tumblr 2 siteleaf

tumblr_2_siteleaf.rb

#!/usr/bin/env ruby

require 'tumblr_client'
require 'siteleaf'
require 'sanitize'
require 'pry'
require 'json'


##
## User Configuration
##

CONF = {

  ## Tumblr configuration
  total_posts: 300, # number of posts to collect.
  t_user: 'your tumblr blog name',
  t_key: 'your tumblr api key',
  t_secret: 'your tumblr api secret',

  ## Siteleaf configuration
  sl_page_title: 'the siteleaf page name to import into',
  site_id: 'your siteleaf site id',
  sl_key: 'your siteleaf api key',
  sl_secret: 'your siteleaf api secret',

  ## Useful if the process fails allong the way…
  offset: 0,
}



##
## Client Configuration
##

Tumblr.configure do |c|
  c.consumer_key = CONF[:t_key]
  c.consumer_secret = CONF[:t_secret]
end

Siteleaf.api_key    = CONF[:sl_key]
Siteleaf.api_secret = CONF[:sl_secret]



##
## Post Class
##

class Post
  attr_reader :post, :published_at, :type

  def initialize(post)
    @post = post
    @published_at = Time.at(post['timestamp'])
    @type = post['type']
  end

  def title
    return @caption if defined? @caption
    # Lots of junk characters…
    @caption = Sanitize.clean(post['caption']).to_s.gsub(/\s+/, ' ').strip
    @caption = unless caption.empty?
         caption.sub(/( —.+| -.+| \(.+\))$/, '').strip
      else
        post['id'].to_s
      end
  end

  def body
    return @body if defined? @body
    caption = Sanitize.clean(post['caption'], {
      elements: %w[b em i strong u a],
      attributes: { 'a' => ['href'] },
      add_attributes: { 'a' => {'rel' => 'nofollow'} },
      protocols: { 'a' => {'href' => ['http', 'https', :relative]}}
    })
    @body = if caption
        caption.sub(/(\(.+via.+\))/, '').strip
      elsif post['text']
        post['text']
      end
  end

  def meta
    return @meta if defined? @meta
    @meta = [{ key: 'type', value: type }]
    add_source_meta
    add_cite_meta
    add_embed_meta
    @meta
  end

  def photos
    return @photos if defined? @photos
    @photos = if post['photos']
        post['photos'].map { |p| p['original_size']['url'] }
      else
        []
      end
  end

  def taxonomy
    return @taxonomy if defined? @taxonomy
    @taxonomy = unless post['tags'].empty?
        { taxonomy: [{ key: 'Tags', values: post['tags'] }] }
      else
        {}
      end
  end

  def params
    return @params if defined? @params
    @params = {
      title: title.to_s,
      body: body.to_s,
      meta: meta,
      published_at: published_at
    }.merge!(taxonomy)
  end

  private

    def add_source_meta
      if post['source_url']
        @meta << { key: 'source', value: post['source_url'] }
      elsif post['caption']
        post['caption'].match(/\(<a href=\"(.+)\" target=\"_blank\">via<\/a>\)/) do |m|
          @meta << { key: 'source', value: m[1] }
        end
      end
    end

    def add_cite_meta
      @meta << { key: 'cite', value: post['source'] } if post['source']
    end

    def add_embed_meta
      if post['permalink_url']
        service, id = parse_video_url(post['permalink_url'])
        @meta << { key: "embed_#{service}", value: id } if service && id
      end
    end

    def parse_video_url(video_url)
      if m = video_url.match(/^.*(?:youtu.be\/|v\/|u\/\w\/|embed\/|watch\?v=|\&v=)([^#\&\?]*).*/)
        ['youtube', m[1]]
      elsif m = video_url.match(/^.+vimeo.com\/(?:.*\/)?([^#\?]*)/)
        ['vimeo', m[1]]
      end
    end
end



##
## Fetch Tumblr posts
##

client = Tumblr::Client.new
total_pages = (CONF[:total_posts] / 20).floor
limit = CONF[:total_posts] > 20 ? 20 : CONF[:total_posts]

tumblr_posts = (0..total_pages).map { |i|
  posts = client.posts("#{CONF[:t_user]}.tumblr.com", {
    limit: limit,
    offset: i * 20 + CONF[:offset]
  })
  posts['posts']
}.flatten


##
## Siteleaf
##

site = Siteleaf::Site.find(CONF[:site_id])
unless sl_page = site.pages.detect { |p| p.title == CONF[:sl_page_title] }
  sl_page = Siteleaf::Page.create({
    site_id: CONF[:site_id],
    title:   CONF[:sl_page_title]
  })
end

tumblr_posts.each_with_index do |p, i|
  post = Post.new(p)

  # I’m skipping link and text posts. Sorry.
  next if %w[text link].include? post.type

  puts "Importing post #{p['id']} (#{i+1+CONF[:offset]} of #{tumblr_posts.length+CONF[:offset]})"

  sl_post = Siteleaf::Post.create(post.params.merge({
    parent_id: sl_page.id
  }))

  post.photos.each_with_index do |photo, i|
    puts "    creating asset #{i+1} of #{post.photos.length}"

    Siteleaf::Asset.create({
      post_id: sl_post.id,
      url: photo
    })
  end
end