libr-2/app/workers/process_post_worker.rb

class ProcessPostWorker
  include Sidekiq::Worker

  sidekiq_options queue: :post, retry: 3

  def perform url
    post = Post.find_by url: url
    begin
      page =  MetaInspector.new post.url
      post.title = page.title.try(:strip)
      post.summary = page.description
      post.processed = true
      if page.images.best.present?
        post.create_thumbnail file: URI.parse(page.images.best)
      end
      post.save
    rescue SocketError => e
      Rails.logger.error "出错了： #{e}"
    ensure
      if post.url.include?('//mp.weixin.qq.com/')
        get_content_by_nokogirl post
      else
        readability_extract_content(post)
      end
    end

  end

  private

  def readability_extract_content post
    token = 'f0365ca0cf184783625c676e3373b329d7b7f4f4'
    begin
      response = RestClient.get "https://readability.com/api/content/v1/parser?token=#{token}&url=#{post.url}"
      if response.code == 200
        post_json = JSON.parse response.body

        post.content = HTMLEntities.new.decode post_json['content']
        post.good_format = true if post.type != 'WebPage'
        post.title = post_json['title'] unless post.title
        post.create_thumbnail(file: URI.parse(post_json['lead_image_url'])) if post_json['lead_image_url'].present?
        post.save
      end
    rescue => e
      Sidekiq.logger.error "ReadAbility 获取文章失败：#{e}"
    end

  end

  def get_content_by_nokogirl post
    page = Nokogiri::HTML RestClient.get(post.url)
    content_html = page.css('div#js_content')[0].try(:to_html)
    post.content = content_html
    post.good_format = true if post.type != 'WebPage'
    post.save!
  end

end