Skip to content

Latest commit

 

History

History
324 lines (237 loc) · 13.9 KB

README.md

File metadata and controls

324 lines (237 loc) · 13.9 KB

Fastrack analytics for your business. Use SQueaLy

Deploy
Build Status license

SQueaLy is an open-source, self-deployable application for developers. It is a micro service for business intelligence and analytics which uses SQL queries to generate reporting APIs with fine-grained security. It gives you the power to analyze your organizational data in an environment that is completely owned by you. Hence, making it the most suitable solution for generating charts and reports out of sensitive data.

When to use SQueaLy

When you want charts and reports but don't want to build the APIs

Using SQueaLy, creating APIs for your reports/charts and visualizing the output is super quick. Just write the query to fetch the data from your database and use the SQueaLy's authoring interface to visualize the output. Once satisfied, give your users the permission to view/edit the report and share the URL of your report. That's it!

When you want fine grained security in your reporting APIs

You can provide not only chart level view permissions but also user/parameter level permissions. To make that one liner straighforward, consider the following use case.

You have a chain of hotels based on multiple locations. You create 5 reports which are to be viewed by  
the managers of each hotel but you don't want the manager of one location to view the report of the  
hotels not managed by him. Now if you write this logic in your own code, it is going to take a lot of  
time but using SQueLy, this can be done in seconds.

When you want scheduled emails embedded with reports

SQueaLy also provides you the interface to schedule emails with reports embedded in it. All you have to do is

  • Group the reports you want to send in the email
  • Write a cron expression to schedule the email

Not just that. For white labling, you can even customize the look and feel of the email message by providing the HTML for it

Key Features

One-Click Deploy to Heroku

Just click on the 'deploy-to-Heroku' button, login to your Heroku account, enter environment variables and the application will be deployed under your account and domain.

JSFiddle like interface to test and visualize the output
SQueaLy provides you a user-friendly editor like interface to test and debug your queries/APIs.

Your data is always secure

All the database credentials and secret keys are stored as environment variables in your own server architecture (Heroku).

More than just SQL

SQueaLy uses JinjaSQL behind the scenes which gives you power to use JINJA inside the SQL query. For more details, check out JinjaSQL.

Permission based model

SQueaLy provides you the ability to assign view/edit permission to any user for each chart/report.

Multiple Databases

You can set up multiple database connections and specify the database to use while generating the chart/report.

Scheduled emails embedded with reports

SQueaLy comes with email solutions as well. You can schedule emails embedded with the charts/reports you had created by just writing a cron expression

Requirements

  1. An account on Heroku
  2. A database accessible from the internet

Database Support

Athena, Redshift, Mysql, Postgres, Sqlite,

Getting started

  • Click on the 'Deploy-to-Heroku' button on the top of this documentation.
  • Login/signup to your Heroku account.
  • Enter the environment variables and click deploy (Make sure to note the admin credentials entered at this step).

The app is deployed and the first chart has already been created for you. Hit the Run button and you can see the data/visualization in the results section.

Usage

Managing Permissions (Chart-level Authorization)

SQueaLy uses Django's default permission model for managing chart level authorization. Everytime a new chart is created, 2 permissions are added in the backend automatically corresponding to view and edit mode access.

For example, if the chart is named as 'foo' then, permissions 'can-view-foo' and 'can-edit-foo' are added automatically.

The administrator just needs to add the required permissions to users or groups from the Django admin panel. ( located at https://<your_domain>/admin )

Also, there are two other high-level permissions - can-add-chart and can-delete-chart that are required for creation and deletion of charts, respectively.

For non-Django folks: By default, the admin user has all permissions.

Using the APIs

SQueaLy generates APIs in real-time corresponding to each chart. Just hit the API corresponding to your chart_url, NOTE: The chart_url is the chart name in lower case, replacing spaced by hyphen ('-')

<your_domain>/squealy/<chart_url>?<param_name>=<value>&<param_name>=<value>

This API will return the data in the format that is compatible with GoogleCharts.

Query Parameterization

To add a parameter, you need to use the keyword object "params" inside your jinja template. For example, to include a parameter named foo,

SELECT * from some_table
WHERE some_value = {{params.foo}};

In the view mode, these parameter values would be extracted from the filters, but if you are using the API, then you need to pass these parameters in the url like this:

<your_domain>/squealy/<chart_url>?<param_name>=<value>&<param_name>=<value>

User-level Authorization

SQueaLy allows you to use user parameters inside the query template, the values of which would be extracted from the logged in user while running the query.

Use the keyword object "user" in the jinja template to access the request.user object provided by Django. For further details, checkout Django User Objects.

For example,

--- To see the bookings of the current user only.
SELECT bookings from bookings_table
WHERE some_value = {{params.foo}}
AND name = {{user.username}}

Swagger Documentation

Swagger is an integrated, cloud based API development platform that combines the core capabilities of Swagger with advanced features to build, document, and deploy APIs.

