Dockerfile

# Base image with Python and common dependencies
FROM python:3.10-slim

# Set working directory
WORKDIR /app

# Set environment variables
ENV DEBIAN_FRONTEND=noninteractive \
    PYTHONUNBUFFERED=1 \
    PYTHONDONTWRITEBYTECODE=1 \
    TRANSFORMERS_CACHE=/app/cache \
    HF_HOME=/app/cache

# COPY requirements file and install Python dependencies
COPY requirements.txt requirements.txt 
COPY extract_img_pdf.py extract_img_pdf.py
COPY live_streaming_flask.py live_streaming_flask.py
COPY templates/ /app/templates 
COPY .env .env
COPY test_streaming.py test_streaming.py 
COPY app_main.py app_main.py

# Install system dependencies
RUN apt-get update && apt-get install -y \
    build-essential \
    libglib2.0-0 \
    libsm6 \
    libxext6 \
    libxrender-dev \
    tesseract-ocr \
    poppler-utils \
    libgl1 \
    ffmpeg \
    libopencv-dev \
    && apt-get clean && rm -rf /var/lib/apt/lists/*

RUN pip install -r requirements.txt
# Fontconfig Warning Suppression (optional)
ENV FONTCONFIG_PATH=/etc/fonts
ENV FONTCONFIG_FILE=/etc/fonts/fonts.conf

# Set writable path for nltk data
ENV NLTK_DATA=/app/nltk_data
RUN mkdir -p /app/nltk_data

# Pre-download required NLTK models
RUN python -m nltk.downloader -d /app/nltk_data \
    punkt averaged_perceptron_tagger averaged_perceptron_tagger_eng

RUN mkdir -p /app/cache /app/data && chmod -R 777 /app/cache /app/data
RUN mkdir -p /app/OUTPUTS
RUN chmod -R 777 /app

RUN mkdir -p outputs/DETECTED_IMAGE outputs/SCANNED_IMAGE outputs/EXTRACTED_JSON

# Expose the required port for HF Spaces
EXPOSE 7860

# Set the command to run your Flask app
# CMD ["python", "extract_img_pdf.py"]
# CMD ["python", "live_streaming_flask.py"]
CMD ["python", "app_main.py"]