This project is a minimal implementation of the nbgallery server and APIs in order to understand the set of endpoints in nbgallery/controllers and facilitate development of Jupyter notebook extensions.

I am not 100% comfortable reading/writing/understanding either Javascript or Ruby, and if you aren't either, than this project might help you better understand the nbgallery server and the Jupyter nbextensions that connect Jupyter to the Gallery.

Endpoints

So far in this package the endpoints defined are:

• GET to /, returns links to /environments and /notebooks
• GET to /environments, returns a json list of environments ({name, url})
• POST to /environments with {name, url} to create a new environment in environments.json table
• GET to /notebooks, returns an html list of notebooks
• GET to /notebooks/<uuid>, returns a json blob of notebook metadata and content
• POST to /stages with {title, description, notebook}, creates a new notebook in notebooks.json table and returns a uuid

Databases

When the Docker container is first started, it creates two tinydb tables (environments.json and notebooks.json) with example records. The databases are stored in the same container as the webserver for this minimal example project (unlike nbgallery).

Docker container

To run this Docker container in a development / interactive mode, clone this repo and build/run the Docker file.

git clone https://github.com/nbgallery/nbgallery-api-python
cd nbgallery-api-python
docker build -t nbgallery_api . && docker run -it -p 5000:5000 nbgallery_api

Test notebook

Once the Docker container is running, the API test notebook.ipynb should work to get a list of environments, register a new environment in the database, get a list of notebooks, and create a new notebook in the database.