feat(core): CKB_TX_MESSAGE_ALL

[XuJiandong]

described in: nervosnetwork/rfcs#446

[changeset-bot]

🦋 Changeset detected

Latest commit: 396dbb4

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 20 packages

Name	Type
@ckb-ccc/core	Patch
@ckb-ccc/eip6963	Patch
@ckb-ccc/joy-id	Patch
@ckb-ccc/lumos-patches	Patch
@ckb-ccc/nip07	Patch
@ckb-ccc/okx	Patch
@ckb-ccc/rei	Patch
@ckb-ccc/shell	Patch
@ckb-ccc/spore	Patch
@ckb-ccc/ssri	Patch
@ckb-ccc/udt	Patch
@ckb-ccc/uni-sat	Patch
@ckb-ccc/utxo-global	Patch
@ckb-ccc/xverse	Patch
@ckb-ccc/ccc	Patch
ckb-ccc	Patch
@ckb-ccc/connector	Patch
@ckb-ccc/examples	Patch
@ckb-ccc/ccc-playground	Patch
@ckb-ccc/connector-react	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[XuJiandong]

integration tests: xxuejie/ckb-tx-message-all-test-vector-utils#2

[Hanssen0]

This PR provides a good start for CCC to support MESSAGE_ALL signing in the future. However, some API designs need to be discussed more before merging.

[Hanssen0]

I noticed that the logic of hashBytes[Opt]ToHasher can actually be replaced by mol.Bytes[Opt]. I suggest using mol.Bytes[Opt].encode directly to save some code here.

By the way, please add a changeset using pnpm changeset in the project directory.

Anyway, I think this code is good enough to be merged. Thank you so much for the contribution. Should we merge it directly, or wait until the proposal is confirmed?

[XuJiandong]

I noticed that the logic of hashBytes[Opt]ToHasher can actually be replaced by mol.Bytes[Opt]. I suggest using mol.Bytes[Opt].encode directly to save some code here.

According to the RFC:

The length of the witness field, packed in little-endian encoded unsigned 32-bit integer.
The full witness field, or the full content returned by Load Witness syscall, given the correct index and source

The hashBytesToHasher function follows the RFC mentioned above(Although It uses the same encoding as Molecule bytes, it doesn't mention molecule encoding). I believe this version is much cleaner and easier to understand.

The hashBytesOptToHasher function adds an extra header, based on BytesOpt encoding. When BytesOpt is none, it results in zero bytes, so we include an extra header to differentiate it. Here, I've added a molecule encode, according to the RFC.

[Hanssen0]

LGTM