First commit
This commit is contained in:
commit
6da0e73c5e
1
.gitignore
vendored
Normal file
1
.gitignore
vendored
Normal file
@ -0,0 +1 @@
|
||||
/target
|
||||
6
Cargo.toml
Normal file
6
Cargo.toml
Normal file
@ -0,0 +1,6 @@
|
||||
[package]
|
||||
name = "sac"
|
||||
version = "0.1.0"
|
||||
edition = "2021"
|
||||
|
||||
[dependencies]
|
||||
99
src/main.rs
Normal file
99
src/main.rs
Normal file
@ -0,0 +1,99 @@
|
||||
// https://marknelson.us/posts/2014/10/19/data-compression-with-arithmetic-coding.html
|
||||
use std::collections::HashMap;
|
||||
|
||||
type Model = HashMap<u8, (f64, f64)>;
|
||||
|
||||
fn get_symbol(model: &Model, d: f64) -> Option<u8> {
|
||||
// Brute force
|
||||
for (symbol, (start, end)) in model {
|
||||
if d >= *start && d < *end {
|
||||
return Some(*symbol);
|
||||
}
|
||||
}
|
||||
return None;
|
||||
}
|
||||
|
||||
fn encode(data: &[u8], model: &Model) -> f64 {
|
||||
let mut high: f64 = 1.0;
|
||||
let mut low: f64 = 0.0;
|
||||
for symbol in data {
|
||||
let p = model.get(symbol).expect("Invalid/Unsupported data");
|
||||
let range = high - low;
|
||||
high = low + range * p.1;
|
||||
low = low + range * p.0;
|
||||
}
|
||||
return low + (high - low) / 2.0;
|
||||
}
|
||||
|
||||
fn decode(message: f64, model: &Model) {
|
||||
let mut high: f64 = 1.0;
|
||||
let mut low: f64 = 0.0;
|
||||
loop {
|
||||
let range = high - low;
|
||||
let d = (message - low) / range;
|
||||
let c = match get_symbol(&model, d) {
|
||||
Some(c) => c,
|
||||
None => {
|
||||
println!("");
|
||||
eprintln!("Decode error: d={d}");
|
||||
return;
|
||||
}
|
||||
};
|
||||
if c == b'-' {
|
||||
println!("");
|
||||
return;
|
||||
}
|
||||
print!("{}", c as char);
|
||||
let p = model.get(&c).expect("Decode error");
|
||||
high = low + range * p.1;
|
||||
low = low + range * p.0;
|
||||
}
|
||||
}
|
||||
|
||||
fn make_model(probabilities: &[(u8, f64)]) -> Model {
|
||||
let mut model = HashMap::new();
|
||||
let mut end: f64 = 0.0;
|
||||
for (symbol, probability) in probabilities {
|
||||
let start: f64 = end;
|
||||
end = start + probability;
|
||||
model.insert(*symbol, (start, end));
|
||||
println!("{}: [{}, {})", *symbol as char, start, end);
|
||||
}
|
||||
return model;
|
||||
}
|
||||
const ENGLISH: &[(u8, f64)] = &[
|
||||
(b'a', 0.08),
|
||||
(b'b', 0.01),
|
||||
(b'c', 0.02),
|
||||
(b'd', 0.04),
|
||||
(b'e', 0.12),
|
||||
(b'f', 0.02),
|
||||
(b'g', 0.02),
|
||||
(b'h', 0.06),
|
||||
(b'i', 0.07),
|
||||
(b'j', 0.01),
|
||||
(b'k', 0.01),
|
||||
(b'l', 0.04),
|
||||
(b'm', 0.02),
|
||||
(b'n', 0.06),
|
||||
(b'o', 0.07),
|
||||
(b'p', 0.01),
|
||||
(b'q', 0.01),
|
||||
(b'r', 0.06),
|
||||
(b's', 0.06),
|
||||
(b't', 0.09),
|
||||
(b'u', 0.02),
|
||||
(b'v', 0.01),
|
||||
(b'w', 0.02),
|
||||
(b'x', 0.01),
|
||||
(b'y', 0.02),
|
||||
(b'z', 0.01),
|
||||
(b' ', 0.01),
|
||||
(b'-', 0.02),
|
||||
];
|
||||
fn main() {
|
||||
let model: Model = make_model(ENGLISH);
|
||||
let message = encode(b"hello world-", &model);
|
||||
println!("{message}");
|
||||
decode(message, &model);
|
||||
}
|
||||
Loading…
x
Reference in New Issue
Block a user