Solution for Scraping HN 1: Building Objects

Scraping_HN_1:_Building_Objects_solution.rb

# Solution for Challenge: Scraping HN 1: Building Objects. Started 2013-08-14T20:47:25+00:00

require 'nokogiri'
require 'rubygems'
require 'open-uri'

class Post
  attr_reader :comments, :comment_objects

  def initialize(url)
    @page = Nokogiri::HTML(open(url))
    get_comments
    get_comment_id
    get_users
    create_comment_objects
  end

  def title
    @page.search('.title > a:first-child').map { |link| link.inner_text }
  end

  def url
    @page.search('.title > a:first-child').map { |link| link['href'] }
  end

  def points
    @page.search('.subtext > span:first-child').map { |span| span.inner_text }
  end

  def item_id
    @page.search('.subtext > a:nth-child(3)').map { |link| link['href'] }
  end

  def get_comments
    unless @comments
      @comments = @page.search('.comment > font:first-child').map { |font| font.inner_text }
    end
  # BUG::NEEDS TO BE KEEP 'P' together. Currently breaks it up based on <p>
      #     @page.search('td.default').each do |td_tag|
      #   td_tag.search('.comment')
      #   end
      # p @comments
      # comments.map { |comment| comment.inner_text }
  end

  def get_users
    @commented_users = @page.search('.comhead > a:first-child').map { |user| user.inner_text }
  end

  def get_comment_id
    @all_comment_id = @page.search('.comhead > a:nth-child(2)').map { |link| link['href'] }
  end

  def create_comment_objects
    @comment_objects = []

    @comments.length.times do |i|
      @comment_objects << Comment.new(@comments[i], @commented_users[i], @all_comment_id[i])
    end

  end

  def add_comment(comment, user, id)
    @comment_objects << Comment.new(comment, user, id)
  end
end

class Comment

  attr_reader :comment, :user, :id

  def initialize(comment, user, id )
    @comment = comment
    @user = user
    @id = id
  end
end

url = ARGV
post = Post.new(url.pop)
p post.title
p post.url

# p post.add_comment("Hello", "Elaine", "123")
# p post.points
# p post.item_id
# p post.comments
# p post.add_comment("I like cats.")
# p post.comments
 p post.comment_objects[0].user
# p post.comment_objects.last.inspect

# page = Nokogiri::HTML(open('post.html'))
# puts page.class

# def extract_usernames(doc)
#   doc.search('.comhead > a:first-child').map do |element|
#     element.inner_text
#   end
# end

# def extract_post_points(doc)
#   doc.search('.subtext > span:first-child').map { |span| span.inner_text }
# end

# def extract_post_id(doc)
#   doc.search('.subtext > a:nth-child(3)').map { |link| link['href'] }
# end

# def extract_title(doc)
#   doc.search('.title > a:first-child').map { |link| link.inner_text }
# end

# def extract_title_link(doc)
#   doc.search('.title > a:first-child').map { |link| link['href'] }
# end

# def extract_comments(doc)
#   doc.search('.comment > font:first-child').map { |font| font.inner_text }
# end

# def extract_comment_direct_link(doc)
#   doc.search('.comhead > a:nth-child(2)').map { |link| link['href'] }
# end

# p extract_usernames(page)
# p extract_post_points(page)
# p extract_post_id(page)
# p extract_title(page)
# p extract_article_link(page)
# p extract_title_comment_font(page)
# p extract_comment_direct_link(page)


# Data structure is an array.
# Yes, you can call ruby methods on the returned data structure.

# links = page.css("table").css("tr").css("td").css("img")