pydantic>=2.0.0
pyyaml>=6.0
click>=8.0.0
rich>=13.0.0
typing-extensions>=4.0.0

[all]
chunking-strategy[detection,dev,document,hardware,ml,multimedia,text,tika,vectordb]

[detection]
python-magic>=0.4.27
chardet>=5.2.0
langdetect>=1.0.9

[dev]
pytest>=7.0.0
pytest-cov>=4.0.0
pytest-asyncio>=0.21.0
black>=23.0.0
ruff>=0.1.0
isort>=5.12.0
mypy>=1.5.0
pre-commit>=3.0.0

[document]
PyPDF2>=3.0.0
pdfminer.six>=20220524
PyMuPDF>=1.23.0
python-docx>=0.8.11
beautifulsoup4>=4.12.0
lxml>=4.9.0
markdown>=3.5.0

[hardware]
psutil>=5.9.0
GPUtil>=1.4.0
nvidia-ml-py3>=7.352.0

[ml]
scikit-learn>=1.3.0
numpy>=1.24.0
torch>=2.0.0
transformers>=4.30.0
bitsandbytes>=0.47.0
fsspec<2025.0.0,>=2023.1.0

[multimedia]
opencv-python>=4.8.0
Pillow>=10.0.0
librosa>=0.10.0
pydub>=0.25.0
imageio>=2.31.0
moviepy>=1.0.3

[text]
spacy>=3.6.0
nltk>=3.8
sentence-transformers>=5.1.0
huggingface-hub>=0.20.0
tiktoken>=0.5.0

[tika]
tika>=2.6.0
python-magic>=0.4.27

[vectordb]
qdrant-client>=1.6.0
weaviate-client>=3.24.0
pinecone-client>=2.2.0
