feat: add keccak256 merkle tree

[alxiong]

This PR:

• introduce macro rules impl_mt_hash_256! for implementing boilerplate for node and digest for MT that has 32 bytes output size
• impl for both Sha3 and Keccak256 under this macro
• expose a MT and lightweight MT for keccak256

@mrain do you remember why we pick rate 3 for MT? is that an assumption for our current MT internals?

---

Before we can merge this PR, please make sure that all the following items have been
checked off. If any of the checklist items are not applicable, please leave them but
write a little note why.

• Targeted PR against correct branch (main)
• Linked to GitHub issue with discussion and accepted design OR have an explanation in the PR that describes this work.
• Wrote unit tests
• Updated relevant documentation in the code
• Added relevant changelog entries to the CHANGELOG.md of touched crates.
• Re-reviewed Files changed in the GitHub PR explorer

[mrain]

@mrain do you remember why we pick rate 3 for MT? is that an assumption for our current MT internals?

Rescue hash has a rate of 3.

@sveitser I think it's better to instantiate a MT in the application since it's simple enough. If we do it here, people may think that we are enforcing some default instantiations while we are not.

[alxiong]

Rescue hash has a rate of 3.

but our Sha3MT doesn't have to use rate of 3, right? also our implementation is generic of ARITY. I don't know why our Sha3Tree also choose Arity of 3

@sveitser I think it's better to instantiate a MT in the application since it's simple enough. If we do it here, people may think that we are enforcing some default instantiations while we are not.

Softly disagree, the fact that our API is generic already offers flexibility. prelude::* offers good out-of-box options that can be safely ignored.
I don't see downside of adding Keccak256MerkleTree in jellyfish actually.

[mrain]

but our Sha3MT doesn't have to use rate of 3, right? also our implementation is generic of ARITY. I don't know why our Sha3Tree also choose Arity of 3

I just picked a random number because it was supposed to be only an example. If we are making it public we'd better make it generic to ARITY.

Softly disagree, the fact that our API is generic already offers flexibility. prelude::* offers good out-of-box options that can be safely ignored. I don't see downside of adding Keccak256MerkleTree in jellyfish actually.

We could go either way.

[mrain]

LGTM. I think we could include this change in version 0.2.0, and make a new tag after merge.

[mrain]

LGTM!