#!/usr/bin/ruby

# ----------------------------------------------------
# This software is provided AS-IS, with NO WARRANTY. 
# Don't blame me if you have problems.
# ----------------------------------------------------
#
# Pebble-to-WordPress conversion script.
#
# Original Author: Vamsi Krishna
# Author of This Version: John Casey
#
# Based on:
#    http://code.google.com/p/pebble2wordpress/
#
# For more information on running the original version, see:
#    http://www.spritle.com/blogs/?p=107#more-107
#
# For more information running THIS script, see:
#    http://www.johnofalltrades.name/?p=140
# ----------------------------------------------------

require 'rubygems'
require 'rexml/document'
require 'active_record'
require 'yaml'
require 'time'

basedir = "inputXML"

dbconfig = YAML::load( File.open( 'database.yml' ) )

ActiveRecord::Base.establish_connection( dbconfig )

class Term < ActiveRecord::Base
  set_table_name "wp_terms"
end

class TermRelationship < ActiveRecord::Base
  set_table_name "wp_term_relationships"
end

class Posts< ActiveRecord::Base
  set_table_name "wp_posts"
end

class Comments< ActiveRecord::Base
  set_table_name "wp_comments"
end

def time_of( element, xpath )
  txt = element.get_text( xpath )
  return DateTime.now unless txt
  
  # 17 Jul 2006 14:55:15:637 +0000
  value = ''
  value = txt.value.split( /:\d\d\d/ )[0]
  
  return DateTime.strptime( value, "%d %b %Y %H:%M:%S" )
end

def text_of( element, xpath )
  txt = element.get_text( xpath )
  
  return txt && txt.value ? txt.value : ''
end

def count_of( element, xpath )
  elts = element.elements.to_a( xpath )
  return elts ? elts.size : 0
end

def path_part( title )
  result = ''
  title.downcase.split( /[^a-zA-Z0-9]+/ ).each do |part|
    result << '-' if ( result.length > 0 )
    result << part
  end
  
  return result
end

term_map = {}
Term.find(:all).each do |term|
  term_map[term.slug] = term.term_id
end

count = 0
skipped = []
Dir.chdir( basedir ) do
  files = Dir.glob("*.xml").each do |file|
    if File.file?( file )
      # begin
        document = REXML::Document.new File.new( file )
        
        title = text_of( document.root, 'title' )
        if ( title.length < 1 )
          puts "Skipping post with empty title: #{file}"
          skipped << file
          next
        end
        
        puts "Adding: '#{title}'..."
        
        post = Posts.new

        date = time_of( document.root, 'date' )
        body = text_of( document.root, 'body' )
        body.gsub!( "img src=\"images/", "img src=\"/wp-content/uploads/" )
        
        post.post_author = 1
        post.post_date = date
        post.post_date_gmt = date
        post.post_content = body
        post.post_title = title
        post.post_excerpt = text_of( document.root, 'excerpt' )
        post.post_status = "publish"
        post.comment_status = "open"
        post.ping_status = ""
        post.post_password = ""
        post.post_name = path_part( title )
        post.to_ping = ""
        post.pinged = ""
        post.post_modified = date
        post.post_modified_gmt = date
        post.post_content_filtered = ""
        post.post_parent = 0
        
        post.menu_order = 0
        post.post_type = "post"
        post.post_mime_type = ""
        
        comment_count = count_of( document.root, '//comment' )
        post.comment_count = comment_count
        post.save
        post.reload
        
        id = post.ID
        
        post.guid = "/?p=#{id}"
        post.save
        
        document.root.each_element( 'category' ) do |category_element|
          category = category_element.text.downcase
          if ( category )
            category = category[1..-1] if category[0..0] == '/'
            category = path_part( category )
            
            term_id = term_map[category]
          
            puts "...adding to category '#{category}' with term_id '#{term_id}'"
          
            rel = TermRelationship.new
            rel.object_id = id
            rel.term_taxonomy_id = term_id
            rel.term_order = 0
          
            rel.save
          end
        end
        
        if ( comment_count > 0 )
          puts "...adding #{comment_count} comments..."
        
          document.root.each_element( '//comment' ) do |element|
          
            comment = Comments.new
            comment.comment_post_ID = id
            comment.comment_author = text_of( element, 'author' )
            comment.comment_author_email = text_of( element, 'email' )
            comment.comment_author_url = text_of( element, 'website' )
            comment.comment_author_IP = text_of( element, 'ipAddress' )
            comment.comment_date = time_of( element, 'date' )
            comment.comment_date_gmt = time_of( element, 'date' )
            comment.comment_content = text_of( element, 'body' )
            comment.comment_karma = 0
            comment.comment_approved = text_of( element, 'state' ) == 'approved' ? 1 : 0
            comment.comment_agent = ""
            comment.comment_type = ""
            comment.comment_parent = 0
            comment.user_id = 0
            comment.save
          
          end
        end
        
        puts "...done."
        
        count = count.next
      # rescue
      #   puts "Failed to parse #{file}. Error: " << $!
      # end
    else
      puts "Skipping #{file}"
    end
  end
end

puts "Processed #{count} xml files successfully!"
puts "\nSkipped the following due to processing errors:\n\n#{skipped.join( '\n- ' )}" if skipped.size > 0