CompactU16 Readme

Overview

CompactU16 is a compact, variable-length encoding for 16-bit unsigned integers (u16). It is designed to save space when serializing small integers by using fewer bytes for smaller values. The implementation is inspired by the variable-length quantity (VLQ) encoding used in formats like Protocol Buffers.

Features

• Compact Encoding: Efficiently encodes u16 integers using 1 to 3 bytes, depending on the size of the value.
• Optional Borsh Support: When the use-borsh feature is enabled, CompactU16 implements the BorshSerialize and BorshDeserialize traits, allowing it to be used with the Borsh serialization format.
• Error Handling: Provides detailed error handling to catch cases where the encoded value exceeds the bounds of u16.

Example Usage

Deserialize CompactU16 from Bytes

use compact_u16::CompactU16;

let mut input = &[0x80, 0x01][..];
let value = CompactU16::deserialize(&mut input).unwrap();
assert_eq!(value.0, 0x0080);

Serialize CompactU16 to Bytes

use compact_u16::CompactU16;
use borsh::BorshSerialize;

let value = CompactU16(0x3fff);
let mut output = Vec::new();
value.serialize(&mut output).unwrap();
assert_eq!(output, vec![0xff, 0x7f]);

Details

Encoding

• Values from 0x0000 to 0x007F are encoded in a single byte.
• Values from 0x0080 to 0x3FFF are encoded in two bytes.
• Values from 0x4000 to 0xFFFF are encoded in three bytes.

Borsh Integration

The CompactU16 struct supports Borsh serialization and deserialization when the use-borsh feature is enabled. This allows easy integration with systems that use the Borsh format.

Testing

Unit tests are included to verify the correctness of the encoding and decoding. These tests cover a range of values and ensure that the encoding and decoding process is accurate.

Running Tests

cargo test

License

This project is licensed under the MIT License. See the LICENSE file for details.