root/lang/ruby/misc/tumblr_getimages.rb

#
# 指定したIDのTumblrの全画像をカレントディレクトリに保存する
#

require 'open-uri'
require 'rexml/document'

if ARGV.size != 1
  $stderr.puts "Usage: #{$0} id "
  exit 64
end

id = ARGV[0]

api = URI.parse("http://#{id}.tumblr.com/api/read")
api.query = "type=photo"
doc = REXML::Document.new(api.read)
total = REXML::XPath.first(doc, "/tumblr/posts").attributes['total'].to_i

$stdout.sync = true

num = 50
index = 0
while index < total
  num = [num, total - index].min
  api.query = "type=photo&num=#{num}&start=#{index}"
  doc = REXML::Document.new(api.read)
  posts = REXML::XPath.match(doc, "/tumblr/posts/post")
  photo_list = posts.map {|post| URI.parse(post.elements['photo-url'].text)}
  limit = index + num
  while index < limit
    print "\r#{index + 1}/#{total}"
    begin
      photo_url = photo_list.shift
      open(photo_url.path.split(/\//).last, "wb") do |image_file|
        image_file.write(photo_url.read)
      end
    rescue
      $stderr.puts "Error: #{photo_url}"
    end
    index += 1
  end
end
puts