Scrape Unsplash

unsplash.rb

require 'open-uri'
require 'nokogiri'
require 'fileutils'

class Scraper

  attr_accessor :url, :selector, :page

  def initialize(url, selector)
    @url = url
    @selector = selector
    @page = 1
  end

  def self.start
    scraper = Scraper.new('http://unsplash.com', '.photo_img')
    20.times do |page|
      scraper.page = (page + 1)
      scraper.images.each{|img| scraper.download(img.attr('src')) }
    end
  end

  def images
    document(page).css(selector)
  end

  def download(url)
    FileUtils.mkdir_p(File.join(Dir.pwd, 'photos'))
    path = File.join(Dir.pwd, 'photos', url.split('/').last)
    return if File.exist?(path)

    open(url) do |response|
      File.open(path, "wb") do |file|
        file.puts response.read
      end
    end
  end

  def document(page)
    raise "Invalid page" if page <=0
    Nokogiri::HTML(open("#{url}/page/#{page}"))
  end

end

Scraper.start