feat(pkg/mmr): Add basic MMR support

[dimartiro]

Changes

Add basic MMR support following the 1st iteration implementation proposed here

Features

• Leaves accumulation
• Root calculation
• Commit changes to underlying storage
• Basic testing

Tests

go test ./pkg/mmr

Issues

Closes: #4152

[EclesioMeloJunior]

LGTM! Would like to see more tests over getElement, also can the storage be generic over MMRElement?

[dimartiro]

Would like to see more tests over getElement

I'm gonna add more, thanks for the review 👍

also can the storage be generic over MMRElement?

You mean to use a generic instead of using MMRElement? basically MMRElement is an alias for []byte I wanted to keep it simple since it is an MVP and using []byte is easier for us since we don't need any special serializer or deserializer.
I think we can improve it later if we have any future special requirement to use a generic

[jimjbrettj]

Really nice work! I already looked over it, but if you could do a once over of the GP appendix E to make sure it seems compatible before merging that would be great.

[dimartiro]

Really nice work! I already looked over it, but if you could do a once over of the GP appendix E to make sure it seems compatible before merging that would be great.

I've been checking, and if I'm not mistaken, it aligns well with the JAM GP.

The Append function describes how to add a new element in a generic way and how to merge it with the current peaks. If I understand correctly, it considers an MMR, which isn't necessarily limited to binary tries, which is why the function P is described based on n.

Regarding the encoding function, it essentially describes the process of bagging the peaks and calculating the final MMR root so we are ok with that too.

This is all based on my interpretation of the functions described in the GP, but in the future, we can compare our implementation with Parity’s to ensure we have the same behavior 😄

[EclesioMeloJunior]

I think this comment is not needed, the func signature is self-explanatory imo, maybe you can use it to explain how the merge is implemented.

[dimartiro]

Agree, but I wanted to add a comment since it is an exported type

[dimartiro]

Closed in favor of https://github.com/ChainSafe/go-jam/pull/38