John Cleese Groovy Specs

[jhb50]

This feed yields an xml page but the links are not in standard rss format, so it requires a groovy to be programmed which I do not yet know how to do.

http://www.johncleesepodcast.co.uk/cleeseblog/rss.xml as found in http://johncleesepodcast.co.uk/cleeseblog/
He is an excellent english comedian.

The items contain the valid links within <description><A HREF=" tags as follows. Some of the items are audio (.mp3) so only the video (.mp4) items should be displayed.

  Code:

 <item>
  <guid isPermaLink="false">tag:blogger.com,1999:blog-17675311.post-5983787872073372504</guid>
  <pubDate>Tue, 09 Jun 2009 15:30:00 +0000</pubDate>
  <atom:updated>2009-06-09T15:33:09.837Z</atom:updated>
  <category domain="http://www.blogger.com/atom/ns#">video</category>
  <category domain="http://www.blogger.com/atom/ns#">downloads</category>
  <title>A Blatant Audio Advert for Headcast Videos</title>
  <description><A HREF="http://thejohncleese.com/podules/jcpod_june_2009.mp3">A short audio piece extolling the merits of John's new videos</A>.<br /><br />Visit <A HREF="http://headcast.co.uk/">headcast.co.uk</A> for more moments.<div class="blogger-post-footer"><img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/17675311-5983787872073372504?l=www.johncleesepodcast.co.uk%2Fcleeseblog' alt='' /></div></description>
  <link>http://www.johncleesepodcast.co.uk/cleeseblog/2009/06/blatant-audio-advert-for-headcast.html</link>
  <author>noreply@blogger.com (Dean Whitbread)</author>
  <thr:total xmlns:thr="http://purl.org/syndication/thread/1.0">6</thr:total>
  </item>
- <item>
  <guid isPermaLink="false">tag:blogger.com,1999:blog-17675311.post-2025308770889487821</guid>
  <pubDate>Fri, 03 Oct 2008 19:53:00 +0000</pubDate>
  <atom:updated>2008-10-03T19:58:02.144Z</atom:updated>
  <category domain="http://www.blogger.com/atom/ns#">education education education</category>
  <category domain="http://www.blogger.com/atom/ns#">speech</category>
  <category domain="http://www.blogger.com/atom/ns#">Headmaster</category>
  <title>John Cleese Podcast #34: The Headmaster</title>
  <description><A HREF="http://funkwarehouse.com/jcpods/john_cleese_podcast_34.mp4">John reminds us of that much-loved first day of a new academic year, as he greets the school with the traditional Headmaster's speech.</A><div class="blogger-post-footer"><img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/17675311-2025308770889487821?l=www.johncleesepodcast.co.uk%2Fcleeseblog' alt='' /></div></description>
  <link>http://www.johncleesepodcast.co.uk/cleeseblog/2008/10/john-cleese-podcast-34-headmaster.html</link>
  <author>noreply@blogger.com (Dean Whitbread)</author>
  <thr:total xmlns:thr="http://purl.org/syndication/thread/1.0">4</thr:total>
  </item>


[zip]

This is Atom feed which Serviio can parse. However the links don't point to the mp3 files (which is normally the case with audio podcasts), but to a blogger web page. I assume the mp3 link on that page can then be used in a different way for every episode so a plugin would be quite unreliable.

[jhb50]

I don't understand. The blogger page contains the links as I've posted. AFAIC this is no different than the YouTube feeds which point to a page containing the links, and only the .mp4 should be displayed.

[zip]

The thing with YouTube and most of other proper video sites it - they have a lot of hidden 'pages' that properly define where the streams are, the only problem is to get to those pages (XML files mostly), we do not parse the actual HTML YouTube page at all. In this case it's free text basically, so we have to depend on the uploader to always specify the link the same way - or alternatively just search for any .mp3 on the page and use the first found.

[jhb50]

You have me totally confused. The feed points to an xml page with a set of <item> tags containing <title> tags and instead of the normal rss content tags which serviio can parse without a groovy it contains <description>&lt;a href="http://funkwarehouse.com/jcpods/john_cleese_podcast_29.mp4" links which a groovy can access.

I thought the whole idea of groovys was to address these non standard video links.

[zip]

Serviio doesn't use <description> element, it only uses <link> elements. Therefore the groovy plugin has to open the HTML page included in the link element, then search the source for a mp3 file. It can be done, but as I said - if the uploader changes they way they embed the mp3 files it'll stop working - unlike proper media websites that publish content using a set of rules the plugin can depend on.

[jhb50]

Here is how I interpreted your development guide:

A plugin is need if the Rss/Atom feed XML doesn't include content URL (down to the media file level), but only URL to web resource
that describes/advertises/plays the content.
Below is an example of a RSS feed that doesn't require a plugin. Serviio can find the content URL in the media:content or enclosure or other elements..

Seemed to me that http://www.johncleesepodcast.co.uk/cleeseblog/rss.xml
is an Rss/Atom feed XML that does include content URL (down to the media file level) and as such Serviio should be able to find the content URL in the other element...

  Code:

<description><a href="http://funkwarehouse.com/jcpods/john_cleese_podcast_30.mp4">

within each item in the absence of media:content or enclosure tags.

I was not suggesting scraping the link html since the rss xml contains all the links required to access these podcasts, but if the rss xml parsing is going to stop after looking for media:content and enclosure tags and revert to the link tag then a groovy is the only way to go back and access the rss xml and return those media tags.

Hence my request for the groovy, but it would be better if the rss xml parsing for media content was complete.

[zip]

The description element is a free text and can include hundreds of URLs for what it's worth. So that's not the way forward.

[jhb50]

OK, that makes sense. So for this feed, since description is the place with the media content urls, I need a groovy to access them, right?

So do I need to change my original request?

[zip]

yes you need groovy, but as I said the plugin might fail from time to time as serviio doesn't use the description element - you'd have to parse the linked html page to find a mp3 file.

[jhb50]

OK. That's probably true of any groovy ie: the youtube and cnn fixes already. That's the reality of software and on-line sources.

I'm going to take a crack at this groovy myself. Is there anyway to test a groovy on my pc without installing it in serviio? ie: run it and see the output of the feed and thumbnail and verify the accuracy of the regex's?

[zip]

they have a console, you just have to add the serviio libraries (.jar) files to the classpath. If you need more help create a topic in the plugin dev forum.