#!/usr/bin/env ruby

require 'open-uri'
require 'rubygems'
require 'html/tree'
require 'html/xpath'
require 'feed_tools'

DIR_PREFIX = '/home/joelmichael/stuff.ediblepet.net/feeds'

uri = URI.parse('http://70.86.201.113/imageserv2/temporary/archivelist.html')
remote_html = uri.read

cached_html_file = File.open("#{DIR_PREFIX}/cache/pbf.html", 'r+')

unless remote_html == cached_html_file.read
    cached_html_file.truncate 0
    cached_html_file.seek 0
    cached_html_file.write remote_html

    parser = HTMLTree::Parser.new
    parser.feed remote_html
    anchors = parser.tree.rexml_match('//html/body/font/center/a')

    feed = FeedTools::Feed.new
    feed.id = 'http://stuff.ediblepet.net/feeds/pbf.xml'
    feed.link = 'http://cheston.com/pbf/archive.html'
    feed.title = 'Perry Bible Fellowship'
    feed.feed_type = 'atom'
    feed.author = 'Nicholas Gurewitch'

    # No dates are provided, so sadly this is a one-entry feed
    # unless I decide to make a cacheing system
    anchor = anchors.shift
    item = FeedTools::FeedItem.new
    item.author = 'Nicholas Gurewitch'
    filename = anchor.attribute('href').value
    dir = "http://70.86.201.113/imageserv2/temporary"
    item.id = "#{dir}/#{filename}"
    item.link = item.id
    image_page_uri = URI.parse(item.link)
    image_page_html = image_page_uri.read
    doc = REXML::Document.new image_page_html
    image = REXML::XPath.match(doc, '//img').shift
    image_src = image.attribute('src').value
    item.title = anchor.text
    item.content = "<img src=\"#{dir}/#{image_src}\" alt=\"#{item.title}\"/>"
    feed << item

    xml_output_file = File.open("#{DIR_PREFIX}/pbf.xml", 'w')
    xml_output_file.write(feed.build_xml)
    xml_output_file.close
end

cached_html_file.close