Created
December 5, 2012 18:39
-
-
Save hollanddd/4218286 to your computer and use it in GitHub Desktop.
screen scraping practice with nokogiri
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
require 'rubygems' | |
require 'nokogiri' | |
require 'open-uri' | |
def wine_woot(url) | |
doc = Nokogiri::HTML(open(url)) | |
data = doc.xpath('//*[@id="summary"]/div') | |
puts doc.xpath('html/head/title').text + ' as site_name' | |
puts url + data.xpath('hgroup/a/@href').to_s + ' as site_href' | |
puts doc.xpath('//*[@id="todays-deal"]/a/img/@src').to_s + ' as img_src' | |
puts data.at_css('h2.fn').text + ' as wine_name' | |
puts data.at_css('.price').text + ' as price' | |
end | |
def cinderella_wine(url) | |
doc = Nokogiri::HTML(open(url)) | |
puts doc.xpath('html/head/title').text.split('-').first.strip + ' as site_name' | |
puts "#{url} as site_href" | |
puts doc.xpath('//*[@id="bottle-shot"]/a/img/@src').to_s + 'as img_src' | |
puts doc.xpath('//*[@id="title"]/h2/a').text + ' as wine_name' | |
puts doc.xpath('//*[@id="product-dollars"]').text.strip + "." + doc.xpath('//*[@id="product-cents"]').text.strip + ' as price' | |
end | |
def last_call_wines(url) | |
doc = Nokogiri::HTML(open(url)) | |
puts doc.xpath('html/head/title').text.split('|').last.strip + ' as site_name' | |
puts "#{url} as site_href" | |
puts "#{doc.xpath('//*[@id="divProductPic"]/img/@src').to_s} as img_src" | |
puts "#{doc.at_css('h3.ProductNameText2').text} as wine_name" | |
puts "#{doc.at_css('td.SalePrice/div.price').text.strip} as price" | |
end | |
def bacchus_selections(url) | |
doc = Nokogiri::HTML(open(url)) | |
puts doc.xpath('html/head/title').text.split('-').last.strip | |
puts "#{url} as site_href" | |
puts "#{url}#{doc.xpath('//*[@id="content"]/div[1]/div[2]/a/img/@src').to_s.split('?').first} as img_src" | |
puts "#{doc.xpath('//*[@id="content"]/div[1]/div[1]/h2/a').text} as wine_name" | |
puts "#{doc.xpath('//*[@id="content"]/div[1]/div[3]/p[1]').text.strip} as price" | |
end | |
def wine_spies(url) | |
doc = Nokogiri::HTML(open(url)) | |
puts doc.xpath('html/head/title').text.split('-').first.strip | |
puts "#{url} as site_href" | |
puts "#{url}#{doc.xpath('//*[@id="wine-thumb"]/@src').to_s.split('?').first} as img_src" | |
puts "#{doc.at_css('h2#wine-name').text.strip.split.join(" ")} as wine_name" | |
puts "#{doc.at_css('td.our-price/div.inner').text.strip} as price" | |
end | |
url = %w{ http://www.wine.woot.com http://www.cinderellawine.com http://www.lastcallwines.com http://www.bacchusselections.com http://www.thewinespies.com http://www.cellarthief.com } | |
url.each do |u| | |
case u | |
when 'http://www.wine.woot.com' | |
wine_woot u | |
when 'http://www.cinderellawine.com' | |
cinderella_wine u | |
when 'http://www.lastcallwines.com' | |
last_call_wines u | |
when 'http://www.bacchusselections.com' | |
bacchus_selections u | |
when 'http://www.thewinespies.com' | |
wine_spies u | |
when 'http://www.cellarthief.com' | |
cellar_thief u | |
end | |
end |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment