[NONEVM-1091] - Refactor/Setup CR init and cfg and codec init #991
Conversation
ilija42
force-pushed
the
NONEVM-702-cleanup-codec-usage
branch
from
9c4e95d to
382f401
Compare
ilija42
had a problem deploying
to
integration
GitHub Actions
Failure
ilija42
changed the title
|
| // This is necessary because AddressModifier cannot be serialized and must be applied at runtime. | ||
| func injectAddressModifier(outputModifications codeccommon.ModifiersConfig) { | ||
| func injectAddressModifier(inputModifications, outputModifications commoncodec.ModifiersConfig) { |
There was a problem hiding this comment.
Can you elaborate on this a little bit more? Why would someone use commoncodec.AddressBytesToStringModifierConfig instead of codec.SolanaAddressModifier{}?
There was a problem hiding this comment.
Because this AddressBytesToStringModifierConfig is the actual serializeable modifier config, the actual chain specific impl. which is a function and therefore cannot be serialized is injected during CR startup
ilija42
changed the title
ilija42
changed the title
| ReadType ReadType `json:"readType,omitempty"` | ||
| InputModifications commoncodec.ModifiersConfig `json:"inputModifications,omitempty"` | ||
| OutputModifications commoncodec.ModifiersConfig `json:"outputModifications,omitempty"` | ||
| RPCOpts *RPCOpts `json:"rpcOpts,omitempty"` |
There was a problem hiding this comment.
Are the RPCOpts necessary anymore? We are explicitly using finalized commitment and encoding will probably always be base64. The only remaining option is for data slices. Do we use that or are we expecting to use that?
There was a problem hiding this comment.
Idk, I haven't looked into this yet, just kept it because it was already there
There was a problem hiding this comment.
it can most likely be removed
Description