fetch_user.rb

#!/usr/bin/ruby
require "twitter"

user = ARGV[0]

File.open("data/users/#{user}", "w") do |o|
  Twitter.all_tweets(user) do |tweet|
    o.puts tweet["text"]
  end
end

twitter.rb

require "rubygems"
require "httparty"

class TwitterSearch
  include HTTParty
  base_uri "http://search.twitter.com"
  format :json


  def self.search(string, params={})
    params[:q] = string
    get "/search.json", :query => params
  end

  # Fetch everything in the search history for this query.
  def self.full_search(string, modifiers = {}) 
    modifiers[:rpp] = 100

    results = search(string, modifiers)
    
    all_results = results["results"]
    
    while (next_page = results["next_page"])
      results = get("/search.json" + next_page)
      all_results += results["results"]
    end

    all_results
  end
end

class TwitterException < Exception  
end

class Twitter
  include HTTParty
  base_uri "http://twitter.com"
  format :json
   
  def self.tweets(screen_name)
    # this is the lamest caching in the entire world
    # TODO: Pls fix me
    @tweets_cache ||= {}

    results, time = @tweets_cache[screen_name]
    if !time || (Time.now - time > 300) 
      results = get("/statuses/user_timeline.json", :query => {"id" => screen_name, :count => 50})
      @tweets_cache[screen_name] = [results, Time.now]
    else 
      STDOUT.puts "hit the cache. No need to bother twitter"
    end
    
    raise TwitterException.new(results["error"]) if results.empty? && results["error"]
    results
  end
  
  def self.friends(screen_name)
    @friends_cache ||= {}

    results, time = @friends_cache[screen_name]
    if !time || (Time.now - time > 300) 
      results = get("/statuses/friends.json", :query => {"id" => screen_name})
      @friends_cache[screen_name] = [results, Time.now]
    else 
      STDOUT.puts "hit the cache. No need to bother twitter"
    end
    
    raise TwitterException.new(results["error"]) if results.empty? && results["error"]
    results.map{|f| f["screen_name"]}
  end
  
  def self.user(screen_name)
    @users_cache ||= {}
    
    results, time = @users_cache[screen_name]
    if !time || (Time.now - time > 300) 
      results = get("/users/show.json", :query => {"id" => screen_name})
      @users_cache[screen_name] = [results, Time.now]
    else 
      STDOUT.puts "hit the cache. No need to bother twitter"
    end
    
    raise TwitterException.new(results["error"]) if results.empty? && results["error"]
    results
  end
  
  def self.all_tweets(user, params={})
    if !block_given? 
      results = []
      all_tweets(user, params) do |tweet| 
        results << tweet
      end
      return results
    end
  
    new_tweets = nil

    page = 0 
  
    while !(new_tweets = get("/statuses/user_timeline.json", :query => {"id" => user, :page => page, :count => 200})).empty?
      new_tweets.each{|tweet| yield(tweet) } 
      page += 1
    end 
    nil;
  end

end