A demonstration of the different ways inputs can be provided in a Refinery visualization tool, and, potentially, a starting point for other simple tools.
build_run_test.sh encapsulates the whole process of building and deploying a visualization tool.
- In a real application, the tool image would already be built and would be available on DockerHub,
- and, in a real application, django_docker_engine or some other API would be used to start the containers, rather than just using the Docker CLI.
$ pip install -r requirements.txt
$ ./build_run_test.sh`
...
OK
browse: http://localhost:32787/
clean up: docker ps -qa | xargs docker stop | xargs docker rm
PASS!
At the end, you can hit the server it spun up for a manual check. It should show JSON from three different sources.
Successful Github tags and PRs will prompt Travis to push the built image to Dockerhub. For a new version number:
git tag v0.0.x && git push origin --tags