Many services now use RSS feeds to distribute frequently updated information like news stories and status reports. ZXTM's powerful TrafficScript language lets you process RSS XML data, and this article describes how you can embed several RSS feeds into a web document.

It illustrates ZXTM's response rewriting capabilities, XML processing and its ability to query several external datasources while processing a web request.

In this example, we'll show how you can embed special RSS tags within a static web document. ZXTM will intercept these tags in the document and replace them with the appropriate RSS feed data:

<!RSS http://www.zeus.com/news/rss.xml !>

We'll use a TrafficScript rule to process web responses, seek out the RSS tag and retrieve, format and insert the appropriate RSS data.

Check the response

First, the TrafficScript rule needs to obtain the response data, and verify that the response is a simple HTML document. We don't want to process images or other document types!

# Check the response type
$contentType = http.getResponseHeader( "Content-Type" );
if( ! string.startsWith( $contentType, "text/html" ) ) break;

# Get the response data
$body = http.getResponseBody();

Find the embedded RSS tags

Next, we can use a regular expression to search through the response data and find any RSS tags in it:

(.*?)<!RSS\s+(.*?)\s+!>(.*)"

ZXTM supports Perl compatible regular expressions (regexs). This regex will find the first RSS tag in the document, and will extract:

1. the text before the tag
2. the RSS URL within the tag
3. the text after the tag

It will assign these three strings to the variables $1, $2 and $3:

The following code searches for RSS tags:

while( string.regexmatch( $body, "(.*?)<!RSS\\s+(.*?)\\s*!>(.*)" )) {
   $start = $1;
   $url = $2;
   $end = $3;

Note that you need to escape '\' characters in TrafficScript strings.

Retrieve the RSS data

An asynchronous HTTP request is sufficient to retrieve the RSS XML data:

   $rss = http.request.get( $url );

Transform the RSS data using an XSLT transform

The following XSLT transform can be used to extract the first 4 RSS items and format them up as an HTML <UL> list:

<?xml version="1.0"?>
<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
   <xsl:template match="/">
    <ul>
      <xsl:apply-templates select="//item[position()&lt;5]"/>
    </ul>
   </xsl:template>
   
   <xsl:template match="item">
      <xsl:param name="URL" select="link/text()"/>
      <xsl:param name="TITLE" select="title/text()"/>
      <li><a href="{$URL}"><xsl:value-of select="$TITLE"/></a></li>
   </xsl:template>
</xsl:stylesheet>

You can apply the XSLT transform to the XML data using the xml.xslt.transform function. The function returns the result with HTML entity encoding; use string.htmldecode to remove these:

   $xsl = resource.get( "rss.xslt" );
   $html= string.htmldecode( xml.xslt.transform( $rss, $xsl ) );

Insert the new data into the response body

We'll assemble the response body piece by piece, but you could just as easily use string.regexsub to insert the $html in place of the RSS tag that was identified:

   $new = $new . $start . $html;
   $body = $end;
}

Send the new response body

Finally, send the new response in place of the original one:

http.setResponseBody( $new . $body );

The entire rule

The entire response rule, with a little additional error checking, looks like this:

$contentType = http.getResponseHeader( "Content-Type" );
if( ! string.startsWith( $contentType, "text/html" ) ) break;

$body = http.getResponseBody();

$new = "";
$changed = 0;
while( string.regexmatch( $body, "(.*?)<!RSS\\s+(.*?)\\s*!>(.*)" )) {
   $start = $1;
   $url = $2;
   $end = $3;
   
   $html = "<ul><li><b>RSS: ".$url."</b></li></ul>";

   $rss = http.request.get( $url );
   if( $1 != 200 ) {
      $html = "<ul><li><b>Failed to retreive RSS feed</b></li></ul>";
   } else {
      $xsl = resource.get( "rss.xslt" );
      $html = string.htmldecode( xml.xslt.transform( $rss, $xsl ) );
      if( $html == -1 ) {
          $html = "<ul><li><b>Failed to parse RSS feed</b></li></ul>";
      }
   }

   $new = $new . $start . $html;
   $body = $end;
   $changed = 1;
}

if( $changed )
  http.setresponsebody( $new . $body );

Store the XSLT file in the ZXTM conf/extra directory, naming it 'rss.xslt', so that the rule can look it up using resource.get().

Owen Garrett [Zeus Dev Team] 12 January 2006