sitemapからurlをランダムに取得して画像付きツイート

Dockerfile

FROM ruby:2.4

RUN gem install twitter nokogiri
COPY sitemap_to_tweet.rb /
CMD ["ruby", "sitemap_to_tweet.rb"]

sitemap_to_tweet.rb

require 'open-uri'
require 'twitter'
require 'kconv'
require 'zlib'
require 'nokogiri'

def generate_url_image_title
  if ENV['sitemap'] =~ /\.gz/
    xml = Nokogiri::XML Zlib::GzipReader.open(open(ENV['sitemap'])).read
  else
    xml = Nokogiri::XML open(ENV['sitemap'])
  end
  urls = xml.css('loc').map(&:text).select{|u| !(u =~ /tag/)}
  url = urls.sample
  if url =~ /\.xml/
    xml = Nokogiri::XML open(url)
    urls = xml.css('loc').map(&:text).select{|u| !(u =~ /tag/)}
    url = urls.sample
  end

  html = open(url, "r:binary").read
  page = Nokogiri::HTML(html.toutf8, nil, 'utf-8')

  image = title = nil
  page.search('meta').each do |meta|
    next unless meta.attribute('property')
    image = meta.attribute('content').text if meta.attribute('property').text == 'og:image'
    title = meta.attribute('content').text if meta.attribute('property').text == 'og:title'
  end
  [url, image, title]
end

def update(client)
  url, image, title = generate_url_image_title
  p url, image, title
  client.update_with_media(title + ' ' + url, open(image))
  #rescue TypeError
  #  'TypeError'
  #rescue => e
  #  p e
  #  sleep 60
  #  retry
end


client = Twitter::REST::Client.new do |config|
  config.consumer_key        = ENV['consumer_key']
  config.consumer_secret     = ENV['consumer_secret']
  config.access_token        = ENV['access_token']
  config.access_token_secret = ENV['access_token_secret']
end


p update(client)