Geeklog Forums

I am a webmaster for an aerospace company and enjoy mixing and matching all manner of portal tools and technologies.

I recently brought Geeklog to my workplace to solve problems for our department websites, and found that when I tried to tie in RSS/RDF news feeds to my local portal, that a connection error occured. Try as I might, nothing worked until I hit on a simple solution.

This seemed odd to me, since I had could reach the news feed URL via any browser, so I researched the often helpful discussions here on the Geeklog.net site and found that many others have experienced the same problem.

I saw the references to the "allow_url_fopen = On" in .htaccess or the php.ini control file, of which my particular site had already be configured to match.

Still... no luck in getting the feeds to work correctly, until I hit on the idea of using my trusty ColdFusion scheduler and a simple script to solve the problem.

Essentially, I created a cache of feeds with the use of the Coldfusion scheduler and the use of a direct call to the news feed url, which then was captured and stored to a local URL on my own server.

Be aware that this is only one potential way to solve the problem. I used Coldfusion since it was already on my development server, but anyone could easily replicate the technique in other languages and to use a scheduler like cron or some similar tool to automate the process.

In Coldfusion, the adminsitrator can submit jobs to the scheduler on a recurring basis. My local site plugs into NASA and CNN based news feeds that my audience finds interesting, so I have one scheduler task for each call. It would be very easy to alter the code to use a loop with a call to a web service or similar to reduce the number of items in the scheduler.

Our RSS caching script uses the following w URL structure:

<a href="http://somecfserver.com/feeds/xmlfeed.cfm?urlAddress=http://rss.cnn.com/rss/cnn_tech.rss">http://somecfserver.com/feeds/xmlfeed.cfm?urlAddress=http://rss.cnn.com/rss/cnn_tech.rss</a>

The important element is to use the cgi argument "urlAddress" to get feed of your choice to record to the local story cache.

The script is represented below:
------------------------------------------------
<!--- RSS news feed utility for automatic feed collection --->

<!--- Set the URL address - Default to the Macromedia news feed --->
<cfparam name="urlAddress" default="<a href="http://www.macromedia.com/desdev/resources/macromedia_resources.xml"&gt">http://www.macromedia.com/desdev/resources/macromedia_resources.xml"></a>;

<!--- Grab the last name in the submitted list --->
<cfset filename = ListLast(urlAddress,"/&quot>

<!--- Go to the remote URL and capture it's content --->
<cfhttp url="#urladdress#" method="GET" proxyserver="yourproxyaddress.com" proxyport="8080"></cfhttp>
<!--- Write the XML file to cache for use with internal Geeklogs --->
<cffile action="write" file="c:inetpubwwwrootfeeds#filename#" output="#CFHTTP.FileContent#">

------------------------------------------------

Remember, this is a "down and dirty" solution... It is not meant to be elegant. In my case, I did not feel like a big investigation on how to get our network operations to change some rules or to investigate a problem I could handle on my own... As far as the code design goes, I decided to keep the initial process simplified until I had some more experience with it and could write a more elegant and portable solution.

I hope this helps some of you. If you wish to contact me, drop an email to me and I'll be glad to get in touch with you.

Jon F. Almada

---

Jon F. Almada