Ruby module for backing up/archiving public Tumblr data.

tungle_client.rb

require 'typhoeus'
require 'json'

module TungleClient

  EXAMPLE_API_KEY = 'fuiKNFp9vQFvjLNvx4sUwti4Yb5yGutBN4Xh10LXZhhRKjWlV4'
  BASE_BLOG_URL   = 'https://api.tumblr.com/v2/blog'
  BASE_TAG_URL    = 'https://api.tumblr.com/v2/tagged'

  # NOTE: ssl_verifypeer is turned off here due to eternal font of Windows
  #       SSL glitches.
  def self.request(url, params={})
    params[:api_key] = EXAMPLE_API_KEY unless params[:api_key]
    Typhoeus::Request.new(
      url,
      method: :get,
      ssl_verifypeer: false,
      followlocation: true,
      params: params
    )
  end


  # -- FETCH HASHES
  # 
  # 

  def self.posts(blog_id, params={})
    JSON::parse(posts_json(blog_id, params))['response']
  end

  def self.blog_info(blog_id, params={})
    JSON::parse(blog_info_json(blog_id, params))['response']
  end

  def self.tagged(tag, params={})
    JSON::parse(tagged_json(tag, params))['response']
  end



  # -- FETCH JSON
  # 
  # 

  def self.posts_json(blog_id, params={})
    request(posts_api_url(blog_id), params).run.body
  end

  def self.blog_info_json(blog_id, params={})
    request(blog_info_api_url(blog_id), params).run.body
  end

  def self.tagged_json(tag, params={})
    params[:tag] = tag
    request(tagged_api_url, params).run.body
  end



  # -- FETCH AVATAR
  # 
  # 

  # Returns the effective url, not the contents of the file.
  def self.avatar(blog_id, params={})
    request(avatar_api_url(blog_id), params).run.options[:effective_url]
  end



  # -- URL FORMATTING
  # 
  # 

  def self.posts_api_url(blog_id)
    blog_base_api_url(blog_id) + '/posts'
  end

  def self.blog_info_api_url(blog_id)
    blog_base_api_url(blog_id) + '/info'
  end

  def self.avatar_api_url(blog_id)
    blog_base_api_url(blog_id) + '/avatar'
  end

  def self.blog_base_api_url(blog_id)
    'https://api.tumblr.com/v2/blog/' + blog_id
  end

  def self.tagged_api_url
    'https://api.tumblr.com/v2/tagged'
  end

end

tungle_client_readme.md

TUNGLE_CLIENT "DOCUMENTATION"

This is like tumblr_client except it doesn't do OAuth and thus is easier to configure, but only accesses public data, making it appropriate for making backups or archiving tags.

The API key supplied here appears in the examples in the official Tumblr API V2 documentation, and you can your own in the params hash if they pull it, or you just feel like doing that or something, I don't care. I am many hundreds of years old.

The methods accept all the parameters described in the appropriate sections in the Tumblr API V2 documentation:

• TungleClient::tagged and TungleClient::tagged_json

• TungleClient::posts and TungleClient::posts_json

• TungleClient::blog_info and TungleClient::blog_info_json

• TungleClient::avatar

Below are some very clear examples of how to use this module. These examples could not possibly be any clearer.

INDIVIDUAL BLOGS

Get a hash representing the first 20 posts from demo.tumblr.com (except it's only got 7 posts total):

TungleClient::posts('demo.tumblr.com')

Get the first 5 posts tagged 'doge' from snarp.tumblr.com:

TungleClient::posts('snarp.tumblr.com', tag: 'doge', limit: 5)

Get the second 5 posts tagged 'doge' from snarp.tumblr.com:

TungleClient::posts('snarp.tumblr.com', tag: 'doge', limit: 5, offset: 5)

Get all 10 of those doge posts again, this time with notes and reblog info:

TungleClient::posts('snarp.tumblr.com', tag: 'doge', limit: 10,  notes_info: true, reblog_info: true)

SITE-WIDE TAGS

Get the first 20 posts tagged 'doge' site-wide:

TungleClient::tagged('doge')

Get the latest posts tagged 'homestuck' as of 11:59 PM, October 25th 2011:

TungleClient::tagged('homestuck', before: 1319587199)

OTHER STUFF

Gets the userinfo for demo.tumblr.com:

TungleClient::blog_info('demo.tumblr.com')['blog']

Gets the url for the avatar for demo.tumblr.com:

TungleClient::avatar('demo.tumblr.com')

JSON

Get demo.tumblr.com posts as a JSON string rather than a hash:

TungleClient::posts_json('demo.tumblr.com')['posts']

And the userinfo as a JSON string rather than a hash:

TungleClient::blog_info_json('demo.tumblr.com')

And exactly two posts from the 'bats' tag:

TungleClient::tagged_json('bats', limit: 2)