• Stars
    star
    165
  • Rank 228,906 (Top 5 %)
  • Language
    Python
  • Created over 14 years ago
  • Updated about 14 years ago

Reviews

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

Repository Details

A command-line twitter client with smart filtering and statistical classification
tc - Twitter Commander
(a tribute to http://en.wikipedia.org/wiki/Midnight_Commander)
======================

=====
About
=====

tc is a command-line twitter client with smart filtering and statistical classification.

Capabilities:
 * whitelists for users whose content you always want to see and prioritize
 * blacklist for users you follow whose content you never want to see
 * blacklist for regular expressions to filter out of your stream entirely
 * topic classification (simple right now, I'm working on this)
 * search
 * the command-line UI rules the skool

Dependencies:

Database:

mongodb - should be running on localhost (it's easy) - http://www.mongodb.org/downloads

Python packages:

pymongo - the python interface to mongo
tweepy - twitter API client
nltk - for simple topic classification and filtering

=======
Install
=======

Install dependencies, and configure your username in settings.py. Then, run load_tweets.py. If it's your first time, you'll
have to go through the oath authorization process.

=====
Usage
=====


load_tweets.py - loads tweets from twitter and insert into the db. run this first.

t.py - twitter client
  options:
  -n # -- number of tweets to display (default = 10)
  -s time|antitime|rel|inf -- sort unread tweets by time (newest first), antitime (oldest first), rel (arbitrary relevance), inf ('influence' - using klout score) (default = time)
  -m -- don't mark displayed tweets as read (for debugging)
  -d -- print debug output to STDOUT (for debugging)


s.py - search tweets
  options:
  args -- search terms
  -n # - number of results to display (default = 10)
  -u val -- search username and user's name fields for val, display username and bio