Skip to content

lchapo/dash-google-auth

Repository files navigation

Deprecation Warning

This project is no longer actively maintained and is incompatible with newer versions of Dash (>=1.0). Please consider forking this repo or see the Dash docs on officially supported authentication mechanisms.

Dash Google Auth

Dash Google Auth is a simple library using Google OAuth to authenticate and view a Dash app.

This Library uses Flask Dance and a modified version of Plotly's own dash auth for authentication.

Basic Use

Authentication can be added to your Dash application using the GoogleOAuth class, i.e.

from dash import Dash
from flask import Flask
from dash_google_auth import GoogleOAuth

server = Flask(__name__)
server.config.update({
  'GOOGLE_OAUTH_CLIENT_ID': ...,
  'GOOGLE_OAUTH_CLIENT_SECRET': ...,
})

app = Dash(__name__, server=server, url_base_pathname='/', auth='auth')

authorized_emails = [...]
additional_scopes = [...]
auth = GoogleOAuth(app, authorized_emails, additional_scopes)

# your Dash app here :)
...

Example

Steps to try this out yourself:

  1. Install the dash-google-auth library using pip:

    $ pip install dash-google-auth
  2. Follow the Flask Dance Guide to create an app on the google admin console

  3. Make a copy of app.py and set the variables (or set the corresponding environment variables):

    server.config["GOOGLE_OAUTH_CLIENT_ID"] = ...
    server.config["GOOGLE_OAUTH_CLIENT_SECRET"] = ...

    with values from the Google OAuth 2 client you should have set up in step 1. If you've set these up properly, you can find them at APIs & Services > Credentials under the section OAuth 2.0 client IDs.

  4. Replace authorized_emails in app.py with whatever Google emails you want to grant access to your app. In production, I'd recommend getting these from a database instead.

  5. Run python app.py and open localhost in a browser window to try it out! If the app loads automatically without prompting a Google login, that means you're already authenticated -- try using an incognito window in this case if you want to see the login experience for a new user.