tumblr_photo_backuper.rb

#!/usr/bin/env ruby
Dir.chdir(File.dirname(__FILE__))
require 'digest/sha2'
require 'uri'
require 'open-uri'
require 'json'

API_KEY = 'api_key'
TUMBLELOG = 'scrap.sorah.jp'

def get_photos(offset=0)
  response = open("http://api.tumblr.com/v2/blog/#{TUMBLELOG}/posts/photo?api_key=#{API_KEY}&offset=#{offset}", 'r', &:read)
  JSON.parse(response)
end

def generate_filename(post_url, img_url)
  m = post_url.match(%r{/post/(\d+?)(\Z|/)})
  post_id = m ? m[1] : ''

  img_ext = File.extname(URI.parse(img_url).path)

  "#{post_id}_#{Digest::SHA256.hexdigest(img_url)}#{img_ext}"
end

downloaded = true
offset = 0
while downloaded
  downloaded = false
  payload = get_photos(offset)
  puts "============ Offset: #{offset}"

  payload['response']['posts'].each do |item|
    link = item['post_url']
    puts "=> #{link}"

    img_urls = item['photos'].map { |_| _['original_size']['url'] rescue nil }.compact
    img_urls.each do |url|
      filename = generate_filename(link, url)
      if File.exist?(filename)
        next
      end

      puts "*  #{url} => #{filename}"
      system "curl", "--progress-bar", "-o", filename, url
      downloaded = true
    end
  end
  offset += 20
end