Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/totomanov/sortnet
Gas-optimized sorting networks in Solidity.
https://github.com/totomanov/sortnet
ethereum solidity sorting-network
Last synced: 21 days ago
JSON representation
Gas-optimized sorting networks in Solidity.
- Host: GitHub
- URL: https://github.com/totomanov/sortnet
- Owner: totomanov
- License: mit
- Created: 2024-04-18T16:24:05.000Z (7 months ago)
- Default Branch: master
- Last Pushed: 2024-04-19T21:08:10.000Z (7 months ago)
- Last Synced: 2024-04-20T18:39:30.070Z (7 months ago)
- Topics: ethereum, solidity, sorting-network
- Language: Solidity
- Homepage:
- Size: 35.2 KB
- Stars: 6
- Watchers: 1
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
## Sortnet
Sort fixed-size Solidity arrays in an extremely gas-efficient way.
For arrays of size 2-10 `LibSortNet` consumes 30–60% less gas than insertion sort in [Solady](https://github.com/Vectorized/solady).
## Introduction
A _sorting network_ is a sequence of operations that sorts an array in place. Each operation compares two elements and swaps them conditionally.
### Example
`(0,2),(0,1),(1,2)` is a sorting network for an array of length 3.
```solidity
[1 2 0] --(0,2)-> [0 2 1] --(0,1)-> [0 2 1] --(1,2)-> [0 1 2]
[2 0 1] --(0,2)-> [1 0 2] --(0,1)-> [0 1 2] --(1,2)-> [0 1 2]
[2 1 0] --(0,2)-> [0 1 2] --(0,1)-> [0 1 2] --(1,2)-> [0 1 2]
```
## Resources
[Introduction to Sorting Networks (Doug Hoyte)](https://hoytech.github.io/sorting-networks/)
[List of Sorting Networks (Bert Dobbelaere)](https://bertdobbelaere.github.io/sorting_networks.html)
## Installation
To install with [**Foundry**](https://github.com/foundry-rs/foundry):
```sh
forge install totomanov/sortnet
```
## Usage
#### `function sort*(uint256[] memory a) internal pure`
Sorts a `uint256[] memory` array in-place. The size must be known beforehand and must be between 3 and 32.
#### `function median*(uint256[] memory a) internal pure returns (uint256 m)`
Returns the median of a `uint256[] memory` array with known size. The size must be known beforehand and must be between 3 and 32.
## Benchmark
```sh
forge test --match-test Bench -vvv
```
### Gas table
| size | `@sortnet/LibSortNet::sort*` | `@solady/LibSort::insertionSort` | `@solady/LibSort::sort` |
| ---- | ---------------------------- | -------------------------------- | ----------------------- |
| 2 | 55 | 298 | 475 |
| 3 | 153 | 489 | 848 |
| 4 | 268 | 709 | 1190 |
| 5 | 445 | 970 | 1485 |
| 6 | 627 | 1257 | 1796 |
| 7 | 837 | 1577 | 2126 |
| 8 | 1076 | 1940 | 2495 |
| 9 | 1331 | 2323 | 2907 |
| 10 | 1548 | 2731 | 3321 |
| 11 | 2376 | 3194 | 3771 |
| 12 | 2724 | 3690 | 4287 |
Data
Compiled with:
```toml
solc = "0.8.25"
optimizer = true
optimizer_runs = 20_000
via-ir = true
```
```
[PASS] testBench_sort10() (gas: 79892301)
Logs:
LibSortNet::sort10 (min: 1301, max: 1747, ~: 1548)
LibSort::insertionSort (min: 1544, max: 3916, ~: 2731)
LibSort::sort (min: 2187, max: 4465, ~: 3321)
[PASS] testBench_sort11() (gas: 82209683)
Logs:
LibSortNet::sort11 (min: 2038, max: 2795, ~: 2376)
LibSort::insertionSort (min: 1925, max: 4640, ~: 3194)
LibSort::sort (min: 2464, max: 5195, ~: 3771)
[PASS] testBench_sort12() (gas: 83912304)
Logs:
LibSortNet::sort12 (min: 2242, max: 3083, ~: 2724)
LibSort::insertionSort (min: 2137, max: 4970, ~: 3690)
LibSort::sort (min: 2895, max: 5973, ~: 4287)
[PASS] testBench_sort2() (gas: 69401423)
Logs:
LibSortNet::sort2 (min: 40, max: 71, ~: 55)
LibSort::insertionSort (min: 268, max: 327, ~: 298)
LibSort::sort (min: 397, max: 553, ~: 475)
[PASS] testBench_sort3() (gas: 70648831)
Logs:
LibSortNet::sort3 (min: 118, max: 179, ~: 153)
LibSort::insertionSort (min: 397, max: 582, ~: 489)
LibSort::sort (min: 463, max: 1103, ~: 848)
[PASS] testBench_sort4() (gas: 71668024)
Logs:
LibSortNet::sort4 (min: 196, max: 347, ~: 268)
LibSort::insertionSort (min: 526, max: 904, ~: 709)
LibSort::sort (min: 529, max: 1479, ~: 1190)
[PASS] testBench_sort5() (gas: 72985853)
Logs:
LibSortNet::sort5 (min: 353, max: 556, ~: 445)
LibSort::insertionSort (min: 655, max: 1293, ~: 970)
LibSort::sort (min: 595, max: 1928, ~: 1485)
[PASS] testBench_sort6() (gas: 74107595)
Logs:
LibSortNet::sort6 (min: 469, max: 769, ~: 627)
LibSort::insertionSort (min: 784, max: 1749, ~: 1257)
LibSort::sort (min: 933, max: 2444, ~: 1796)
[PASS] testBench_sort7() (gas: 75553028)
Logs:
LibSortNet::sort7 (min: 670, max: 1000, ~: 837)
LibSort::insertionSort (min: 1031, max: 2272, ~: 1577)
LibSort::sort (min: 1479, max: 2893, ~: 2126)
[PASS] testBench_sort8() (gas: 76872860)
Logs:
LibSortNet::sort8 (min: 889, max: 1303, ~: 1076)
LibSort::insertionSort (min: 1101, max: 2736, ~: 1940)
LibSort::sort (min: 1732, max: 3435, ~: 2495)
[PASS] testBench_sort9() (gas: 78513806)
Logs:
LibSortNet::sort9 (min: 1113, max: 1523, ~: 1331)
LibSort::insertionSort (min: 1356, max: 3192, ~: 2323)
LibSort::sort (min: 1963, max: 3977, ~: 2907)
```
## Safety
This is **experimental software** and is provided on an "as is" and "as available" basis.
We **do not give any warranties** and **will not be liable for any loss** incurred through any use of this codebase.
Please always include your own thorough tests when using Sortnet to make sure it works correctly with your code.
## Acknowledgements
[hoytech](https://github.com/hoytech) and [Solady](https://github.com/Vectorized/solady)