- From: Brian Suda <brian.suda@gmail.com>
- Date: Wed, 18 Jul 2007 12:49:53 +0000
- To: public-rdf-in-xhtml-tf@w3.org
I had a search through the archives and the RDFa syntax document, but i couldnīt find the answer, so hopefully it can easily be addressed here. In the microformats community there is a discussion about how and where to extract data when there are child elements inside the element that has not been explicitly marked-up as microformated connet. A similar issue appears with RDFa as well, so i went to see how you deal with it, but i couldnīt find anything. Here is an example: <ul class="vcard"> <li class="fn"><img src="b.png" alt="B" />rian</li> </ul> Right now, (IMHO) the microformats spec should only extract "rian" as the value for FN, others think it should look at the @alt on the image as well and produce "Brian" (microformats already have a way to solve this problem, so my current feeling is that this is an un-needed default behaviour) I had a look at the RDFa syntax document[1] to see if there was an opinion and reason why or why not the IMG element should be parsed. And the best i could find was the examples with Literals, which isnīt what we want <head about=""> <meta property="dc:title" content="E = mc<sup>2</sup>: The Most Urgent Problem of Our Time" /> </head> and the BOLD <span about="http://example.org/foo" property="dc:creator" datatype="xsd:string"> <b>M</b>ark <b>B</b>irbeck </span>. where the <b> elements are dropped altogether, but iīm not sure how that would translate to an IMG with @alt I think the problem with "peeking" inside child elements, becomes, where does it end? ABBR, DFN, OBJECT, etc. Then things get VERY harry! Does RDFa have a stance on this? if so, then microformats should attempt to come into line as well or atleast agree in some form how to parse instances and edge cases like this. Thanks, -brian [1] - http://www.w3.org/2006/07/SWD/RDFa/syntax/ -- brian suda http://suda.co.uk
Received on Wednesday, 18 July 2007 13:07:11 UTC