My fifteen minutes of fame (or how to get OPML into HTML)


A few days back as I was skimming through my plethora of rss feeds on Bloglines I came across a post from Tod Maffin looking for a programmer to write a simple script which would parse an OPML file of podcasts into HTML which could be included in a web page. The task was straightforward enough and given the ubiquity of OPML in RSS readers, I knew someone else would have at least written a basic parser in PHP that I could start with. I wrote back to Tod telling him that I was the man for the job and set about working on the script. I offered to do the script pro-bono as a way to help spread the word about public-radio podcasts which keep me entertained and amused throughout the day.

A quick google search brought me to Joe Grossberg’s PHP script for including his OPML Bloglines subscription list as an HTML un-ordered list on his blog. His code was very straightforward and well writen so I decided to use it as a base.

My first task was to strip out all the code that was particular to Bloglines OPML format and customize the script to the format of Tod’s OPML file (you can find an example here). This proved to be more difficult than I originally anticipated and I had to consult the PHP manual to understand how the PHP function, xml_parse_into_struct function worked in order to make sure I was pulling out the parsed podcast feed title and URL. However, after a few hours of trial and error, I was able to parse the OPML file correctly into an ordered list. I even added a small bit of code to ignore the first entry in his OPML file which is a link to add new podcasts. (Tod had hardcoded that link into his website and he only provided the link in the OPML file so that it could be used in the Ipodder directory to have podcast listeners contact him about new podcasts).

With the parsing complete, I turned to the task of displaying the podcast information in an HTML table. I took my cue from Tod’s orginial page which he had been manually updating as new podcasts were added. In that layout, podcasts were organized by categories and alternating categories had the same color so that it was easy to distinguish between categories. I wanted my script to generate the table entirely using PHP so I had to reaquaint myself with the process of writing code by hand because my script would have to add the table tags dynamically depending on how many podcasts were in the OPML file. After a few hours of tweaking the code and a night’s worth of sleep I finally had it generate a basic table which I could include in another PHP file.

In order to make the table a little easier to read, I used the trick of alternating colors for table rows that’s become popular around the Net and in applications like Apple’s iTunes. The easiest way I found to do this was to keep track of the number of podcasts in each category. Every time my script would go to write out the table cell definition for a podcast it would check to see if the podcast number was divisible by two meaning that it was even. Even numbered podcasts would be colored using one color and odd numbered podcasts would use a different color. To make things even more interesting, I kept track of the category number as well and wrote code to create two sets of even/odd color pairs - one for even numbered categories and other for odd-numbered categories. You can admire the results at http://todmaffin.com/feeds/ . http://publicradiofeeds.comhttp://web.archive.org/web/20050605025057/http://publicradiofeeds.com/.

With Tod’s permission I’m releasing the source code for my file under GPL so that others might benefit and improve on it. The file is called ipodder.opml.class.php. To use it, you’ll need to add the following code snippet as a table row in another .php file:

Anyways here is the link to the ipodder.opml.html.php.

I hope you find it useful in and of itself or as a point of departure for building your own OPML parser. Please email me with any feedback you might have at projects@rasheqrahman.com.


comments powered by Disqus

Date

Thursday, May 26, 2005

Estimated Reading Time

3 minutes

Categories

Previous Article

Next Article