Arabic NLP Toolkit

A complete, production-ready Python package for processing Arabic text following MLOps best practices. Designed for use in machine learning pipelines, it includes utilities for cleaning, normalization, tokenization, and basic sentiment analysis.

Features

• Text Cleaning: Strip diacritics, remove special characters, and normalize Arabic text.
• Tokenization: Simple and rule-based tokenizers and sentence splitters tailored for Arabic text.
• Sentiment Analysis: Basic rule-based sentiment scoring (positive, negative, neutral).
• MLOps Ready: Designed with modularity, testing, and continuous integration in mind.

Installation

From Source

Ensure you have Python 3.8+ installed. You can install it directly by cloning the repository and running:

git clone https://github.com/Melad98/MLOps-package-structure.git
cd MLOps-package-structure
pip install .

For development (including dependencies for testing and building):

pip install -e ".[dev]"

Usage Example

from arabic_nlp_toolkit.cleaning import normalize_arabic

text = normalize_arabic("النّص العَرَبِيُّ!")
print(text)
# Output might be "النص العربي!" depending on the normalization functions

Testing

To run the unit tests, install development requirements and use pytest:

pip install -e ".[dev]"
pytest tests/

Packaging

Ensure your package is buildable using build:

python -m build