## GPU-ready base image with CUDA 12 + cuDNN 9 runtime
# If you don't have an NVIDIA GPU or the NVIDIA Container Toolkit, this image still runs on CPU.
# For smaller CPU-only images, you can switch back to python:3.11-slim.
# Base image is configurable so you can choose CPU-only or CUDA+cuDNN runtime.
# Default to Python slim for maximum compatibility; override via build-arg
# BASE_IMAGE (or docker-compose build args) to use an NVIDIA CUDA runtime.
ARG BASE_IMAGE=python:3.11-slim
FROM ${BASE_IMAGE}

# Keep python fast/quiet and pip lean
ENV PYTHONDONTWRITEBYTECODE=1 \
    PYTHONUNBUFFERED=1 \
    PIP_NO_CACHE_DIR=1 \
    LANG=C.UTF-8 \
    LC_ALL=C.UTF-8 \
    # sensible defaults (can be overridden by .env)
    WHISPER_MODEL=large-v3 \
    WHISPER_PRECISION=int8

# System deps: ffmpeg for media, curl for healthcheck, jq for scripts, poppler-utils for PDFs
RUN apt-get update && apt-get install -y --no-install-recommends \
    python3 python3-pip python3-venv \
    ffmpeg \
    curl \
    jq \
    poppler-utils \
 && rm -rf /var/lib/apt/lists/*

WORKDIR /app

# Upgrade pip toolchain then install Python deps
COPY requirements.txt .
RUN python3 -m pip install --upgrade pip setuptools wheel \
 && pip3 install --no-cache-dir -r requirements.txt \
 && pip3 check || true

# App code
COPY app.py worker.py scanner.py ./
RUN pip3 install --no-cache-dir gunicorn==22.0.0

# Healthcheck against the app's /health endpoint

EXPOSE 8080
CMD ["gunicorn", "-b", "0.0.0.0:8080", "app:app", "--workers", "2", "--threads", "4"]