- From: Matthew Smith <matt@kbc.net.au>
- Date: Fri, 27 Jun 2003 08:30:20 +0930
- To: WAI Interest Group <w3c-wai-ig@w3.org>
David Woolley wrote: >>To answer this one from my angle, the ability to apply XSLT translations >>to XHTML is an advantage, and a big one. > > What features of HTML parse trees make them incompatible with XSLT? > It's not so much a question of "can't work" as "shouldn't work". My understanding is that at the heart of a programme which applies the XSLT translation, there is an XML parser. With HTML not being well-formed, the parser should not, if complying with specification, be able to read the HTML file. Even the best-designed, valid, HTML should not parse because, basically, it is not XML and that is what the parser is expecting. That is not to say that one could not write a translation utility based on an SGML/HTML parser. One would have to be very careful that HTML fed to this is valid, otherwise unpredictable results would occur - this is the very reason that XML parsers are designed to stop as soon as they see something that is not well formed. Cheers M -- Matthew Smith Kadina Business Consultancy South Australia http://www.kbc.net.au
Received on Thursday, 26 June 2003 19:00:23 UTC