Solidity Library

Stop paying for resolution you don't need.

A Solidity library for shift-based quantization of unsigned integers, enabling tighter struct packing and lower gas costs.

View on GitHub Documentation Playground

Solidity EVM Gas Optimization MIT License

Quick Start

Terminal
forge soldeer install uint-quantization-lib

MyContract.sol
import {Quant, UintQuantizationLib} from "uint-quantization-lib/src/UintQuantizationLib.sol";

Quant private immutable SCHEME = UintQuantizationLib.create(32, 24);

uint24 stored = uint24(SCHEME.encode(largeValue));   // compress
uint256 restored = SCHEME.decode(stored);             // decompress

How It Works

Original

1e18 wei

kept: 224 bits discarded: 32 bits

256 bits total (1 token, 18 decimals)

>> 32

encode

Stored

232,830,643 wei

fits in uint224

<< 32

decode

Restored

≈ 1e18 wei

preserved: 224 bits zero-filled: 32 bits

max error < 2^32 wei (~4.3 gwei)

Ask an AI

Explain how the uint-quantization-lib Solidity library (https://github.com/0xferit/uint-quantization-lib) works. How does its shift-based quantization reduce gas costs, and what are the tradeoffs?

API

create(discardedBitWidth, encodedBitWidth)Create a quantization scheme

encode(value)Floor-encode a value (reverts on overflow)

encode(value, true)Strict encode (reverts if not step-aligned)

decode(encoded) / decodeMax(encoded)Restore lower-bound or upper-bound value

fits(value) / fitsEncoded(encoded)Check if value is representable

isAligned(value) / remainder(value)Check step alignment and lost bits

requireAligned(value) / requireMinStep(value)Guard helpers that revert on bad input

floor(value) / ceil(value)Snap to nearest step boundary

stepSize() / max()Query scheme parameters