Image scraping in Ruby

问题:

How do I scrape an image present on a particular URL using Nokogiri? If there are better options than Nokogiri please suggest. The css image tag is .profilePic img

回答1:

If it is just an <img> with a URL:

PAGE = "http://site.com/page.html"
require 'nokogiri'
require 'open-uri'
html = Nokogiri.HTML(open(PAGE))
src  = html.at('.profilePic img')['src']
File.open("foo.png", "wb") do |f|
  f.write(open(src).read)
end

If you need to turn a relative image path into an absolute, see:
https://stackoverflow.com/a/4864170/405017

回答2:

The lazy way is to use mechanize as it will figure out the urls and filenames for you:

require 'mechanize'
agent = Mechanize.new
doc = agent.get(url)
agent.get(doc.parser.at('.profilePic img')['src']).save