Guest User

Untitled

a guest
Feb 12th, 2014
120
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
Ruby 1.54 KB | None | 0 0
  1. require 'nokogiri'
  2. require 'open-uri'
  3. require 'sequel'
  4. require 'sqlite3'
  5. require 'pry'
  6. require 'uri'
  7. require 'cgi'
  8.  
  9.  
  10.  
  11. clubland = "http://www.clublandlv.com/forum.php"
  12.  
  13. doc = Nokogiri::HTML(open(clubland))
  14.  
  15. puts doc.css('title').text
  16. puts
  17.  
  18.  
  19. doc.css("#cat10").each do |grab|
  20.   genres = grab.css(".forumtitle").map(&:text)
  21.   genreslink = grab.css(".forumtitle"){[:href]}
  22.   subgenre = grab.css(".subforum a").map(&:text)
  23.   subgenrelink = grab.css(".subforum a"){[:href]}
  24.    
  25.  
  26.   puts "This is an array of the genres:"
  27.   puts genres
  28.   puts
  29.   puts "These are the genre links:"
  30.   puts genreslink
  31.   puts
  32. #  puts "This is an array of the subgenres:"
  33. #  puts subgenre
  34. #  puts
  35. #  puts "These are the subgenre links:"
  36. #  puts subgenrelink
  37.  
  38.  
  39.    
  40. # Table has 3 columns. ID(auto), genre, forum_id
  41. #Connects to the local DB
  42.         DB = Sequel.connect('sqlite:///D:/Users/RyanOConnor/workspace/testing/clublandlv.sqlite')
  43. # Create a dataset
  44.         main_genre = DB[:genres]
  45. # Insert the results
  46.  
  47.  
  48.    
  49. #  genres.each do |genre|
  50. #   main_genre.insert(:genre => genre)
  51. #    unless db_has_genre?
  52. #    genre;
  53.  
  54.  
  55.         genreslink.each do |g_link|
  56.         link_id = g_link.children.first["href"]
  57.         link_cut = CGI.parse(URI.parse(link_id).query)['f']
  58.         #   main_genre.insert(:forum_id => link_cut)
  59.         g_link;
  60.     binding.pry      
  61.  
  62.  
  63.  
  64.        
  65.              
  66.      #   main_genre.insert(:forum_id => genrelink)
  67.       #  unless db_has_genre?
  68.      #   genrelink;
  69.  
  70.       end          
  71.     end      
  72. #  end
Advertisement
Add Comment
Please, Sign In to add comment