Struct FastHasher 

pub struct FastHasher { /* private fields */ }

A fast hasher for computing multiple hash values efficiently.

This struct uses a family of permutation hash functions suitable for MinHash. The hash functions have the form: h_i(x) = (a_i * x + b_i) mod p where p is a large prime and a_i, b_i are random odd coefficients.

Implementations

impl FastHasher

pub fn new(num_hashes: usize, seed: u64) -> Self

Create a new hasher with the given number of hash functions.

A count of 0 returns a no-op hasher with empty coefficients. Counts above [crate::config::MAX_SIGNATURE_SIZE] are clamped to it: the coefficients are two u64 vectors of this length, so an unbounded count (for example usize::MAX from a hostile or mistaken caller) would abort the process on allocation failure.

pub fn hash_shingle(&self, shingle: u64) -> Vec<u32>

Compute MinHash signature for a single shingle value.

Returns a vector of hash values, one per hash function. For MinHash, we take the minimum across all shingles.

pub fn update_signature(&self, signature: &mut [u32], shingle: u64)

Update a signature in-place with a new shingle using MINimum update.

For each hash function, updates signature[i] = min(signature[i], hash_i(shingle)).

pub fn update_signature_batch(&self, signature: &mut [u32], shingles: &[u64])

Batch hash multiple shingles and update signature.

This is more cache-efficient than calling update_signature repeatedly.

pub const fn num_hashes(&self) -> usize

Get the number of hash functions.

Trait Implementations

impl Clone for FastHasher

fn clone(&self) -> FastHasher

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for FastHasher

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.