Re: more elegant way to process element with default v

> The parsing does look for "element start tags, etc" in a sense (at least, it
> looks for the "<" delimiter, the beginning of a start tag) - it just
> signals an error if they are found because they are illegal there.

Of course the terminology is confusing as CDATA is PCDATA without a P to
indicate that it's not parsed (or at least not parsed as normal)
in SGML you could have a < in CDATA attributes (just as you could
in CDATA elements) but in XML CDATA elements don't exist and allowing <
or not in an attribute value depending on its DTD declaration would have
gone against the whole point of XML that an instance ought to be able to
be parsed as a tree without using a DTD at all, so < is uniformly banned
in all attribute values.

CDATA was especially confusing in SGML as for elements it made both <
and & into normal characters (ie like a CDATA marked section)
but for attributes it made < normal but & is still special so you can
go & # 1234; in CDATA attributes and the & has its special meaning.


David

________________________________________________________________________
This e-mail has been scanned for all viruses by Star Internet. The
service is powered by MessageLabs. For more information on a proactive
anti-virus service working around the clock, around the globe, visit:
http://www.star.net.uk
________________________________________________________________________

 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list