# MarkItDown API — Dependencies (Lightweight)
markitdown[all]>=0.1.5
fastapi>=0.111.0
uvicorn[standard]>=0.30.0
pydantic>=2.7.0
python-multipart>=0.0.9
httpx>=0.27.0

# OCR — RapidOCR with onnxruntime CPU backend (English + Chinese)
rapidocr-onnxruntime>=1.4.4
onnxruntime>=1.18.0
pillow>=10.0.0

# Scanned-PDF fallback (page rendering before OCR)
pypdfium2>=4.30.0

# Data processing for JSON extraction
pandas>=2.0.0

# spaCy NER extraction for non-tabular file formats
spacy>=3.7.0

# Recursive nested-JSON key extraction
jsonpath-ng>=1.8.0

# CLI banner
pyfiglet>=1.0.2
rich>=13.0.0

# PDF rendering and conversion
pdf2image>=1.17.0
pypdf>=5.1.0

# Structured logging and metrics
structlog>=24.4.0
prometheus-client>=0.21.1

# Pydantic settings
pydantic-settings>=2.7.1

# Rate limiting (optional — requires Redis)
redis[hiredis]>=6.4.0