# 1. Use an official lightweight Python image
FROM python:3.11-slim

# 2. Install system dependencies required for 'unstructured' PDF OCR
RUN apt-get update && apt-get install -y \
  tesseract-ocr \
  poppler-utils \
  libmagic-dev \
  libgl1 \
  libglib2.0-0 \
  libgomp1 \
  libgthread-2.0-0 \
  && rm -rf /var/lib/apt/lists/*

# 3. Create a non-root user (Required by Hugging Face for security)
RUN useradd -m -u 1000 user
USER user
ENV PATH="/home/user/.local/bin:$PATH"
ENV HF_HUB_DISABLE_XET=1 \
  HF_HUB_ETAG_TIMEOUT=10 \
  HF_HUB_DOWNLOAD_TIMEOUT=60

# 4. Set the working directory
WORKDIR /app

# 5. Copy requirements and install them
COPY --chown=user:user requirements.txt .
RUN pip install --no-cache-dir -r requirements.txt

# 6. Copy the rest of your project files
COPY --chown=user:user . .

# 7. Optional ML asset pre-build. Disabled by default so HF builds do not hang
# on unauthenticated model downloads; runtime code falls back safely if needed.
ARG PREBUILD_ML_ASSETS=0
ARG MORPHEUS_BUILD_ASSETS_MODE=skip
RUN if [ "$PREBUILD_ML_ASSETS" = "1" ]; then \
  MORPHEUS_BUILD_ASSETS_MODE=$MORPHEUS_BUILD_ASSETS_MODE python -m backend.core.build_ml_assets ; \
  else \
  echo "Skipping ML asset pre-build"; \
  fi

# 8. Start FastAPI (7860 is the HF standard, but Railway uses $PORT)
ENV PORT=7860
CMD ["sh", "-c", "uvicorn backend.main:app --host 0.0.0.0 --port ${PORT}"]