Curate Tumblr Ruby Gem

Reblog contents, extract infos and follow contacts in your Tumblr, from a file of tumblr links.

Grow your Tumblr audience by automating the boring tasks !

gem install curate_tumblr

require 'curate_tumblr'

CurateTumblr.reblog( "kubricklove", "/home/tumblr" )

reblog links in the queue of tumblr kubricklove

where links to reblog are in file /home/tumblr/kubricklove/links/kubricklove_links

and config (with oauth) is in /home/tumblr/kubricklove/kubricklove_config.yaml

• add tumblr links to your tumblr queue
• follow tumblr source and author : in separate task to not to waste Tumblr requests
• polite with tumblr : random sleep and stop if detect tumblr errors
• extract informations from posts : other tumblrs to follow, external links to see...
• add visibility to your reblogs : add tags and link to your tumblr
• config in realtime : you can stop it or change parameters when running

Please note before all you have to config oauth for your tumblr (and put the codes in the config file, see below).

You can see an example of reblog and follow for a kubrick tumblr in the example directory.

First create a test tumblr if you have not

ex : mytumblrtest.tumblr.com

Follow the tumblr process to have oauth authenficiation : https://www.tumblr.com/login?redirect_to=%2Foauth%2Fapps

more infos : http://www.tumblr.com/docs/en/api/v2#blog-likes

This application must be authorized by your tumblr

At the end you must have these codes :

consumer_key: XX        
consumer_secret: XX     
token: XX               
token_secret: XX  

Curate Tumblr uses the name of your tumblr to find the directory for links (to reblog and follow) and config file

Create a directory where you will have your config and links files :

ex : /home/tumblr

Create in this directory another directory with your test tumblr name :

ex : /home/tumblr/mytumblrtest

In this directory create a directory links and a directory logs

ex :

/home/tumblr/mytumblrtest/links

/home/tumblr/mytumblrtest/logs

Create a config file in this directory.

You can copy kubricklove_config.yaml from the example and rename it with your tumblr name :

ex /home/tumblr/mytumblrtest/mytumblrtest_config.yaml

Put your oauth codes in the config file

oauth:

consumer_key:        
consumer_secret:     
token:               
token_secret: 

Congratulations! :-)

The easy way is to render tumblr links from file, with CurateTumblr::Render module :

require "curate_tumblr"

CurateTumblr.reblog( "mytumblrtest", "/home/tumblr" )

CurateTumblr.follow( "mytumblrtest", "/home/tumblr" )

You can also use CurateTumblr::Curator :

require "curate_tumblr"

curator = Curator.new( "mytumblrtest", "/home/tumblr" )

curator.reblog_and_extract( "http://joshbrooksonfilm.tumblr.com/post/58294445822/barry-lyndon-1975-kubrick-1st-viewing-if-i" )

curator.reblog_and_extract( "http://oh-lesedi.tumblr.com" )

client: basic config for manage Tumblr

• is_running : if false stop the application
• sleep_before_client_min and max : random sleep before each request (important to not be blocked by Tumblr)
• sleep_before_follow_min and max : random sleep before following, can simulate a queue
• max_requests_and_posts : max of all requests before stop
• max_posted : max of posted (text, image) before stop
• max_reblogged : max of reblogged before stop
• max_followed : max of followed before stop
• oauth: consumer_key, consumer_secret, token, token_secret

infos: infos to add to your posts

• tags : a string with list of tags
• title : to add a link to your tumblr

Before testing code you have to set oauth and create a config file (you can copy the example config file)

In spec/factories.rb, change get_tumblr_name and get_tumblr_directory to your tumblr name and your test directory

in the config file set oauth oauth:

consumer_key:        
consumer_secret:     
token:               
token_secret: 

The application manage these tumblr errors :

• Rate limit exceeded : too much requests, try to increase the sleep time in config file => the application stop
• No authorized : problem with oauth in the config file, tokens may be bad => the application stop
• Bad request : perhaps the link has been deleted
• Too much bad requests : there is no enough place in the queue, or there is a problem with your tumblr => the application stop

The application display little characters to tell what it's doing You can change it (or display nothing) in lib/curate_tumblr/values

• (4) : sleep 4 seconds
• 2> : reblog second post
• 3: : follow third tumblr
• ? : send request to have infos about a post

Config file is read before each operation.

You can stop a cron reblog or follow by changing is_running in config file (ex /home/tumblr/mytumblrtest/mytumblrtest_config.yaml)

is_running: false

you can also change other parameters (max to reblog, sleep time...)

If strange things happen :

• please note by default reblogged posts are published in queue (not visible)
• check there is enough place in the queue (max 300 posts in queue)
• check your tumblr name (ex : mytumblrtest if you have created mytumblrtest.tumblr.com)
• check oauth is ok in your config file
• check log file (ex in /home/tumblr/mytumblrtest/logs)

To reblog a post you need its reblog key.

To have it you have to send a request to tumblr with the post tumblr and id. In infos hash post from Tumblr you'll have the reblog key

You can use :

get_hash_post in CurateTumblr::Publish::Client to have the reblog key

or CurateTumblr::Tumblr::ExtractLinks.get_reblog_key_from_reblog_url if the url contains the reblog key (www.tumblr.com/reblog/...)

and reblog_post_key to reblog it in CurateTumblr::Publish::Reblog

• memory (to not follow tumblrs already followed)
• hidden oauth (are you enough paranoid ?)

The Curate Tumblr gem is Copyright (c) 2013 David Tysman and is licensed under the MIT License. Feel free to use it for any project.

Tumblr is Copyright (c) Tumblr, Inc. The Curate Tumblr gem is NOT affiliated with Tumblr, Inc.

www.davidtysman.com