Download all the images from a Tumblr blog

tumblr-photo-ripper.rb

# Usage:
#   [sudo] gem install mechanize
#   ruby tumblr-photo-ripper.rb

require 'rubygems'
require 'mechanize'

# Your Tumblr subdomain, e.g. "jamiew" for "jamiew.tumblr.com"
site = "doctorwho"


FileUtils.mkdir_p(site)

concurrency = 8
num = 50
start = 0

loop do
  puts "start=#{start}"

  url = "http://#{site}.tumblr.com/api/read?type=photo&num=#{num}&start=#{start}"
  page = Mechanize.new.get(url)
  doc = Nokogiri::XML.parse(page.body)

  images = (doc/'post photo-url').select{|x| x if x['max-width'].to_i == 1280 }
  image_urls = images.map {|x| x.content }

  image_urls.each_slice(concurrency).each do |group|
    threads = []
    group.each do |url|
      threads << Thread.new {
        puts "Saving photo #{url}"
        begin
          file = Mechanize.new.get(url)
          filename = File.basename(file.uri.to_s.split('?')[0])
          file.save_as("#{site}/#{filename}")
        rescue Mechanize::ResponseCodeError
          puts "Error getting file, #{$!}"
        end
      }
    end
    threads.each{|t| t.join }
  end

  puts "#{images.count} images found (num=#{num})"
  if images.count < num
    puts "our work here is done"
    break
  else
    start += num
  end

end

Awesome, exactly what i needed, thanks

Very nice code, it is exactly what i need!

I will give you just some advises, if u wanna spread this beautiful piece of code.

1. Make a .exe .

2. A very simple Graphical Interface

3. Be happy

Why have constants when you could append the link at the end of the command, and have it strip off everything but the blog name

Nice tool. Helped me save 500+ images. Thanks.

Great script! Thank you so much!

Successfully done for 16290 images (3.8G). Thank you!

Thanks a lot. Downloaded 5800 images.

Hi, I'm really sorry, I don't know anything about programming, but I'm desperately looking for some software or way of downloading images from a tumblr. Is there some way I can use this script with my ignorance? Thanks in advanced.

Version in Perl to grab every single image instead, surpassing XML parsing etc.:

#!/usr/bin/perl

use strict;

use LWP::Simple;

my $site = 'foo';

foreach (my $i = 0; $i < 9999; $i += 50) {
    my $url = "http://$site.tumblr.com/api/read?type=photo&num=50&start=$i";
    warn "Retrieving $url\n";

    my $src = get($url);

    foreach my $image( $src =~ m{https://[^\s]+\.jpg}g ) {
        my ($filename) = $image =~ m{/([^/]+\.jpg)};
        print "found $filename\n";
        
        getstore($image, "pictures/$filename");
    }
}