Fairly Fast & Flexible Reed-Solomon Coding
This library uses a variation Intel's Slicing-by-8 algorithm for the encoding routines, achieving some hefty speeds depending on the code sizes. See Benchmarks.
The core implementation is done in C++ and exported to Python with pybind11.
import ffrs
message = bytearray(b'Hello World!')
# Create encoder/decoder for 8 parity bytes
rs8 = ffrs.RS256(ecc_len=8)
# Allocate space for correction code
data = message + bytearray(rs8.ecc_len)
# Compute parity bytes
rs8.encode(data)
# Corrupt some bytes (up to ecc_len/2)
data[3] ^= 0xaa
data[9] ^= 0x55
# Restore partially corrupted message
rs8.decode(data)
# Remove parity bytes
recovered_message = data[:-rs8.ecc_len]
assert message == recovered_messagegit clone --recursive https://github.com/zydd/libffrs.git
pip install ./libffrsAuto-generated documentation can be found at:
Documentation for this project is generated automatically using Sphinx based on class/method signatures and doc strings.
cd ./libffrs
pip install 'sphinx>=7.2' 'furo==2023.9.10'
make -C ./docs/ htmlDocumentation files will be created in ./build/html
libffrs is licensed under the Apache License, Version 2.0