A couple of comments.

First, why are you bringing in CGI? You're not making any use of it at all. I could see using LWP instead of lynx -dump, but use CGI is just making your program bigger for no reason.

Second, you're using dot-star unnecessarily in your regexes. In your if statement, you can omit them all and your results will be the same, but it will simplify your code (and make life easier for the regex engine).

But I do agree that this is a pretty cool idea. jcwren did it for the monastery. Take a look at Stats Whoring - The VXML Way.

#!/usr/bin/perl -w
use strict;

#
# A script to convert slashdot headlines to VXML.
# (C) 2001, Jaldhar Vyas
# Licensed under the Crowley Public License ("Do what thou wilt 
# shall be the whole of the license.")
#

use LWP::Simple qw(get);
use XML::LibXML;
use XML::LibXSLT;

my $content = get('http://slashdot.org/slashdot.rdf');

unless (defined ($content)) # undef means something went wrong.
{
  $content = <<'-EOT-';
<?xml version="1.0"?>
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#">
  <item>
    <title>Failed to retrieve headlines</title>
  </item>
</rdf:RDF>
-EOT-
}

my $xml = XML::LibXML->new();
my $xslt = XML::LibXSLT->new()->parse_stylesheet($xml->parse_fh(*DATA)
+);
print "Content-type: text/xml\n\n" ,
  $xslt->output_string($xslt->transform($xml->parse_string($content)))
+;

__DATA__
<?xml version="1.0"?>

<xsl:stylesheet xmlns=""
    xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="1.0"
    xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
    xmlns:rss="http://my.netscape.com/rdf/simple/0.9/">

  <xsl:namespace-alias stylesheet-prefix="#default" result-prefix='rss
+' />

  <xsl:output method="html" media-type="text/xml" indent="yes" />

  <xsl:template match="rdf:RDF">
    <vxml version='2.0'>
    <form><block>
      <xsl:apply-templates select="rss:item/rss:title" />
    </block></form>
    </vxml>
  </xsl:template>

  <xsl:template match="rss:title">
    <xsl:value-of select="." /><xsl:text>.
    </xsl:text>
  </xsl:template>

</xsl:stylesheet>
[download]

The code in the original post only addresses Slashdot, or at best Slashcode based sites, and the implementation breaks if the Slashcode project authors ever change the way they format data for the Palm OS.

A more leverageable solution would be to use XML::RSS to extract title entities from each item of a site's RSS file and rewrite the data in VoiceXML. This would allow you to apply your resulting script to any site that produces an RSS file. Syndic8.com knows about over 2,600 such sites or subsites at the moment.

Another solution would be to leverage Using XSL Transformations to Produce VoiceXML. That's example 118 on the Tellme Studio web site.

The original solution offered is interesting, but it is only readily understandable by people using UNIX. Although it is possible to run Lynx on PCs and even Macs, nobody I know is doing that. Furthermore, most developers on those platforms do not even know what Lynx is.

Dave Aiello
Chatham Township Data Corporation

But it still has a very high cool factor.

Kudos!

metadoktor

"The doktor is in."