feat: rlp encoding for logs with generic event data #553
Conversation
|
@Rjected ptal |
| impl<T> alloy_rlp::Encodable for Log<T> | ||
| where | ||
| for<'a> &'a T: Into<LogData>, | ||
| { |
There was a problem hiding this comment.
ah I see, can just be converted to Log<LogData> when serializing
There was a problem hiding this comment.
right. and because we are usually using sol! there is no problem
This is significantly less ergonomic with DynSolEvent tho, as a decoded sol event can't be assumed to be validly encoded the way a decoded sol! struct with SolEvent can be. That
assumption allows the infallible reserialize
cc @DaniPopes do you think that we should clone + attach the DynSolEvent as a prop on the DecodedEvent in alloy-dyn-abi so that we can assume valid construction and encoding? This would let us impl From<&DecodedEvent> for LogData
|
cc @DaniPopes what's up with the tests? |
|
You can update them by running with env |
|
this also needs to be rebased |
|
clippy issue seems to have been introduced by merge of main. why? Please do not merge until we know why this is the case |
the new nightly clippy doesnt like things that are never constructed, including in tests |
|
hmm unsure, was this struct ever used? doing yields nothing except changelog and the struct def itself weird that it wouldn't be caught before this? edit: not weird given dan comment |
why are we using nightly clippy? didnt we decide that was a mistake with ethers as it introduced unstable lints and situations like this where CI breaks for no reason? |
prestwich
force-pushed
the
prestwich/blanket-log-encode
branch
from
2ff92d4
to
ee44732
Compare
|
CI is unrelated. |
|
This breaks RLP encoding for |
closes #544
Motivation
Allow rlp encoding of logs after their data has been ABI decoded.
Solution
Add a set of methods and a blanket Encode implementation for
Log<T> where for<'a> &'a T: Into<LogData>,Note that this change does not allow decoding, as that would require embedding the ABI error into the RLP error decode impl
PR Checklist