fix: bring digest generation in sync with HSReplay.net #106

	name: Continuous Integration

	on: [push]

	jobs:
	tox:
	strategy:
	fail-fast: false
	matrix:
	toxenv: [py39-django32, flake8, mypy]
	name: Tox
	runs-on: ubuntu-latest
	env:
	TOXENV: ${{matrix.toxenv}}
	services:
	postgres:
	image: postgres:10.8
	env:
	POSTGRES_DB: postgres
	POSTGRES_USER: postgres
	POSTGRES_PASSWORD: ""
	ports:
	- 5432/tcp
	# needed because the postgres container does not provide a healthcheck
	options: --health-cmd pg_isready --health-interval 10s --health-timeout 5s --health-retries 5
	steps:
	- uses: actions/checkout@v2
	- uses: actions/setup-python@v2
	with:
	python-version: 3.9
	- name: Install dependencies
	run: \|
	python -m pip install --upgrade pip
	python setup.py install
	pip install tox
	- name: Prepare database
	env:
	PGPORT: ${{ job.services.postgres.ports['5432'] }}
	run: python manage.py migrate --settings tests.settings
	- name: tox (${{matrix.toxenv}})
	env:
	PGPORT: ${{ job.services.postgres.ports['5432'] }}
	run: tox

	deploy:
	runs-on: ubuntu-latest
	if: ${{ github.ref == 'refs/heads/master' }}
	needs: [tox]
	steps:
	- name: Trigger Deploy
	env:
	BUILDTRIGGER_URL: ${{ secrets.BUILDTRIGGER_URL }}
	run: \|
	curl -k -X POST $BUILDTRIGGER_URL

Provide feedback