eRDF List Mode? [code examples]
Tom Morris
24 April 2007

<ul class="triplestore">
	<li class="triple"><a href="http://www.google.com" 
class="subject">Google</a>
		<ul class="annotations">
			<li class="dc-title">Google</li>
			<li class="dc-lang">en</li>
		</ul>
	</li>
</ul>

Becomes:
<http://www.google.com> dc:title "Google";
	dc:lang "en".

Alternatively, we could use 'subject' to declare within a particular 
node:
<p><a href="http://www.google.com" class="subject">Google</a>
<br /><span class="dc-title">Google</a><span 
class="dc-lang">en</span></p>

That would work, as would declarations lower in the node tree, but not 
higher.
Basically, we would use an XPath like this:
*[contains(concat(' ', @class, ' '), ' subject ')]/../*
to grab all the data and then declare triples as being as follows:
<xsl:element name="rdf:Description">
<xsl:attribute name="rdf:about">
<xsl:value-of select="*[contains(concat(' ', @class, ' '), ' subject 
')]/@href" />
</xsl:attribute>
[... then declare all the predicates and objects ...]
</xsl:element>
If we are restricting ourselves to HTML, we can rely on @href. I don't 
think we should generalise this to other XML structures, but we can, for 
instance, implement a similar system for those other XML structures 
(OPML comes to mind).

And for styling (not very semantic, I know):
<p class="hidepredicates"><a href="http://www.google.com" 
class="subject">Google</a>
<br /><span class="dc-title predicates">Google</a><span 
class="dc-lang">en</span></p>

.hidepredicates > .predicates {
	display: none;
}

---
Things to consider:
* styling
* validation
* Don't Repeat Yourself
* how to fit other formats in within this - microformats etc.
* how to implement this so as to maximise compatibility with eRDF

tom [at] opiumfield [dot] com