Parsing DTD 's with CDATA elements 2003-01-14 - By Joseph Kesselman
This looks like the standard confusion between CDATA, #PCDATA, and
<![CDATA[]]>
<![CDATA[]]> (a "CDATA Section") is actually a kind of #PCDATA content.
It's just a syntactic convenience, as an alternative to escaping
individual characters. It does not have to be specially declared in the
DTD; just say that the element can contain #PCDATA and then use
<![CDATA[]]> in your document, and everything should work properly.
#CDATA is the datatype used for *attributes* which contain unconstrained
text. You can't define an element as containing #CDATA, and you don't
have to do so.
Yes, the fact that CDATA means two different things, and is so similar to
PCDATA, is confusing. All I can say is that it makes sense to a parser
developer. (It doesn't make us _happy_, but it does make sense.)
______________________________________
Joe Kesselman / IBM Research
---------------------------------------------------------------------
To unsubscribe, e-mail: xerces-j-user-unsubscribe@(protected)
For additional commands, e-mail: xerces-j-user-help@(protected)
|
|
