requirements.txt

# PDF handling
PyMuPDF                   # provides fitz; replace ‘fitz’ to avoid the frontend conflict :contentReference[oaicite:0]{index=0}  
pdf2image                 # PDF→PIL images wrapper :contentReference[oaicite:1]{index=1}  

# OCR & layout
paddleocr                 # PaddleOCR toolkit :contentReference[oaicite:2]{index=2}  
paddlepaddle              # Paddle backend for PaddleOCR  

# Table extraction
camelot-py[base]          # Camelot’s core + cv dependencies :contentReference[oaicite:3]{index=3}  

# Data processing
numpy  
pandas  

# NLP & ML
spacy  
transformers  
torch  
tqdm  

# Vision (if using any OpenCV ops)
opencv-python  

# HTML parsing (if needed)
beautifulsoup4  

# System dependency wrapper (note: poppler-utils is a system package, not pip)
# poppler-utils          ← install via apt/conda, not pip

# Install spaCy small English model
en_core_web_sm @ https://huggingface.co/spacy/en_core_web_sm/resolve/main/en_core_web_sm-any-py3-none-any.whl