Support for Keccak

[kumavis]

We need support for base Keccak as well as the already added SHA-3 standard subset of Keccak.
Ethereum calls its hash algo of choice "SHA-3" but was chosen before the standard was settled, and is more accurately called "Keccak-256".

Instance	Definition
eth_sha3(M)	Keccak[256](M, 256)
SHA3-224(M)	Keccak[448](M
SHA3-256(M)	Keccak[512](M
SHA3-384(M)	Keccak[768](M
SHA3-512(M)	Keccak[1024](M
SHAKE128(M, d)	Keccak[256](M
SHAKE256(M, d)	Keccak[512](M

table from sha-3 wiki

further reading:

• http://ethereum.stackexchange.com/questions/550/which-cryptographic-hash-function-does-ethereum-use
• Rename/alias sha3 to minimize confusion with SHA-3 standard ethereum/EIPs#59
• https://github.com/emn178/js-sha3
• http://emn178.github.io/online-tools/keccak_256.html
• http://emn178.github.io/online-tools/sha3_256.html

[kumavis]

related to:

• sha3/keccak variants and clarification #11
• Clarify sha3 #12
• add compatibility note on sha3 as alias for sha3-512 #22

[kumavis]

cc @jbenet @diasdavid

[jbenet]

Yeah absolutely, we should add all these. @diasdavid is revamping multicodec table, maybe he can assign codes for them?

[daviddias]

Was just replying through IRC, but better to answer here

tl;dr; the multicodec table inside the multihash table is out of date

See https://github.com/multiformats/multicodec/tree/spec/update, it already has codes for the algos you are looking for. If I'm reading it correctly, there is a direct mapping in implementation between sha3 and Keccak, so, should we recognize both names, or will it be acceptable for devs to use the sha3 names, since their are standard by NIST?

With regards to eth_sha3, since it is a special variant, we can add it, can we call it sha3-eth, so that it is adequately namespaced like all the other sha3?

@jbenet can we get a review on multiformats/multicodec#16, there has been some proposals with regards to multicodecs

[Kubuxu]

Its official name AFAIK, is keccak-256 so I would go with it.

[kumavis]

@diasdavid to be clear, NIST standard SHA-3 is a subset of keccak. eth-sha3 uses "base" keccak-*. ( eth-sha3 is just vanilla keccak-256).

• you can perform NIST standard SHA-3 if you have keccak-*.
• you can perform eth-sha3 if you have keccak-*.
• you can not perform eth-sha3 if you (only) have NIST standard SHA-3.

[kumavis]

Here is the PR for selecting a codec code multiformats/multicodec#19

[axic]

NIST SHA3 and Keccak differ in the padding byte only, otherwise they are the same. Of course this results in entirely incompatible hashes.

The best source for Keccak is http://keccak.noekeon.org/.

[axic]

@diasdavid:

With regards to eth_sha3, since it is a special variant, we can add it, can we call it sha3-eth, so that it is adequately namespaced like all the other sha3?

It is not a special variant. Keccak is the hashing algorithm winning the SHA3 competition and for some reason the standard got the padding byte changed.

@Kubuxu is right, it should be called keccak-<n>

[daviddias]

@axic understood, at first from a conversation, it sounded like eth_sha3 was not straight Keccak or sha3.

I believe we can close this one now, correct?

[Kubuxu]

I think yes.