Django Chat!

Django Chat is a project built with Daphne and Channels to demonstrate Django capabilities beyond the HTTP protocol! It contains a real-time chatting application via sockets protocol. Moreover, it also brings a CHAT BOT logic so that user can use some commands!

This app contains:

• Real-time chatting application with rooms and users powered with Django and Channels;
• Persistence of chat messages, users and rooms;
• CHAT BOT logic that allows users to use commands such as /now;
• Task queue example with the /sms command;
• Async web socket consumer;
• Async HTTP consumer.

By using async consumers, it's very easy to use Python's coroutines with asyncio or trio!

Application Demonstration

Technologies used

• Python 3.8;
• Django;
• Django Channels;
• Daphne;
• Pytest;

Chat Rooms and Users

Since this is a demo application, to create and join a chat room, just follow the appropriate endpoints:

/chat/<chat_room>/<chat_user_name>

Hence, to join a room named myroom with your username being conker, just go to the url:

/chat/myroom/conker

And all the messages will be sent with the username conker! Hyphens are NOT supported yet in the chat_room and chat_user_name.

PS: This a demo application just to show Django Channels and chatting logic, so no real user authentication, login and other security enforcements were developed (not yet... at least!).

Chat BOT commands syntax

To use a command, start the sentence with a slash /! Hence, the syntax to use a command is as follows:

/command arg1 arg2

where command is the desired command, and arg1, arg2, etc. are the arguments for such command.

Chat BOT available commands

Currently two commands available:

• /now: prints the current server datetime;
• /sms <user_name>: fakely sends an SMS to the user (demonstrates channels simple task queue)

Task Queue

The command /sms <user_name> sends a message (task) to a fixed channel which is listed by a worker that can process such messages! This demonstrates simple task queue that channels allows! (Simple because there's no task retrying, etc. -- not Celery!)

The worker starts listening on the channel named sms_channel by the following command:

python manage.py runworker sms_channel

Running locally

To run the app, just start it with docker-compose:

docker-compose up

This will start Django's development server on localhost listening on port 8080.

Running tests

Soon to come!