Dockerfile.unstructured 927 B

1234567891011121314151617181920212223242526272829
  1. FROM python:3.12-slim AS builder
  2. # Install system dependencies (including those needed for Unstructured and OpenCV)
  3. RUN apt-get update && apt-get install -y --no-install-recommends \
  4. gcc g++ musl-dev curl libffi-dev gfortran libopenblas-dev \
  5. tesseract-ocr libtesseract-dev libleptonica-dev pkg-config \
  6. poppler-utils libmagic1 pandoc libreoffice \
  7. libgl1-mesa-glx libglib2.0-0 \
  8. && apt-get clean && rm -rf /var/lib/apt/lists/*
  9. ENV TESSDATA_PREFIX=/usr/share/tesseract-ocr/5/tessdata
  10. ENV PYTHONDONTWRITEBYTECODE=1
  11. ENV PYTHONUNBUFFERED=1
  12. WORKDIR /app
  13. RUN pip install --no-cache-dir unstructured "unstructured[all-docs]"
  14. RUN python -c "from unstructured.partition.model_init import initialize; initialize()"
  15. RUN pip install gunicorn uvicorn fastapi httpx
  16. COPY core/integrations/unstructured/main.py .
  17. EXPOSE 7275
  18. CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "7275", "--workers", "8"]