gistfile1.rb

# -*- coding: utf-8 -*-

# This script will download the meaning of all names found on http://www.misabueso.com/nombres/
# Date: July 20, 2009, at 2:14AM
# Author: Andrés Mejía <andmej@gmail.com>

# The format of the output file is as follows.
# For each name fetched, there are two lines:
# Line 1: The name itself
# Line 2: A valid HTML chunk containing the meaning of the name and some other stuff

require 'rubygems'
require 'hpricot'
require 'open-uri'
require 'iconv'

def element_from_url(url)
  repeat = true
  while repeat
    repeat = false
    begin
      f = open(url)
      doc = Hpricot(Iconv.conv('UTF-8', f.charset, f.read))
      f.close
    rescue Exception => e
      print "Exception caught: "
      puts e.inspect
      puts "Retrying in 1800 seconds..."
      sleep(1800)
      repeat = true
    end
  end
  doc
end

def do_name(name, url, output_file)
  puts "  About to fetch '#{url}'..."
  doc = element_from_url(url)
  puts "   received."

  header = doc.search("html > body > div#contenido > div.incont > ul.li1").first
  header = header.inner_html.gsub(/[\n\t\r]+/, ' ').gsub(/<li>/, '').gsub(/<\/li>/, '')

  content = doc.search("html > body > div#contenido > div.incont > div > div.r3 > div.inr3").first
  content = content.inner_html.gsub(/[\n\t\r]+/, ' ').gsub(/h6/, "h3")

  if content && header
    output_file.puts name
    output_file.puts "<div class=\"name_meaning\">#{header}</div> <div class=\"name_analysis\">#{content}</div>"
  else
    puts "[ERROR] Problems fetching. Name = #{name}, content = #{content}, header = #{header}"
  end
  sleep(20)
end

def do_names_that_start_with(letter)
  url = "http://www.misabueso.com/nombres/nombre_#{letter.upcase}.html"
  puts "About to fetch '#{url}'..."
  doc = element_from_url(url)
  puts " received."

  output_file = File.new(letter + ".txt", "w")
  links = doc.search("a") # finds all links
  links.each do |link|
    uri = link["href"]
    name = link.inner_html
    if uri and Regexp.new("nombre_.{2,}\.html").match(uri)
      do_name(name, "http://www.misabueso.com/nombres/" + uri, output_file)
    end
  end
  output_file.close
end


def do_everything
  "A".upto "Z" do |letter|
    do_names_that_start_with(letter)
  end
end

do_everything

Iconv.conv('UTF-8', f.charset, f.read)

Iconv es la peor librería que podrías estar utilizando para manejo de UTF-8 =S Esa mierda no sirve para nada! Yo tuve muchos problemas con eso y el plugin para los permalinks automáticos. Utilizá esto: https://github.com/svenfuchs/stringex

¡Está una elegancia esa librería! ¡Gracias!