Block Hashing for Large Files

A custom cryptographic hash function algorithm built from scratch in Python, designed to process large files (>1MB) deterministically in blocks with chaining mechanisms, outputting a unique fixed-length 256-bit integrity value.

The decision for the name aguirre_arias_galvis_hash.py was a simple creative process of joining the first surnames of the three authors who developed the idea for the encryption process: Diego Fernando Aguirre Tenjo, Camilo Andres Arias Tenjo, and Katlyn Jennelis Galvis Rodriguez.

---

Main Features

• Custom Compression Function: 32-round mixing with XOR operations, bitwise rotations, and modular additions for secure data diffusion.
• Block Processing Architecture: Reads files in 512-bit (64-byte) blocks to handle large files efficiently without memory overflow.
• Merkle-Damgård Padding: Implements length-extension attack prevention with deterministic padding scheme.
• Avalanche Effect Validation: Automatically verifies that 1-bit input changes produce ~50% output bit changes.
• Performance Benchmarking: Compares execution times against Python's native SHA-256 implementation.
• Automated Test Generation: Creates 1MB, 5MB, and 10MB test files with controlled modifications for validation.

---

Execution and Development Guide

1. Clone the Repository

git clone https://github.com/CamiloAT/aguirre-arias-galvis-hash.git
cd aguirre-arias-galvis-hash

2. Requirements

• Python 3.8 or higher
• Standard built-in Python libraries (os, time, hashlib, struct, json, random)
• No external dependencies or pip installs required
• Multi-platform (Windows, macOS, Linux)

3. Run the Program

python aguirre_arias_galvis_hash.py

Note: The program will automatically generate test files, compute hashes, validate the avalanche effect, and benchmark against SHA-256. All results are displayed in the console.

---

Project Structure

aguirre-arias-galvis-hash/
├── aguirre_arias_galvis_hash.py  ← Main algorithm implementation
├── README.md                     ← Project documentation
└── .gitignore                    ← Git ignore rules

---

Tech Stack

Layer	Technology
Language	Python 3.8+
Hashing	Custom block cipher (512-bit blocks, 256-bit output)
Compression	Davies-Meyer construction with 32-round mixing
Padding	Merkle-Damgård compatible scheme
Testing	Automated file generation and validation
Benchmarking	Python hashlib.sha256 comparison

---

Authors

Name	GitHub
Diego Fernando Aguirre Tenjo	@elcokiin
Camilo Andres Arias Tenjo	@CamiloAT
Katlyn Jennelis Galvis Rodriguez	@Katlyng

Cryptography & Algorithm Development