Refactoring of CoverCrypt API #134
Conversation
| Ok(res) | ||
| } | ||
| plaintext: &[u8], | ||
| ) -> Result<(Encapsulation, Vec<u8>), Error>; |
There was a problem hiding this comment.
I wonder if we should already serialize the encapsulation: the current solution seems to be somewhere in-between returning a struct containing each parts (encapsulation, mac, none and ciphertext) and serializing and concatenating all parts.
I don't have any strong opinion on it but this solution is surprising.
|
Missing only implementation of theEncryptionHandler methods. Now they as trait's functions |
|
@tbrezot I think I've addressed all your comments. Maybe miscommunication, but idk if I have to do a new pull request or not. |
src/core/api.rs
Outdated
| String::from_str("SEED_LENGTH >= KEY_LENGTH").unwrap(), | ||
| )); | ||
| } | ||
| let sym_key = SymmetricKey::<KEY_LENGTH>::try_from_slice(&seed[..KEY_LENGTH])?; |
There was a problem hiding this comment.
We may want to implement a key derivation from a secret:
impl<const KEY_LENGTH: usize> SymmetricKey<KEY_LENGTH> {
fn derive<const SEED_LENGTH: usize>(&mut impl CryptoRngCore, seed: &Seed<SEED_LENGTH>, info: &[u8]) -> Result<Self, Error> {
if SEED_LENGTH < KEY_LENGTH {
Err(...) // insufficient entropy
} else {
let mut key = Self::default();
kdf!(&mut key, seed, info);
Ok(key)
}
}
}```There was a problem hiding this comment.
This has to be implemented in crypto_core, right? The type is not defined in Covercrypt. @tbrezot
There was a problem hiding this comment.
yes, maybe you could push a commit to develop ?
No description provided.