get_tweets

#!/usr/bin/env ruby
 
require 'rubygems'
require 'json'
require 'net/http'
require 'uri'
require 'time'

module TwitterAPI
  
  def api_call(method,page,username,password)
    url = URI::parse('http://twitter.com')
    req = Net::HTTP::Get.new("#{method}?screen_name=#{username}&page=#{page}&count=200")
    req.basic_auth(username, password) if password
    res = Net::HTTP.start(url.host, url.port) {|http| http.request(req) }
    JSON.parse(res.body)
  end
  
  def all_tweets(method,username,password=nil)
    page = 1
    loop do
      begin
        tweets = api_call(method,page,username,password)
        break if tweets.length == 0
        tweets.each { |tweet| yield tweet }
        page += 1 
        sleep 5
      rescue
        $stderr.puts "Error getting page #{page}: will retry in 30secs"
        sleep 30
      end
    end
  end
end

class TwitterBackup
 
  include TwitterAPI
 
  def tweets(username,password=nil)
    all_tweets('/statuses/user_timeline.json',username,password) do |tweet|
      print(tweet)
    end
  end
  
  def retweets(username,password=nil)
    all_tweets('/statuses/retweeted_by_me.json',username,password) do |tweet|
      print_rt(tweet)
    end
  end
  
  protected
    def clean_up(tweet)
      tweet.gsub("\n"," ").gsub("\r"," ").gsub("\t"," ")
    end
  
    def print_rt(tweet)
      text = "RT @"+tweet['retweeted_status']['user']['screen_name']+": "+clean_up(tweet['retweeted_status']['text'])
      put_tweet(tweet,text)
    end
  
    def print(tweet)
      put_tweet(tweet,clean_up(tweet['text']))
    end
    
    def put_tweet(tweet,text) 
      id, date = tweet['id'], Time.parse(tweet['created_at']).strftime("%Y-%m-%dT%H:%M:%S%z")
      puts [id,date,text].join("\t")
    end
end

username,password = ARGV[0],ARGV[1]
TwitterBackup.new.tweets(username,password)
TwitterBackup.new.retweets(username,password)