cambiatus/backend

Cambiatus GraphQL API

Backend 👋

New Organizations for the planet regeneration. Clear Objectives. Impactful Collaboration.

For more information on us:

• Wiki
• Blog
• Communities
• FAQs
• Architecture & Data Flow

Table of Contents

• General Information
• Technologies
• Development Environment Setup
• Contributing
• Additional Resources
• License

General Information

Welcome to the Application that serves as one of the backends for the Cambiatus Ecosystem. One of the backend since this is the one that contains the data in a manner than is easy to index and search.

In the context of the diagram below which is a high level view of how the data flows in our application this application serves as the datastore using a Postgres db and as the API using Phoenix running a Graphql Server

Dataflow

At a high level this is a database that is synced to events on an blockchain which then presents a Graphql API that makes it simpler to consume the information from the blockchain. At the moment this is a normal database with the the usual CRUD actions however creation and updating happens as a result of events that trigger writes and updates to our database.

The intention down the line is to make this database a write only database in an Event Sourced structure which will enable us to replay events and give us much more observability.

Technologies

Here we have information on the type of technologies we use on our project, enjoy!

Language & Framework

• Elixir language main documentation page
• Phoenix framework main documentation page

Query language

• Intro to GraphQL

Here are some notable packages related to GraphQL:

• Ecto is Elixir's database wrapper that works around GraphQL

• Absinthe package GraphQL toolkit for Elixir

Here is our GraphQL wiki page

Databases

• Postgres main documentation page

• EOS Blockchain main documentation page

  • Here is our documentation on how we use EOS blockchain

HTTP server

• NGINX main documentation page

Development Environment Setup

To build and run this application locally follow the following steps!

Step 1

Clone this repository by running

git clone [email protected]:cambiatus/backend.git

Step 2

Change into the new repository directory by running

cd backend

Step 3

The system uses two external dependencies that need to be manually installed. Both of them are used for dealing with images

• ImageMagick is used to ensure that the process of image uploads runs properly and Mogrify (dependency package) is dependent on it. Here are instructions for installing it.
• Exiftool is used to remove unnecessary metadata from images, mainly to help protect user privacy. Here are instructions for installing it.

Install dependencies by running

mix deps.get

Step 4

Create a database by running

mix ecto.create

Note: you may need to change the database user and password variables values in the config/test.exs and config/dev.exs to ensure proper connection with Postgres.

Then, run the current database migrations using

mix ecto.migrate

Step 5

Once the ecto migration is done successfully, run tests via the test command below

mix test

Note: Ideally the test results should exit with a status of 0 failed tests

Step 6

Lastly, run the server using the following command

mix phx.server

Now you can visit localhost:4000 from your browser.

💥 Boom! Now you can hack away!

NGINX Setup

This project was written with NGINX in mind and so we need to configure it properly. You can check out how to install and run it on their official site. When running your own servers or setting up a new env for Cambiatus, without those configurations we will not be able to fully operate. We basically need to setup redirects and crawlers

NGINX Setup: API and Mailer redirects

We use pretty straight forward configurations. Make sure we reply to our backend from both /api and /mailer:

        location /api/ {
                proxy_set_header Host $host;
                proxy_set_header X-Real-IP $remote_addr;
                proxy_set_header Upgrade $http_upgrade;
                proxy_set_header X-Forwarded-For $remote_addr;
                proxy_redirect off;
                proxy_http_version 1.1;
                proxy_set_header Connection "upgrade";
                proxy_set_header Origin *;
                proxy_pass http://ADDRESS:PORT;
                client_max_body_size CLIENT_SIZE;
        }
	
	location /mailer/ {
                proxy_set_header Host $host;
                proxy_set_header X-Real-IP $remote_addr;
                proxy_set_header Upgrade $http_upgrade;
                proxy_set_header X-Forwarded-For $remote_addr;
                proxy_redirect off;
                proxy_http_version 1.1;
                proxy_set_header Connection "upgrade";
                proxy_set_header Origin *;
                proxy_pass http://ADDRESS:PORT;
                client_max_body_size CLIENT_SIZE;
        }

NGINX Setup: Crawlers

Cambiatus has builtin support for social media crawlers, in order to dynamically serve rich links (or Open Graphs) in accordance to the Open Graph Protocol. To setup, follow these steps:

Step 0

Locate your NGINX configuration file (nginx.conf) and get ready to edit it

Step 1

Add the following lines right before any of the server blocks are declared:

map $http_user_agent $rich_link_prefix {
	default 0;
	~*(facebookexternalhit|twitterbot|telegrambot|linkedinbot|slackbot|whatsapp)  /api/rich_link;
}

This code is responsible for detecting a crawler from different social media websites based on the user agent issued on the http request.

If a crawler is detected it saves the user agent to the $rich_link_prefix variable.

Step 2

Inside the server block referring to server_name block staging.cambiatus.io *.staging.cambiatus.io; under location / insert:

proxy_set_header Host $http_host;
proxy_set_header X-Real-IP $remote_addr;
if ($rich_link_prefix != 0) {
	proxy_pass http://127.0.0.1:#{backend_listening_port}$rich_link_prefix$request_uri;
}

Note: Replace #{backend_listening_port} with the port to which the backend server is setup to listen.

This code redirects the identified crawler to the correct rich link URL. Nut if a crawler was not detected the server runs normally.

Contributing

When you are ready to make your first contribution please check out our contribution guide, this will get your up to speed on where and how to start.

Once done with the contributing guide, here are some developing tips to help you:

Code Formatting

• To ensure code consistency we use linter testing, static code analysis for the approval of our Pull Requests. Always run mix credo before pushing your commits.

• Another critical formatting command is mix format, which formats a specific file according to the Elixir language formatting rules command. There are IDE specific extensions and settings that you could use to have automated formatting. Here is one Elixir vscode example for this.

Files Not To Commit

Changes related to local Postgres database credentials must not be committed to the repo

Here one way of how to not commit changes related to your local database connection credentials (user & password) to dev.exs and text.exs files:

• After you finished working on a development and before you commit your changes, select dev.exs and text.exs files on your IDE or Git Desktop.

  • IF the only changes to these files are related to the local database connection credentials.

    • Then, revert all the changes on these two files (dev.exs and text.exs).

  • ELSE

    • ONLY revert the changes related to the local database connection credentials on these two files (dev.exs and text.exs).

  • HERE is another alternative using the exclude command.

Additional Resources

• Here is our Frontend (Elm) repo. We use Elm which is an awesome functional language to play with!

• Here is our Smart Contract (EOS) repo. We use EOS which is an awesome blockchain and uses C++ as language to play with!

• Our wiki page has several development resources to help you during your collaboration.