test: DDO onboarding non-market verified data #11603
Conversation
itests/direct_data_onboard_test.go
Outdated
| events := make([]types.ActorEvent, 0) | ||
| go func() { | ||
| for e := range evtChan { | ||
| fmt.Printf("%s Got ActorEvent: %+v", time.Now().Format(time.StampMilli), e) |
There was a problem hiding this comment.
currently doesn't show anything because they get filtered out due to address translation in the event filter
itests/direct_data_onboard_test.go
Outdated
| require.NoError(t, err) | ||
|
|
||
| for _, evt := range evts { | ||
| fmt.Printf("Got ActorEvent: %+v", evt) |
There was a problem hiding this comment.
also doesn't show anything for the same reason
| for _, evt := range evts { | ||
| fmt.Printf("Got ActorEvent: %+v", evt) | ||
| } | ||
|
|
There was a problem hiding this comment.
from here below we use the existing lotus APIs to get receipts and their event logs for the whole chain, which works
rvagg
force-pushed
the
rvagg/ddo-verified
branch
from
96a551c
to
fbd2cd6
Compare
rvagg
force-pushed
the
rvagg/ddo-verified
branch
from
fbd2cd6
to
ff17d7f
Compare
| params, err := actors.SerializeParams(&verifregtypes13.AddVerifiedClientParams{Address: verifiedClientAddr, Allowance: initialDatacap}) | ||
| require.NoError(t, err) | ||
|
|
||
| msg := &types.Message{ |
There was a problem hiding this comment.
we really need to extract some of this logic into a helper to be reused by multiple tests
| fmt.Sprintf("height=%d", event.Height), | ||
| fmt.Sprintf("msg=%s", event.MsgCid), | ||
| fmt.Sprintf("emitter=%s", event.EmitterAddr), | ||
| fmt.Sprintf("reverted=%t", event.Reverted), |
There was a problem hiding this comment.
This information is not persisted on message receipts. How will client be informed about reverted events if they only depend on message receipts ?
|
Addes tests for the events API. |
aarshkshah1992
requested review from
Stebalien and
aarshkshah1992
and removed request for
a team
| func matchEvents(t *testing.T, exp []types.ActorEvent, actual []types.ActorEvent) { | ||
| // height and tipset cid can mismatch because expected events are sourced using APIs that can put in different tipsets | ||
| for i := range exp { | ||
| exp[i].Height = 0 |
There was a problem hiding this comment.
@rvagg The tipset/height we put in the buildActorEventsFromMessages function below on these events is always one ahead of the actual tipset/height these events are actually emitted at (as recorded by the CollectEvents function in lotus/chain/events/index.go).
I tried digging into it and it looks like this is because the node.StateSearchMsg function can return messages from a previous tipset as well. Let me know what you think here. I think this is a non-issue.
There was a problem hiding this comment.
Oh, that's very interesting. Probably a non issue as far as the tests are concerned but this is really helpful to know as I write docs.
|
itests will be green once the FFI/ref-fvm deps changes for NV22 are merged into this branch. |
aarshkshah1992
merged commit 4cfd4f0
into
feat/built-in-actor-events-api
WIP. I'm planning on adding some events testing in this. This could also do with some refactoring, maybe moving some of this into the testkit because it's quite verbose but also some of it is repeated from other verifreg testing.