Skip to content

jupyterlab/jupyterlab-telemetry

jupyterlab_telemetry

Github Actions Status

A JupyterLab library for logging and telemetry of usage data

This extension is composed of a Python package named jupyterlab_telemetry for the server extension and a NPM package named @jupyterlab/jupyterlab-telemetry for the front end library.

Requirements

  • JupyterLab >= 3.0

Install

To install the server extension, execute:

pip install jupyterlab_telemetry

To install the front end library, add "@jupyterlab/jupyterlab-telemetry": "^1.0.0" to dependencies in package.json.

Uninstall

To remove the server extension, execute:

pip uninstall jupyterlab_telemetry

To remove the front end library, remove "@jupyterlab/jupyterlab-telemetry": "^1.0.0" from dependencies in package.json.

Troubleshoot

If you have installed the frontend library, but it is not able to connect to telemetry endpoint, check that the server extension is enabled:

jupyter server extension list

Usage

Define handlers that can receive events from the EventLog

import { EventLog } from '@jupyterlab/jupyterlab-telemetry';

function consoleHandler(el: EventLog, events: EventLog.RecordedEvent[]) {
  console.log(`[Handler1] Received events ${JSON.stringify(events)}`);
}

function consoleHandler2(el: EventLog, events: EventLog.RecordedEvent[]) {
  console.log(`[Handler2] Received events ${JSON.stringify(events)}`);
}

Create an instance of the EventLog and configure the handler and other options from EventLog.IOptions

const el = new EventLog({
  handlers: [consoleHandler, consoleHandler2],
  allowedSchemas: [
    'org.jupyter.foo',
    'org.jupyterlab.commands.docmanager:open'
  ],
  commandRegistry: app.commands,
  commandEmitIntervalSeconds: 2
});

Send custom events via the recordEvents interface. If the commandRegistry instance was passed, then the EventLog will subscribe to commands executed in the JupyterLab application and send the whitelisted ones to each configured handler.

el.recordEvent({
  schema: 'org.jupyter.foo',
  version: 1,
  body: {
    foo: 'bar'
  }
});

Dispose the event log after use

el.dispose();

Contributing

Development install

Note: You will need NodeJS to build the extension package.

The jlpm command is JupyterLab's pinned version of yarn that is installed with JupyterLab. You may use yarn or npm in lieu of jlpm below.

# Clone the repo to your local environment
git clone https://github.com/jupyterlab/jupyterlab-telemetry.git

# Change directory to the jupyterlab-telemetry directory
cd jupyterlab-telemetry

# Install package in development mode
pip install -e .

# Server extension must be manually installed in develop mode
jupyter server extension enable jupyterlab_telemetry

# Build the Typescript source after making changes
jlpm build

Testing

The server extension has python tests which can be updated and tested before pushing the new changes.

# Install the test dependencies
pip install .[test]

# Run the tests
pytest

The front end library is using jest for testing, and is setup to produce test coverage when the test target is run.

# To run the tests with coverage
jlpm test

Development uninstall

# Server extension must be manually disabled in develop mode
jupyter server extension disable jupyterlab_telemetry
pip uninstall jupyterlab_telemetry

Packaging the extension

See RELEASE