Dockerfile

FROM python:3.11.6-slim

# Set environment variables
ENV PYTHONUNBUFFERED=1 \
    PYTHONPATH="/app" \
    DEBIAN_FRONTEND=noninteractive \
    STREAMLIT_SERVER_ADDRESS=0.0.0.0 \
    STREAMLIT_SERVER_PORT=7860

# Install system dependencies
RUN apt-get update && apt-get install -y \
    build-essential \
    libglib2.0-0 \
    libsm6 \
    libxrender1 \
    libxext6 \
    git \
    curl \
    wget \
    procps \
    net-tools \
    && apt-get clean \
    && rm -rf /var/lib/apt/lists/*

# Create non-root user for security
RUN groupadd -r appuser && useradd -r -g appuser appuser

# Set working directory
WORKDIR /app

# Copy requirements first for better caching
COPY requirements.txt /app/
RUN pip install --no-cache-dir --upgrade pip && \
    pip install --no-cache-dir -r requirements.txt

# Copy path configuration first
COPY path_config.py /app/

# Copy validation script
COPY docker_validation.py /app/

# Copy project files
COPY . /app

# Create necessary directories with proper permissions
# The path_config.py will handle environment-specific paths
RUN mkdir -p /app/data /app/model /app/logs /app/cache /app/temp && \
    chmod -R 755 /app/data /app/model /app/logs /app/cache /app/temp

# Ensure logs directory is writable
RUN chmod 777 /app/logs || echo "Could not set logs permissions, will use fallback logging"

# Make scripts executable
RUN chmod +x /app/start.sh

# Copy health check script and make it executable
COPY health_check.sh /app/
RUN chmod +x /app/health_check.sh

# Copy initial datasets if they exist to the correct app structure
# The system will use /app/data/* directly instead of copying to /tmp
RUN if [ -d /app/data/kaggle ]; then \
        echo "Kaggle datasets found in app structure"; \
    fi && \
    if [ -f /app/data/combined_dataset.csv ]; then \
        echo "Combined dataset found in app structure"; \
    fi

# Initialize system with the new path structure
RUN python /app/initialize_system.py

# Test the system initialization
RUN python3 -c "\
import sys; \
sys.path.append('/app'); \
from path_config import path_manager; \
print(f'Container environment: {path_manager.environment}'); \
print(f'Base directory: {path_manager.base_paths[\"base\"]}'); \
print(f'Data directory: {path_manager.base_paths[\"data\"]}'); \
print(f'Model directory: {path_manager.base_paths[\"model\"]}'); \
critical_files = [path_manager.get_combined_dataset_path(), path_manager.get_model_file_path(), path_manager.get_vectorizer_path()]; \
[print(f'✅ {file_path}' if file_path.exists() else f'❌ {file_path}') for file_path in critical_files]; \
print('System check completed')"

# Health check using the proper paths
HEALTHCHECK --interval=30s --timeout=10s --start-period=60s --retries=3 \
    CMD /app/health_check.sh

# Change ownership to appuser
RUN chown -R appuser:appuser /app

# Switch to non-root user
USER appuser

# Expose ports
EXPOSE 7860 8000

# Set environment variable to help the app detect container environment
ENV DOCKER_CONTAINER=1

# Run the startup script
CMD ["./start.sh"]