Struct RangeEncoder 

pub struct RangeEncoder<W: Write> { /* private fields */ }

Streaming range‐encoder for arithmetic coding.

The encoder maintains three key internal values:

• low: the low end of the current coding interval
• range: the size of the current coding interval
• buffer: a byte buffer (flushed in 4 KB chunks) holding the high‐order output bytes

For each symbol you wish to encode, call encode with:

1. cum_freq: cumulative frequency of all symbols less than the one you’re encoding
2. freq: the frequency of the symbol itself
3. tot_freq: the total of all symbol frequencies in the current context

After encoding every symbol, call finish to flush any remaining bytes and retrieve the underlying writer.

Example

use std::fs::File;
use ppmd_core::{RangeEncoder, PpmResult};

fn encode_stream() -> PpmResult<()> {
    let file = File::create("out.ppm")?;
    let mut encoder = RangeEncoder::new(file);

    // Suppose `model` yields (cum, freq, tot) triples for each byte:
    for (cum, freq, tot) in model.symbols() {
        encoder.encode(cum, freq, tot)?;
    }

    // Finalize and get back the file writer
    let _file = encoder.finish()?;
    Ok(())
}

Implementations

impl<W: Write> RangeEncoder<W>

pub fn new(writer: W) -> Self

Create a new range encoder wrapping writer.

Initializes:

• low = 0
• range = u32::MAX (the full 32-bit interval)
• an internal 4 KB output buffer

The encoder will emit one output byte at a time into buffer, flushing to writer whenever the buffer is full.