Module entropy_coding 

Entropy coding primitives.

This module provides simplified implementations of common entropy coding techniques used in video codecs: arithmetic coding, range coding, and Huffman coding.

Structs

AdaptiveFrequencyTracker

Adaptive frequency tracker using a sliding window.

ArithmeticCoder

Simplified binary arithmetic coder.

CabacContext

A single CABAC context model.

CabacEncoder

CABAC encoder with multiple context models.

HuffmanNode

A node in a Huffman tree.

ProbTableEntry

Pre-computed lookup table entry for one probability value.

RangeCoder

Simplified range coder (decoder side).

RangeEncoder

Multi-symbol range encoder that writes bytes to an output buffer.

TableArithmeticCoder

High-throughput table-based arithmetic coder.

TableArithmeticDecoder

Table-based arithmetic decoder counterpart.

Functions

build_huffman_tree

Builds a Huffman tree from a frequency table using a greedy (priority-queue) algorithm.

build_prob_table

Builds a probability lookup table for n_syms symbols.

compare_coding_strategies

Compare two coding strategies and return which uses fewer estimated bits.

compute_huffman_codes

Traverses the Huffman tree depth-first, collecting (symbol, code_bits) pairs at each leaf.

estimate_block_entropy

Estimate the number of bits needed to encode a block of symbols without actually producing a bitstream.

estimate_entropy_from_freqs

Estimate the entropy (bits per symbol) of a frequency distribution.

optimal_code_lengths

Compute optimal code lengths for a set of symbol frequencies.