Habra comments parser

parser.rb

#!/usr/bin/env ruby
# encoding: utf-8
require 'nokogiri'
require 'open-uri'
require 'yaml'

###################################################
# ♥ http://habrahabr.ru/users/bupycnet/comments/

NICK =  unless ARGV.empty? then "#{ARGV.pop}"  else
  puts "Использование: ./parser.rb nick"
  exit
end

CACHE = "cache-#{NICK}.yaml"
###################################################

class Parser
  attr_reader :nick

  def initialize nick
    @nick = nick
  end

  def get_document relative
    Nokogiri::HTML(open("http://habrahabr.ru/#{relative}",
                        "User-Agent" => "Mozilla/5.0"))
  end

  def get_pages
    get_document("users/#{@nick}/comments/page999/").css('#nav-pages li').max.content
  end

  def get_comments page
    get_document("users/#{@nick}/comments/page#{page}/").css('.comment_item .message').map{|html| html.content.strip}
  end

  def get_all_comments
    comments = []
    1.upto(get_pages.to_i).each { |page|
      comments << get_comments(page)
    }
    comments.flatten
  end
end

def run
  comments = []

  # check for cache
  unless File.exist? CACHE
    File.open(CACHE, "w") do |file|
      # do cache
      comments = Parser.new(NICK).get_all_comments
      file.puts YAML::dump(comments)
    end
  else
    comments = YAML::load(open(CACHE, "r"))
  end

  puts "Пользователь: #{NICK}"
  puts "Всего комментариев: #{comments.size}"

  overall = comments.inject{ |sum, e|  sum = e.length + sum.to_i}

  puts "Общая длина: #{overall}"
  puts "Средняя длина комментария: #{overall / comments.size}"
  puts "Длины первых 10 самых длинных:"
  p comments.map{|comment| comment.length }.sort.reverse[0..10]
end

run