#!/bin/sh set -e set -u # DEBUG set -x check_app_is_there() { if [ ! -f "./manage.py" ]; then usage fi } deploy() { # detect if existing deployment (TODO only works with sqlite) if [ -f "${program_dir}/db/db.sqlite3" ]; then echo "INFO: detected EXISTING deployment" ./manage.py migrate else # move the migrate thing in docker entrypoint # inspired by https://medium.com/analytics-vidhya/django-with-docker-and-docker-compose-python-part-2-8415976470cc echo "INFO detected NEW deployment" ./manage.py migrate ./manage.py add_user user@example.org 1234 fi } runserver() { PORT="${PORT:-8000}" if [ "${DEBUG:-}" = "true" ]; then ./manage.py runserver 0.0.0.0:${PORT} else # TODO #./manage.py collectstatic true if [ "${EXPERIMENTAL:-}" = "true" ]; then # TODO # reloading on source code changing is a debugging future, maybe better then use debug # src https://stackoverflow.com/questions/12773763/gunicorn-autoreload-on-source-change/24893069#24893069 # gunicorn with 1 worker, with more than 1 worker this is not expected to work #gunicorn --access-logfile - --error-logfile - -b :${PORT} trustchain_idhub.wsgi:application true else ./manage.py runserver 0.0.0.0:${PORT} fi fi } main() { program_dir='/opt/devicehub-django' cd "${program_dir}" deploy runserver } main "${@}"