Struct EntropyDecoder 

pub struct EntropyDecoder<S: RansParams> { /* private fields */ }

High-level entropy decoder using CDF tables for symbol lookup.

Generic over S: RansParams (RansByte or Rans64).

Implementations

impl<S: RansParams> EntropyDecoder<S>

pub fn new() -> Self

Create a new uninitialized entropy decoder.

pub fn initialize( &mut self, pmf_lengths: &[i32], pmf_offsets: &[i32], pmf_table: &[i32], symbol_bits: u32, bypass_bits: u32, ) -> Result<(), EntropyError>

Initialize the decoder with PMF distribution data.

• pmf_lengths — number of symbols per distribution (including bypass sentinel)
• pmf_offsets — value offsets per distribution
• pmf_table — flat array of symbol frequencies for all distributions
• symbol_bits — number of bits for symbol encoding (e.g. 16)
• bypass_bits — number of bits for bypass encoding (e.g. 4)

pub fn decode( &self, values: &mut [i32], indices: &[i32], data: &[u8], ) -> Result<(), EntropyError>

One-shot decode: decode from data into values.

• values — output buffer (must be same length as indices)
• indices — distribution indices for each value to decode
• data — encoded byte stream

pub fn decode_partial( &self, values: &mut [i32], indices: &[i32], data: &[u8], ) -> Result<usize, EntropyError>

Decode from a slice but do NOT require the source to be fully exhausted.

This is used when decoding from a RansDecoderStream where multiple encoded segments are concatenated. The method decodes values/indices from the beginning of data and returns the number of bytes consumed.

• values — output buffer (must be same length as indices)
• indices — distribution indices for each value to decode
• data — encoded byte stream (may contain extra trailing data)

Returns the number of bytes consumed from data on success.

Trait Implementations

impl<S: RansParams> Default for EntropyDecoder<S>

fn default() -> Self

Returns the “default value” for a type.