8114 into develop Jul 11, 2022
@babolivier babolivier deleted the yoric/spam-errors-5 branch July 11, 2022 16:52
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants