3.1.4 Further document hash_to_F which may collide

This commit is contained in:
Luke Parker
2023-02-23 01:03:53 -05:00
parent cb4ce5e354
commit 686a5ee364
3 changed files with 34 additions and 5 deletions

View File

@@ -17,8 +17,6 @@ macro_rules! dalek_curve {
) => {
use dalek_ff_group::$Point;
#[derive(Clone, Copy, PartialEq, Eq, Debug, Zeroize)]
pub struct $Ciphersuite;
impl Ciphersuite for $Ciphersuite {
type F = Scalar;
type G = $Point;
@@ -37,6 +35,14 @@ macro_rules! dalek_curve {
};
}
/// Ciphersuite for Ristretto.
///
/// hash_to_F is implemented with a naive concatenation of the dst and data, allowing transposition
/// between the two. This means `dst: b"abc", data: b"def"`, will produce the same scalar as
/// `dst: "abcdef", data: b""`. Please use carefully, not letting dsts be substrings of each other.
#[cfg(any(test, feature = "ristretto"))]
#[derive(Clone, Copy, PartialEq, Eq, Debug, Zeroize)]
pub struct Ristretto;
#[cfg(any(test, feature = "ristretto"))]
dalek_curve!("ristretto", Ristretto, RistrettoPoint, b"ristretto");
#[cfg(any(test, feature = "ristretto"))]
@@ -60,6 +66,14 @@ fn test_ristretto() {
);
}
/// Ciphersuite for Ed25519.
///
/// hash_to_F is implemented with a naive concatenation of the dst and data, allowing transposition
/// between the two. This means `dst: b"abc", data: b"def"`, will produce the same scalar as
/// `dst: "abcdef", data: b""`. Please use carefully, not letting dsts be substrings of each other.
#[cfg(feature = "ed25519")]
#[derive(Clone, Copy, PartialEq, Eq, Debug, Zeroize)]
pub struct Ed25519;
#[cfg(feature = "ed25519")]
dalek_curve!("ed25519", Ed25519, EdwardsPoint, b"edwards25519");
#[cfg(feature = "ed25519")]