README.markdown

UPDATES:

• Will process files with either .markdown or .md file extension
• Set language to en_US

This is a very rough script to export your octopress files to Ghost. You will still need to check your data but it will give you a head start. Feel free to fork and optimize the script for your purpose.

The following should be considered:

1. You need to execute the script in the source/_posts directory
2. You need to manually copy all the images to /var/www/ghost/content/images
3. Images with external URLs are not handled
4. Blockquotes have their issues
5. Lists are not exported correctly

For the content I had to migrate it was a good compromise between automation and manual checking and corrections.

Hope this helps anyone ;-)

export_to_ghost.rb

#!/usr/bin/env ruby

require 'json'
require 'yaml'
require 'active_support/core_ext'
 
export_data = {
  meta: {
    version: "000"
  },
  data: {
    posts: [],
    tags: [],
    posts_tags: []
  }
}
 
class Tag
  class << self
    def ids
      @ids ||= 2000.step(1_000_000).to_enum
    end
 
    def next_id
      ids.next
    end
 
    def [](tag_name)
      repository[tag_name] ||= Tag.new(tag_name)
    end
 
    def repository
      @repository ||= {}
    end
  end
 
  attr_reader :name, :id
 
  def initialize(name)
    @name = name
    @id   = Tag.next_id
  end
 
  def slug
    name.parameterize
  end
 
  def to_json(*args)
    {
      id: id,
      name: name,
      slug: slug
    }
  end
end
 
class Post
  class << self
    def next_id
      ids.next
    end
 
    def ids
      @ids ||= 1000.step(1_000_000).to_enum
    end
  end
 
  attr_reader :id, :tags
 
  def initialize(raw)
    @metadata   = YAML.load raw
    @content    = raw.split('---').last.strip
    @id         = Post.next_id
    @tags       = [@metadata["categories"]].flatten.collect { |c_name| Tag[c_name] }
  end
 
  def title
    @metadata["title"]
  end
 
  def slug
    title.parameterize
  end
 
  def content
    remove_more
    fix_images
    strip_newlines
    fix_blockquotes
 
    @content
  end
 
  def timestamp
    @metadata["date"].to_time.to_i * 1000
  end
 
  def as_json(*args)
    {
      id: id,
      title: title,
      slug: slug,
      markdown: content,
      image: nil,
      featured: 0,
      page: 0,
      status: "published",
      language: "en_US",
      author_id: 1,
      created_at: timestamp,
      created_by: 1,
      updated_at: timestamp,
      updated_by: 1,
      published_at: timestamp,
      published_by: 1
    }
  end
 
  private
 
  def remove_more
    @content.gsub! '<!-- more -->', ''
  end
 
  def fix_images
    @content.gsub!(/\{% img.*?(\/.*?) \d+ \d+\s?(.*?) %}/, '![\2](\1)')
    @content.gsub!('images', 'content/images')
  end
 
  def strip_newlines
    @content = @content.strip.gsub(/(.)\n(.)/, '\1 \2')
  end
 
  def fix_blockquotes
    @content.gsub!(/{% blockquote %}(.*?){% endblockquote %}/, '> \1')
  end
end
 
tag_ids = 3000.step(1_000_000).to_enum
 
posts = Dir.glob("*.{markdown,md}")

posts.each do |post_file|

  post = Post.new File.read post_file
 
  export_data[:data][:posts] << post.as_json
 
  post.tags.each do |tag|
    export_data[:data][:posts_tags] << {
      id: tag_ids.next,
      post_id: post.id,
      tag_id: tag.id
    }
  end
end
 
export_data[:data][:tags] = Tag.repository.values
 
puts export_data.to_json