rudra-512-hash 11.10.11

RUDRA512: A 512-bit cryptographic hash function for research and experimental analysis.

RUDRA512

RUDRA512 is a 512-bit cryptographic hash function designed as a research-oriented construction combining ARX-based mixing with structured preprocessing, including tokenisation and input scattering.

⚠️ Warning: This implementation is intended for research and experimentation only. It is not a production-ready cryptographic primitive and should not be used in security-critical applications.

---

✨ Features

• 512-bit hash output
• ARX (Add-Rotate-XOR) based internal mixing
• Structured preprocessing pipeline
• BPE-based tokenisation (fixed vocabulary)
• Data-dependent input scattering
• Deterministic and reproducible design

---

📦 Installation

Install via pip:

pip install rudra512

Or install locally:

git clone https://github.com/Developer-Ayush/rudra512.git
cd rudra512
pip install .

---

🚀 Usage

Basic Example

from rudra512 import rudra512

hash_value = rudra512("hello world")
print(hash_value)

---

CLI Usage (if supported)

python -m rudra512 "hello world"

---

🧪 Test Vectors

"" → <hash_output_here>
"abc" → <hash_output_here>
"hello world" → <hash_output_here>

(Replace with actual outputs from your implementation)

---

🔒 Design Overview

RUDRA512 consists of:

1. Tokenisation Layer

   • Uses a fixed BPE vocabulary
   • Vocabulary is frozen and versioned

2. Input Scattering

   • Data-dependent permutation of token sequence
   • Reduces structured input control

3. ARX Mixing Core

   • Multi-round transformation
   • Combines addition, rotation, and XOR

4. Finalisation

   • Produces a 512-bit digest

---

📚 Specification & Paper

Full specification and analysis available in the paper:

• RUDRA512: A Structured 512-bit Hash Function with Tokenisation and Input Scattering

(Include your ePrint link after submission)

---

🔁 Reproducibility

• The BPE vocabulary is fixed and included in the repository
• SHA-256 hash of vocabulary file is provided in the paper
• Implementation is deterministic across platforms

---

⚠️ Security Notice

• No formal security proof
• No full differential or linear cryptanalysis
• Performance significantly slower than standard hash functions such as SHA-512 and BLAKE2b

👉 This project is intended to encourage community cryptanalysis and research discussion.

---

🤝 Contributing

Contributions, feedback, and cryptanalysis are welcome.

---

📄 License

This project is licensed under the Apache-2.0.

---

👤 Author

Ayush Anand Independent Researcher

---

⭐ Acknowledgment

This work is released to encourage exploration of alternative hash function design approaches.