--- Query for testchart1.
SELECT * from bookings_table
WHERE booking_id = {{ params.booking_id }};

--- Query for testchart2
SELECT * from bookings_table
WHERE booking_date
BETWEEN {{ params.start_date }}
AND {{ params.end_date }};

Once you are done creating your charts, you can get a documentation of all the REST APIs for these charts. To view this documentation, got to

<your-domain>/swagger

Dropdown Filter APIs

SQueaLy provides Filter APIs to add Dropdown options. You just need to write the query to get the data from the database. SQueaLy provides can_edit permission for filters. If user does not have can_edit permission, API will be available as a dropdown option only while adding parameters in the chart.

SQueaLy supports Parameterized filters also. For Example: you have two filters Country and City and you want to show all cities for selected Country.

Query for Country

SELECT Country FROM some_table;

Query for City

SELECT City FROM some_table WHERE Country={{params.country}};

Note: Order plays an important role for Parameterized filter. Parameterized filter should render after it's parameters. For above example, City filter should render after Country filter.

Validations

With every chart, you can attach another SQL query that would validate the API. The API would return a 403 Forbidden response, if the validation query returns no rows.

JWT Authentication

SQueaLy provides a mechanism to log in a user via an access token that would help in sharing some other application's user-base with squealy without the need to import the users in SQueaLy database.

  • JWT_KEY - This is setup in the environment while deploying SQueaLy app. This is the private key that is shared between squealy and the other application which is directing users to SQueaLy.
  • Token encoding: Use the above key to create a jwt token in your application with the following payload:
    {
    "username": "foo", //Squealy assumes the authenticity of this user is handled by the directing application.
    
    "groups": ["g1", "g2", "g3"] //List of Django permission groups that this user belongs to.
    }
  • Sending the token: You can send this token in the request header. SQueaLy would ensure to login the user before handling the request.

Here is an example of hoe the request header should look like:

  Authorization: Bearer <token>

Scheduling emails

The scheduled emails can contain as many reports(charts) as you want. You just have to associate the reports(charts) you want in the email with the scheduled email.

To schedule emails embedded with reports, initialize the following variables in the environment:

  1. EMAIL_HOST
  2. EMAIL_HOST_USER
  3. EMAIL_HOST_PASSWORD
  4. EMAIL_PORT

Then, go to the Django admin and select the Scheduled reports model. On doing so you will see a form like this

Subject

This will be the subject of the email. This subject does not have to be hard coded. You can use the parameters in the query or any macro provided by squealy here. Just wrap the placeholders in {{}} to generate the subject dyanamically

Template

You can customize the look and feel of the email body by writting HTML here. Do not forget to put {% include 'report.html' %} somewhere in the template otherwise there won't be any reports in the email.

Cron expression

Enter a valid cron expression to tell the schedule of the email. For examle, to schedule an email for 11AM everyday, the cron expression should be 0 11 * * *

Report Parameters

This would contain all the values of the parameters used in the database queries. For example

If you have two reports(charts), one of them uses the parameter - facility-id and the other one uses  
two parameters - start_date and end_date. Then you need to provide the names of these parameters with  
their corresponding values to be used while generating the reports for an email

Report Recipients

This would contain the email address of all the recipients of a particular scheduled email.

Associating charts with Scheduled reports.

To associate charts with scheduled reports i.e. to determine which report has to be sent inwhich email, go tho Django admin and select the Scheduled Report Charts model. On doing so you will see a form like this

Here, just select a chart and the scheduled report and hit save. Doing so will associate one chart with the selected scheduled report. Repeat this for as many charts you want to associate to a scheduled report

Development

Backend setup

Activate a virtual environment

virtualenv venv
source venv/bin/activate

Installing dependencies

pip install -r requirements.txt

Database Setup SQuealy needs a postgres database to run. Create a database and a user with Create DB role. (To run test cases). Use the following command on psql to grant this access to a user:

ALTER USER username CREATEDB;

Note the URL of this database for use in the next step. The URL will be of the following format:

<mysql/postgres>://<username>:<password>@<host>:<port>/<database>

For further help regarding postgres setup, click on one of the following links:

Setting up environment variables Before running the server, make sure to set the following environment variables using this command:

export KEY=VALUE
  • DATABASE_URL - This is the database that will be used by squealy internally for user management.
  • QUERY_DB - This is the DB used to run the queries on. Enter comma separated URLs for multiple DBs.
  • ADMIN_USERNAME - This is the default admin's username that would be created with the first migration command.
  • ADMIN_PASS - This is the password for the default admin user.
  • JWT_KEY - This is the shared private key that will be used to decode the JWT token. This is required for optionally enabling JWT Authentication mechanism that would enable the user to login via an access token parameter in the url or request body. For more information refer above.

Running the migrations

python manage.py migrate

Running the server

python manage.py runserver

Running test cases

python manage.py test

Frontend setup

Installing dependencies

cd squealy/squealy-web
npm install

Running the development server

npm start

Running test suites

npm test

Getting production build

npm run build

Copyright

(c) 2017 HashedIn Technologies Pvt. Ltd.