neinomaten/neinomat-radio-chaotica

#!/usr/bin/env ruby
$:.unshift('vendor/ruby-mediawiki/lib')
require 'mediawiki/dotfile'
require 'cgi'
require 'date'
require 'erb'
require 'net/http'

#----------------------------------------------------------------------------
# Constants

RFC822 = "%a, %d %b %Y %T %z" # for Date::strftime
TPL_RSS = '<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" 
  xmlns:dc="http://purl.org/dc/elements/1.1/"
  xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
  xmlns:admin="http://webns.net/mvcb/"
  xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
  xmlns:content="http://purl.org/rss/1.0/modules/content/">

<channel>
<title>radio chaotica podcast</title>
<link>http://entropia.de/wiki/Radio%20Chaotica</link>
<description>ein podcast der sendungen des entropia e.v., chaos computer club karlsruhe auf querfunk, 104.8 mhz</description>
<category>Talk Radio</category>
<generator>MediaWiki neinomat</generator>
<language>de</language>
<copyright>Entropia e.V. Karlsruhe</copyright>

<managingEditor>radio@entropia.de (Radio Chaotica)</managingEditor>
<pubDate><%=Time.now.strftime(RFC822) %></pubDate>
<image>
  <url>http://podcast.entropia.de/chaotica-100x300-crappy.png</url>
  <width>100</width>
  <height>300</height>
  <link>http://entropia.de/wiki/Radio%20Chaotica</link>
  <title>Radio Chaotica</title>
</image>

<sy:updatePeriod>daily</sy:updatePeriod>
<sy:updateFrequency>1</sy:updateFrequency>
<sy:updateBase>2000-01-01T12:00+00:00</sy:updateBase>

<itunes:subtitle xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd">ein podcast der sendungen des entropia e.v., chaos computer club karlsruhe auf querfunk, 104.8 mhz</itunes:subtitle>
<itunes:summary xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd">
ein podcast der sendungen des entropia e.v., chaos computer club karlsruhe auf querfunk, 104.8 mhz
</itunes:summary>
<itunes:image xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd" href="http://podcast.entropia.de/chaotica-300x300-crappy.png"/>

<itunes:category xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd" text="News &amp; Politics"/>
<itunes:keywords xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd">entropia,karlsruhe,ccc,hacking,computer,technology,society,chaosradio,chaos,politics</itunes:keywords>
<itunes:explicit xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd">no</itunes:explicit>

<itunes:author xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd">Entropia e.V. Karlsruhe</itunes:author>
<itunes:owner xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd">
  <itunes:name>Entropia e.V. Karlsruhe</itunes:name>
  <itunes:email>radio@entropia.de</itunes:email>
</itunes:owner>

<%= items %>
</channel></rss>
'
TPL_ITEM = '
<item>
  <title><%= CGI::escapeHTML(episode["title"]) %></title> 
  <link><%= episode["url"] %></link>
  <description><%= CGI::escapeHTML(episode["title"]) %></description>
  <category>Talk Radio</category>
  <enclosure url="<%= episode["enclosure_url"] %>" length="<%= episode["length"] %>" type="<%= episode["type"] %>" />

  <guid isPermaLink="true"><%= episode["url"] %></guid>
  <pubDate><%= episode["pubdate"] %></pubDate>
  <comments><%= episode["discussion"] %></comments>
  <content:encoded><%= CGI::escapeHTML(episode["title"]) %></content:encoded>

  <author>radio@entropia.de</author>

  <itunes:explicit xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd">no</itunes:explicit>
  <itunes:author xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd">Entropia e.V. Karlsruhe</itunes:author>
  <itunes:keywords xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd">entropia,karlsruhe,ccc,hacking,computer,technology,society,chaosradio,chaos,politics</itunes:keywords>
  <itunes:duration xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd"><%= episode["duration"] %></itunes:duration>
  <itunes:subtitle xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd"><%= CGI::escapeHTML(episode["title"]) %></itunes:subtitle>
  <itunes:summary xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd"><%= CGI::escapeHTML(episode["title"]) %></itunes:summary>
</item>
'

#----------------------------------------------------------------------------
# Some helper functions

def parse_template(title, template)
  template_re = Regexp.new("\\\{\\\{" + template + "(.*?)\\\}\\\}")

  fields = {}

  fields_string = @wiki.article(title).text.scan(template_re)[0][0]

  inlink = 0
  field = ""
  fields_string += "|" unless fields_string.match(/\|$/)
  fields_string.split(//).each do |c|
    if c == "|" && inlink == 0
      (key, value) = field.split(/=/)
      fields[key] = value
      field = ""
    else
      inlink += 1 if c == "["
      inlink -= 1 if c == "]"
      field += c
    end
  end

  fields
end

def datum2isodate(datum)
  return datum[6..9] + "-" + datum[3..4] + "-" + datum[0..1]
end

def head(url)
  uri = URI.parse(url)
 
  Net::HTTP.start(uri.host, uri.port) do |http|
    http.head(uri.path)
  end
end

#----------------------------------------------------------------------------
# Get episodes from wiki

@wiki = MediaWiki.dotfile(nil,'entropia')

template = "Vorlage:Radio Chaotica-Sendung"
episodes = []
@wiki.article(template).what_links_here.each do |page|
  # puts page # DEBUG
  if erste_sendung = parse_template(page, template)["erste_sendung"]
    episode = {
      "title"         => page,
      "url"           => @wiki.article_url(page),
      "discussion"    => @wiki.article_url("Diskussion:#{page}"),
      "enclosure_url" => parse_template(page, template)["download"],
      "date"          => Date.today(), # fallback
    } 
    

    # Get a real date
    erste_sendung.gsub!(/\s*um\s*\d+:\d+$/,"")
    if erste_sendung !~ /^[0-9]{2}\.[0-9]{2}\.[0-9]{4}.*$/
      $stderr.puts "Field '#{erste_sendung}' in #{episode["url"]} looks funny, fall back to today."
    else
      episode["date"] = Date.parse(datum2isodate(erste_sendung))
    end
    episode["pubdate"] = episode["date"].strftime(RFC822)
  
 
    # Get content type and length
    head = head(episode["enclosure_url"])
    episode["length"] = head["content-length"]
    episode["type"]   = head["content-type"]

    # We just assume that the episode's length is an hour or so
    episode["duration"] = "00:59:59"

    episodes.push(episode)
  end
  $stderr.puts "DEBUG: Only crawling one episode"; break 
end

episodes = episodes.sort do |x,y| 
  y["date"] <=> x["date"]
end

#----------------------------------------------------------------------------
# Generate RSS

items = ""
episodes.each do |episode|
  items += ERB.new(TPL_ITEM).result(binding);
end

rss = ERB.new(TPL_RSS).result;
#puts rss #DEBUG

rssfile = File.new("/home/neingeist/public_html/chaotica.xml", "w")
rssfile.puts rss
rssfile.close