etherpad-batch-import.rb

#!/usr/bin/env ruby
# lib/etherpad.rb
#
# mechanize => 1.0.0
%w{rubygems nokogiri mechanize json logger}.each {|l| require l }

class Etherpad
  def initialize(base_url, username, password)
    @@base_url = base_url
    @@username = username
    @@password = password
    @@robot = Mechanize.new do |a|
      `rm ../log/etherpad.log`
      log = Logger.new('../log/etherpad.log')
      log.level = Logger::INFO
      log.datetime_format = ""
      a.log = log
      a.user_agent_alias = 'Mac Safari'
    end
  end

  # slug2txt.
  def fetch(slug)
    return Pad.new(slug).to_s
  end

  # Fetches everything in the Etherpad namespace.
  def fetchall(toc_slug)
    todo = [toc_slug]  # List of slugs to fetch.
    done = []      # Stack of slugs already done.
    while !todo.empty?
      slug = todo.pop

      # Save contents to a file.
      begin
        pad = Pad.new(slug)
        mkd = pad.to_s
      rescue
        puts "404 on #{slug}."
        next
      end
      unless mkd.empty?
        mkd_file = File.new("../.cache/#{slug}.mkd", "w")
        mkd_file.write(mkd)
        puts "Wrote #{slug}."
        mkd_file.close
      end
      done << pad

      # Add links to processing.
      new_slugs = pad.slugs.reject {|slug| done.include? slug }
      new_slugs << slug + "-scratch"
      new_slugs << slug + "-sb"
      todo = (todo | new_slugs).uniq
    end
    puts "#{done.length} files processed."
  end

  def import(slug, file)
    pad = Pad.new(slug)
    pad.import(file)
    puts "Successfully imported #{file} into #{slug}"
  end

  def importzipfolder(folder)
    Dir.chdir(folder)
    Dir.foreach('.') do |file|
      next unless file[-5..-1] == ".html"
      import(file[0..-6], file)
    end
  end
end

class Etherpad::Pad < Etherpad
  def initialize(slug, rev = "latest")
    @base_url = @@base_url
    @username = @@username
    @password = @@password
    rev = "rev.#{rev}" unless rev == "latest"
    @slug = slug
    @url  = {
      :base => URI.join(@base_url, slug),
      :txt  => URI.join(@base_url, "ep/pad/export/#{slug}/#{rev}?format=txt"),
      :imp  => URI.join(@base_url, "ep/pad/impexp/import"),
      :imp2 => URI.join(@base_url, "ep/pad/impexp/import2"),
    }
  end

  # Regexp grep of slugs in pad with related @@base_url.
  def slugs
    tbr = []
    to_s.each_line do |line|
      # Takes care of URLs in form of http://etherpad.com/<slug>{...}
      line.gsub(/#{@base_url}(.*?)(\{.*?\})?(>|\s)/) do |match|
        slug = $1
        if $2.nil?
          tbr << slug
        else
          bracket = $2.delete("{}")
          bracket.split(',').each {|suffix| tbr << "#{slug}#{suffix}" }
        end
      end
    end
    return tbr.uniq
  end

  # Return full-text of the pad.
  def to_s
    page = @@robot.get(@url[:txt])
    # Check if we need to login.
    if page.methods.include? 'forms'
      form = page.forms.first
      form.email    = @username
      form.password = @password
      page = form.submit
    end
    return page.body
  end

  # Imports a file.
  def import(file)
    page = @@robot.get(@url[:base])
    # Check if we need to login.
    if page.title.include? "Sign In"
      form = page.forms.first
      form.email    = @username
      form.password = @password
      page = form.submit
    end

    # Check if we need to create the stub.
    form = page.form_with(:action => '/ep/pad/create')
    page = form.submit unless form.nil?


    # Fill out the form
    form = page.form_with(:action => '/ep/pad/impexp/import')
    form.file_uploads.first.file_name = file
    page = form.click_button

    token = page.body.split("'")[3]
    page = @@robot.post(@url[:imp2], {'padId' => @slug, 'token' => token})
    return true if page.body == 'ok'
    return false
  end
end

if __FILE__ == $0
  ep = Etherpad.new('http://YOUR_SUBDOMAIN.titanpad.com/',
                    'YOUR_EMAIL', 'YOUR_PASSWORD')
  ep.importzipfolder(ARGV[0])
end