Trait Transcript

Source

pub trait Transcript:
    Clone
    + Read
    + Write {
    // Required methods
    fn new(label: &[u8]) -> Self;
    fn absorb_raw(&mut self, data: &[u8]);
    fn squeeze_raw(&mut self, buf: &mut [u8]);

    // Provided methods
    fn absorb_serialize(&mut self, obj: &impl CanonicalSerialize) { ... }
    fn squeeze_deserialize<T: CanonicalDeserialize>(&mut self) -> T { ... }
    fn to_rng(self) -> TranscriptRng<Self>
       where Self: Sized { ... }
}

Expand description

Fiat-Shamir transcript with absorb/squeeze interface.

Provides a streaming hash abstraction where data is absorbed into an internal state and arbitrary-length output is squeezed out.

Implements io::Write so that serializable types (points, scalars) can be written directly into the transcript without intermediate buffers.

Implementations do not need to handle domain separation or length-prefixing of variable-length inputs. The protocol layer takes care of this by absorbing domain-separation tags and explicit lengths before variable-length data. Since absorb_raw is a plain concatenation into a single hash stream (absorb then squeeze, no resets), the domain-separation bytes injected by the caller are sufficient to prevent ambiguous parses.

Required Methods§

Source

fn new(label: &[u8]) -> Self

Create a new transcript.

Source

fn absorb_raw(&mut self, data: &[u8])

Absorb raw bytes into the transcript.

This is a plain concatenation into the internal hash state. Domain separation and length-prefixing of variable-length fields are the caller’s responsibility (handled by the protocol layer).