This crate contains a wrapper of Blake2b that produces a 32 byte digest, instead of 64 byte, for convenience.
Similar objects can be placed here.
PseudoMerlin is an adaptor that allows to emulate the append_message API of
a Merlin transcript, using any cryptographic digest. The purpose of this is
to use it with Digestible crate.
The idea is that append_message(label, data) should be implemented as
self.digest.update((label.len() as u32).to_le_bytes());
self.digest.update(label);
self.digest.update((data.len() as u32).to_le_bytes());
self.digest.update(data);
This creates a prefix-free encoding of (label, data) pairs, so when the digest
is ultimately extracted, we know that if an adversary can find two different
sequences of (label, data) pairs that cause PseudoMerlin to give the same hash
it means that the adversary can also find a nontrivial collision in the underlying digest function.
If the digest function is collision resistant in the classical sense, this means that
PseudoMerlin has the same security property that Digestible crate needs of Merlin.
Even though this is available, it is recommended not to use this and to use Merlin instead
if possible. This exists to support any niche cases where data must be digested into a particular
hasher in order to be compatible with a particular API.