installing (in a python 3.11+ environment):
pip install "uvicorn[standard]"
pip install -e .
running:
uvicorn sciwyrm.main:app --reload
access:
navigate browser to:
http://localhost:8000/docs
In order to test, you can submit the following POST request , once you have the service running locally:
curl \
-H 'Content-Type: application/json' \
-d '{
"template_id" : "b32f6992-0355-4759-b780-ececd4957c23",
"parameters" : {
"dataset_pids" : [ "dataset_pid_1" ,"dataset_pid_2" ],
"file_server_host" : "this_file_server",
"file_server_port" : 22,
"scicat_url" : "my.scicat.instance",
"scicat_token" : "my.scicat.token"
}
}' \
-X POST \
http://localhost:8000/notebook
as a one liner, ready to be run, it looks like:
curl -H 'Content-Type: application/json' -d '{ "template_id" : "b32f6992-0355-4759-b780-ececd4957c23", "parameters" : {"dataset_pids":["dataset_pid_1","dataset_pid_2"],"file_server_host": "this_file_server","file_server_port" : 22, "scicat_url":"my.scicat.instance", "scicat_token" : "my.scicat.token"}}' -X POST http://localhost:8000/notebook
This request should result in an populated instance of the template b32f6992-0355-4759-b780-ececd4957c23.ipynb included in the repository.
A docker image is created by a dedicated workflow every time there is a new release by merging a PR in the release branch.
In case a users or developer has the need to create locally a docker image and deploy the container, the following commands can be used, assuming that image name sciwyrm and the container name is scijug:
Docker image creation
docker build -f docker/Dockerfile . -t sciwyrm
Docker container instantiation
docker create --name scijug -p 8000:8000 sciwyrm
The Jupyter creation service will be available on http://localhost:8000
All contributors should create a branch dedicated branch to develop their feature and [performing their work. Once the code is in a ready, the contribution should be done through a PR to the main branch, where maintainers can review, provide feedback and approve it.
Release are managed by maintainers by creating PRs from the main branch to the release branch. Merging a PR in release trigger a git workflow that creates a release in the repository and also the creation of a docker image with the version of the service released.