add L1 support to getCurrentValidators API #3564
Conversation
Signed-off-by: Ceyhun Onur <[email protected]>
…go into add-l1-validators-api
|
|
||
| func (s *Service) getL1Validators(subnetID ids.ID, nodeIDs set.Set[ids.NodeID]) ([]interface{}, error) { | ||
| validators := []interface{}{} | ||
| baseStakers, l1Validators, _, err := s.vm.state.GetCurrentValidators(subnetID) |
There was a problem hiding this comment.
(No action required) The docstring for GetCurrentValidators refers to the stakers returned as 'legacy stakers'. What does that imply in the context of an L1?
| continue | ||
| } | ||
| weight := avajson.Uint64(staker.Weight) | ||
| apiStaker := platformapi.Staker{ |
There was a problem hiding this comment.
(No action required) It appears that that conversion of state.Staker to platformapi.Staker` is defined twice in this file. Maybe avoid this duplication?
Also - and not for this PR - the need to perform this conversion seems a bit strange. Presumably there is room to unify these types?
| @@ -2130,7 +2130,7 @@ func TestGetCurrentValidators(t *testing.T) { | |||
| NodeID: ids.GenerateTestNodeID(), | |||
| PublicKey: otherPK, | |||
| Weight: 1, | |||
| StartTime: now, | |||
| StartTime: now.Add(1 * time.Second), | |||
There was a problem hiding this comment.
(No action required) Why does the start time change in these tests?
| name: "initial stakers", | ||
| initial: []*state.Staker{ | ||
| { | ||
| TxID: ids.GenerateTestID(), |
There was a problem hiding this comment.
(No action required) Given that the following fields are initialized the same way for current tests, maybe move their initialization to the test body to ensure that only relevant fields need be maintained for each test case?
Staker.TxIDL1Validator.ValidationID*.SubnetID*.NodeID
| }, | ||
| } | ||
|
|
||
| remaningBalanceOwner := message.PChainOwner{ |
|
|
||
| require.NoError(service.vm.state.PutL1Validator(l1Validator)) | ||
|
|
||
| if l1Validator.Weight == 0 { |
There was a problem hiding this comment.
(No action required) Why it it desirable to call PutL1Validator with zero weight but not add to l1ValidatorsByVID?
| service.vm.ctx.Lock.Unlock() | ||
|
|
||
| testValidator := func(vdr interface{}) ids.NodeID { | ||
| switch v := vdr.(type) { |
There was a problem hiding this comment.
(No action required) Not for this PR, but maybe these types would benefit from an equality method to simplify this kind of testing?
Why this should be merged
It adds L1 support and shows L1 validators in
platform.getCurrentValidatorsAPI.How this works
validationIDto theplatform.getL1ValidatorAPI outputs.platform.getCurrentValidatorsAPI.PermissionedValidatorstruct as it is no longer used.getL1ValidatorsandgetPrimaryOrSubnetValidatorsmethods to handle fetching validators based on subnet ID.convertL1ValidatorToAPImethod to convert L1 validator data to API format.GetL1Validatorsmethod to theStateinterface to return both base stakers and L1 validators in a converted L1.How this was tested
Locally
Need to be documented in RELEASES.md?
Yes, updated.