gistfile1.py

def collectURLs(tweets):
    """
    Extract all mentioned URLs from a set of tweets.

    @param tweets: A C{dict} of tweets, as returned by L{getHistoricalTweets}.
    @return: A C{dict} with mentioned URL keys and values a time-sorted list
        of tweet URLs.
    """
    URLs = defaultdict(list)
    for id, tweet in tweets.iteritems():
        for mentionedURL in tweet.entities['urls']:
            expandedURL = mentionedURL.get('expanded_url')
            if expandedURL:
                URLs[expandedURL].append(
                    (id, 'http://twitter.com/#!/%s/status/%d' % (
                        tweet.user.screen_name, id)))

    # Sort each list of tweets of each URL by tweet id (i.e., in order of
    # tweeting).
    sortedURLs = {}
    idGetter = itemgetter(0)
    URLGetter = itemgetter(1)
    for URL, mentions in URLs.iteritems():
        sortedURLs[URL] = map(URLGetter, sorted(mentions, key=idGetter, reverse=True))

    return sortedURLs