Drax dashboard

Drax is a python and react implementation of the beautiful dashing app that has minimal dependencies. I created this project to develop my react and tornado skills.

How to use drax

After you installed drax run the following commands:

    $ drax init mydashboard
    $ cd mydashboard
    $ drax start

Now a new dashboard is available on http://localhost:8888/

Every project comes with some base widgets and an example dashboard. The directory is setup as follows:

• Assets - All your images, fonts, and javascript libraries.
• Dashboards - One html file for each dashboard that contains the layout for the widgets. The default dashboard is defined in index.html.
• Jobs - The python jobs for fetching data to sent to your widgets.
• Widgets - All the jsx and css files for individual widgets.

How to create a dashboard

By editing the index.html file in the templates directory:

  React.render(
    <Dashboard widgetHeight={450} widgetWidth={360}>
      <Clock widgetid="clock" row={1} col={1}/>
      <HeartBeat widgetid="hearthbeat" row={1} col={1}/>
      <Text widgetid="mywidget3" row={1} col={4} initialTitle="This is the title" initialText="This is my initial text...."/>
      <List widgetid="mylist" row={1} col={3} sizey={2}/>
      <Number widgetid="valuation" row={2} col={1} initialTitle="Current valuation" initialInfo="In billions"/>
      <Meter widgetid="synergy" initialTitle="Synergy" initialValue={10} min={0} max={100} row={2} col={2}/>
      <Image widgetid="image" row={2} col={4} src="http://dashingdemo.herokuapp.com/assets/logo.png"/>
    </Dashboard>,
    document.getElementById('dashboard'));

If you want to add more dashboards copy index.html to a new html file in the templates directory and edit that file.

If you created a new dashboard status.html it will be available on http://localhost:8888/status (Notice the url without the ".html").

How get data into widgets

Your widgets can be updated directly over HTTP. Post the data you want in json to /widgets/widget_id. Example

curl -d '{ "auth_token": "YOUR_AUTH_TOKEN", "text": "Some new text..."}' http://localhost:8888/widgets/text

or a python example:

import json
import requests

url = 'http://localhost:8888/widgets/text'
data = {
    'auth_token': 'YOUR_AUTH_TOKEN',
    'text': 'Some new text...',
    'id': 'text'
}

requests.post(url, data=json.dumps(data))

How to create a job

Every file with the .py extension in the jobs folder is loaded as a job and executed periodically.

Each job needs to define a function named callback (the actual job) and a variable callback_time that defines the time in milliseconds of the frequency the job is executed.

An example of a job executing every 30 seconds:

import json
from tornado.httpclient import AsyncHTTPClient

# callback function is called every 30 seconds
callback_time = 30000


def get_myapi_data():
    # Add logic to fetch data from somewhere
    data = 'This is some data loaded from somewhere...'
    return data


def callback():
    data = {
        'auth_token': 'YOUR_AUTH_TOKEN',
        # the id of the widget you want to post the data to
        'id': 'text',
        'title': 'Some Info:',
        'icon': 'icon-star',
        'text': get_myapi_data()
    }
    payload = json.dumps(data)
    url = 'http://localhost:8888/widgets/text'

    # Using tornado async client
    client = AsyncHTTPClient()
    client.fetch(url, None, method='POST', headers=None, body=payload)

How to reload your dashboards

You can also perform dashboard-level actions by posting to /dashboards/dashboard_id. Currently there is a single event available, reload, that will refresh the target dashboard in the browser.

curl -i -d '{ "auth_token": "YOUR_AUTH_TOKEN", "event": "reload"}' http://localhost:8888/dashboards/sample

If you want an event to target every dashboard, you can use a wildcard (*).

curl -i -d '{ "auth_token": "YOUR_AUTH_TOKEN", "event": "reload"}' http://localhost:8888/dashboards/*

Installation

Installing drax is easy with pip:

    $ pip install drax

Developing

To run tests:

    $ python test.py

Tested with python 2.7