NbExchange JupyterLab Plugin

A JupyterLab extension which provides the plugins for nbgrader to use NbExchange as an external Exchange service.

It is composed of a Python package named nbexchange_jlab for the server extension and a NPM package named nbexchange_jlab for the frontend extension.

Additional Functionality

In addition to the usual suite of plugins for exchanging files, the plugin provides a History option to the Nbgrader menu - this is a visual render of NbExchanges history endpoint:

The current course is shown in green, and each course indicates the first and last date of any activity for the whole course

Requirements

• JupyterLab >= 4.0.0
• NbGrader

Installation

The primary reference for this should be the nbgrader documentation - but in short:

1. Install nbexchange_jlab_plugin into your jupyter environment [from github, using pip]

• This will also install nbgrader - this release installs 0.9.5 (which makes it compatible with JupyterLab & Notebook 7)

3. Include the following in your nbgrader_config.py file:

c.ExchangeFactory.exchange = 'nbexchange_jlab.plugins.Exchange'
c.ExchangeFactory.list = 'nbexchange_jlab.plugins.ExchangeList'
c.ExchangeFactory.release_assignment = 'nbexchange_jlab.plugins.ExchangeReleaseAssignment'
c.ExchangeFactory.fetch_assignment = 'nbexchange_jlab.plugins.ExchangeFetchAssignment'
c.ExchangeFactory.submit = 'nbexchange_jlab.plugins.ExchangeSubmit'
c.ExchangeFactory.collect = 'nbexchange_jlab.plugins.ExchangeCollect'
c.ExchangeFactory.release_feedback = 'nbexchange_jlab.plugins.ExchangeReleaseFeedback'
c.ExchangeFactory.fetch_feedback = 'nbexchange_jlab.plugins.ExchangeFetchFeedback'

(note the change to plugins, plural)

These plugins will also check the size of releases & submissions

c.Exchange.max_buffer_size = 204800 # 200KB

[or even a more specific c.ExchangeSubmit.max_buffer_size = 204800 # 200KB]

By default, upload sizes are limited to 5GB (5253530000)

The figure is bytes

Configuring the plugins to talk to the NbExchange server

The plugins make http requests to the server, which requires it to prepare several things:

• base_service_url is the http origin of the NbExchange service - we default this to https://noteable.edina.ac.uk, 'cos.... advertising
• base_path is the path-part of requests into the exchange. This needs to match base_url defined in the NbExchange service, and (likewise) defaults to /services/nbexchange/.
• api_plugin_class is the name of the class that sets up headers, cookies, etc for the api_request to call the external exchange
  • Whatever is used here needs to match whatever the nbexchange get_current_user will use to identify the user.
  • This see the test_plugin_exchange_with_bespoke_apiPlugin.py test file for an simplistic example using JSON Web Tokens

eg:

from nbexchange_jlab.plugins import BaseApiPlugin

class AuthApiPlugin(BaseApiPlugin):
    def prep_api_call(self, path):

        # Gets an HTTP Authentication Token
        auth_token = _some_magic_function()
        cookies = dict()
        headers = dict()

        if auth_token:
            headers["Authorization"] = f"Basic {auth_token}"

        url = self.service_url() + path
        return url, cookies, headers
c.Exchange.api_plugin_class = AuthApiPlugin
c.Exchange.base_service_url = 'https://nbexchange.example.com'
c.Exchange.base_path = '/services/exchange'

Contributing

See Contributing.md for details on how to extend/contribute to the code.