Getting Started

1. Install django-carson:

   $ mkvirtualenv --no-site-packages example_website
   $ pip install django-carson

2. Add carson to your INSTALLED_APPS

3. Create the database tables with syncdb (or migrate carson if you use South)

4. To access the Twitter Streaming API, you must first create the appropriate tokens.

   First, create a new application. Then, click “Create my access token.”

   Add the following values to settings.py

   • “Consumer key” → CONSUMER_KEY

   • “Consumer secret” → CONSUMER_SECRET

   • “Access token” → TOKEN_KEY

   • “Access token secret” → TOKEN_SECRET

5. Via the admin interface, add your accounts and hashtags/keywords.

   Note: You’re not required to add both accounts and hashtags/keywords. If you want, you could design a site that only stored hashtag mentions or only tweets sent by your given set of users.

6. If you added any accounts, you must either run ./manage.py lookup_twitter_ids or use the “Lookup Twitter IDs” admin action before the next step will work. If you only added hashtags or keywords, you don’t need to do this.

7. Finally, run:

   $ ./manage.py get_tweets

   This will open a connection to the Twitter Streaming API and immediately after one of your accounts posts a tweet or a tweet is created mentioning one of your tags, that tweet will be stored.

Usage

django-carson is only a bridge between Django and the Twitter Streaming API. It is the web developer’s job to wire up the views and templates needed to actually display the data.

The main entry point for any developer using carson is likely to be carson.models.Tweet. This model holds all tweets stored with get_tweets.

Each carson.models.Tweet object has four attributes:

account

A ForeignKey pointing to the carson.models.Account that created the tweet, if applicable. Will be None if the tweet didn’t come from an account listed in carson.models.Account.

timestamp

The UTC timestamp of the tweet.

status_id

The unique status id for each tweet. This is also in data[‘id’], but this allows an index to be created for it.

data

Stores the complete JSON object associated with the tweet. You can see what all is included in this attribute here.

Attached to carson.models.Tweet are three managers:

objects

Returns a QuerySet of all tweets

trusted

Returns only the tweets associated with a carson.models.Account. In other words, Tweet.account != None.

untrusted

Returns only the tweets not associated with a carson.models.Account. In other words, Tweet.account == None.

A simple index view exists in carson.views.index which grabs the 20 most recent trusted and untrusted tweets and renders carson/index.html (with the context variables trusted and untrusted). Might be useful if your website isn’t too complex.

If you seem to be having problems accessing the Twitter API, you can set HTTP_DEBUG to True in settings.py. By default it is False.

Changelog

0.2 (October 5th, 2011)

• Use SSL for Streaming and REST APIs

• Document and test some utility methods

• Use requests for REST API calls

0.1 (August 9th, 2011)

• Initial release