How do you use environment variables and Docker's --env-file (or docker-compose's environment section) to configure a containerized Node.js application per environment?

Rather than baking environment-specific configuration into the image itself, values are injected at container runtime via -e flags, an --env-file, or docker-compose.yml's environment/env_file sections -- keeping the same built image usable unchanged across development, staging, and production, with only the externally supplied environment variables differing, consistent with the twelve-factor app methodology.

# docker-compose.yml
services:
  app:
    image: my-app:latest
    env_file: .env.production
    environment:
      - NODE_ENV=production
    ports:
      - '3000:3000'