Dwitter

A Twitter clone written in django

Dwitter is made by different apps and modules available for django and python. It uses:

• Celery
• Redis
• Haystack
• Gunicorn
• Gevent
• Gevent-socketio
• django-registration
• django-taggit
• Twitter text lib
• Bootstrap

Purpose

Dwitter is an educational and experimental project. It shows how one can combine different parts available in order to make an application like twitter. It uses websockets, distributed task queue, ajax calls and redis in order to achieve this. It can be used as a starting or reference project for other more complex ones. Implements also various django workflows like registration, authentication and translation.

Features

• User registration
• User profile with name, description and image
• Multilingual, user selectable
• User search in name and description using haystack
• User notification when new dwit arrives
• Direct, Reply and Redwit capability
• Hash tags, tag search and tag watch

Installation

Requirements

Apart from the requirements listed in requirements.txt there are some extra:

• You must have redis installed on your system. Download the latest redis version from redis.io and install it.
• The gevent version listed in requirements.txt is a development one and requires the latest Cython to be installed. If you don't want that download gevent from google code and install it manually. You can find a wiki page for Cython installation.

VirtualEnv

It's advised, but not necessary, to install all packages under a virtual python environment. Virtualenvwrapper is a collection of useful tools for virtualenv.

pip install virtualenv virtualenvwrapper

Decide a name for the directory which will hold all your virtualenvs and add the next 3 lines at the end of your .bashrc. (Here ~/venvs is given as the virtual enviroments home directory)

export WORKON_HOME=~/venvs
source /usr/local/bin/virtualenvwrapper.sh
export PIP_VIRTUALENV_BASE=~/venvs

and do:

source ~/.bashrc

Now you are ready to create the environment you will work on.

mkvirtualenv dwitter

Once created you can activate it with:

workon dwitter

Download and install

Clone or fork this repository to a directory of your liking. Change directory to dwitter's root and do:

pip install -r requirements.txt

This will download and install all requirements.

Configuration

Dwitter needs 5 servers to run in order to function correctly. That's why there is a servers directory containing all start/stop scripts and configuration files needed for those servers to work. Those servers are configured to create all their needed files under the servers directory. So there is no need for you to sudo. Start up scripts and local database dump are provided for redis also. If this is not desired, in case you use redis already, you can delete this two lines from the startall and stopall scripts. There is also a wiki page for redis setup.

servers/servers.conf

This file holds all the necessary variables for celery. Change the following variables to suit your case:

# Full path to root location of project
CEL_CHDIR=path/to/dwitter's root

# Full path of virtualenv or system interpreter
ENVPYTHON=path/to/python

# Username ang group the celery servers will run under.
SERVERS_USER = youruser
SERVERS_GROUP = yourgroup

servers/redis.conf

Configuration file for local redis server. Change the following variables to suit your case:

# Full path of redis pid file. Recomendation: (dwitter_root)/servers/run/redis.pid
pidfile path/to/redis.pid

# Full path of redis log file. Recomendation: (dwitter_root)/servers/log/redis.log
logfile path/to/redis.log

# Full path of redis db directory. Recomendation: (dwitter_root)/servers/redis
dir path/to/redis_db_dir

dwitter/settings.py

The only necessary variable here is:

PYTHON_ENV = 'path/to/python'

where you must provide the same path as in servers/servers.conf ENVPYTHON variable. For a complete list of variables available in settings.py please visit the wiki page.

First time setup

Once everything is installed you must setup the database, change directory to dwitter's root (if not already there) and type the following commands:

./manage.py syncdb                           # To create your db
./manage.py createsuperuser                  # To create your first user
./manage.py migrate djcelery                 # This is needed for celery
./manage.py convert_to_south dwitter.main    # This is optional but useful
./manage.py collectstatic                    # To have your static files ready for serving
./manage.py rebuild_index                    # To create your haystack db

You are now ready to start the servers.

servers/startall

This will start redis, celery, celeryev, celerybeat. To stop the servers do:

servers/stopall

Every server can be started/stopped by its own with its starting script:

servers/celeryd start
etc.

In order to start gunicorn you have two options.

Either run it with:

./manage.py run_gunicorn -c gunicorn

this will start the gunicorn server with the ./gunicorn file as config displaying the log at your console.

Or:

servers/gu start

this will start the gunicorn server with the servers/gunicorn file as config which daemonizes it and saves logs at servers/log/gunicorn.log. In order to stop the server you need to do:

servers/gu stop

First time run

Once everything is running point your browser to: 127.0.0.1:8000/ and test it.

Dwitter's members have an one to one relation with django users. So, in order to create a dwitter member you must first create a django user, then create a new member and relate it with the user just created.

So go to admin page login with the superuser data you provided earlier and create a member for the superuser.

Now you can login and use dwitter.

Enjoy!!

Links

This project was inspired by Flavio Curella's blog post and a series of blog posts by Rick Copeland and especially this.