Struct TableArithmeticCoder 

pub struct TableArithmeticCoder { /* private fields */ }

High-throughput table-based arithmetic coder.

Uses pre-computed probability tables for O(1) symbol lookup, avoiding floating-point division in the inner loop. This matches the approach used in many modern video entropy engines.

Design

The coding interval is maintained as (low, range) in a 32-bit window. After each symbol, the interval is renormalised to keep range in the half-open interval [LOW_RANGE_MIN, LOW_RANGE_MAX) by emitting bytes.

Example

use oximedia_codec::entropy_coding::{TableArithmeticCoder, build_prob_table};

let freqs = [10u32, 30, 20, 5];
let table = build_prob_table(&freqs);
let mut enc = TableArithmeticCoder::new();
enc.encode_symbol(1, &table[1]);
enc.encode_symbol(0, &table[0]);
let bitstream = enc.flush();
assert!(!bitstream.is_empty() || true); // output depends on state

Implementations

impl TableArithmeticCoder

pub fn new() -> Self

Create a new table-based arithmetic coder.

pub fn encode_symbol(&mut self, sym_is_high: bool, entry: &ProbTableEntry)

Encode one symbol using its pre-computed ProbTableEntry.

sym_is_high selects the high partition when true.

pub fn flush(self) -> Vec<u8>

Flush any remaining state, returning the full byte stream.

pub fn bytes_emitted(&self) -> usize

Returns the number of bytes emitted so far (before flush).

Trait Implementations

impl Clone for TableArithmeticCoder

fn clone(&self) -> TableArithmeticCoder

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for TableArithmeticCoder

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

Formats the value using the given formatter.