noir_sort

Efficiently sorts fixed-sized arrays.

Noir version compatibility

This library is tested with all Noir stable releases from v0.36.0.

Usage

1. Basic usage:

use dep::sort::sort;

fn foo(a: [u32; 100]) -> [u32; 100] {
    sort(a) // tadaa
}

2. Usage with a custom sort function

use dep::sort::sort_custom;

struct Entry {
    key: Field,
    value: u32
}
fn sort_entry(a: Entry, b: Entry) -> bool {
    a.value <= b.value
}

fn foo(a: [Entry; 100]) -> [Entry; 100] {
    sort_custom(a, sort_entry)
}

3. Usage with an unconditional lte function

fn sort_u16(a: u16, b: u16) -> bool { a <= b }

fn unconditional_lte(a: u16, b: u16) {
    let diff = (b as Field - a as Field);
    diff.assert_max_bit_size(16);
}

fn foo(a: [u16; 100]) -> [u16; 100] {
    sort_extended(a, sort_u16, unconditional_lte)
}

Comments

The sort_extended method is likely to be the most efficient method as asserting that a <= b costs fewer constraints than determining whether a <= b and assigning a bool to the outcome (e.g. for a u16, the <= operator needs to constrain the case where a <= b and a > b and then conditionally assign the return value to the correct case)

Algorithm Description

The library executes, in an unconstrained function, a quicksort algorithm to determine the sorted array.

The library perform two constrained steps:

1. Validates the sorted array contains the same values as the unsorted array (using the check_shuffle library)
2. Validates that, for the sorted array, successive elements are not smaller than previous elements

The algorithm is highly optimized and the cost is linear in the size of the array.