XOM API for Processing XML with Java

[Elliotte Harold <elharo AT metalab.unc.edu>]

New, Cecil (GE Trans) wrote:

> As long as there is no DTD, then a CDATA seems unecessary.

> But if there is a DTD and it does not allow mixed content for <content>,
> then a CDATA section would effectively sidestep validation issues...

Again, you are asking to do exactly what I maintain is a horrible idea. 
Content in CDATA is *not* markup , even if it looks like it is. Placing 
it in a CDATA section has not fixed your problem. There are two good 
solutions to the problem you encountered, neither of which requires 
hiding markup inside CDATA sections:

1. Change the DTD
2. Put the markup inside the relevant markup, even though the DTD says 
you can't. Validity is an option. It is not required.

Both of those solutions are more robust and maintainable than the very 
ugly hack of hiding unparsed markup in a CDATA section.

--
Elliotte Rusty Harold  elharo AT metalab.unc.edu
XML in a Nutshell 3rd Edition Just Published!
http://www.cafeconleche.org/books/xian3/
http://www.amazon.com/exec/obidos/ISBN=0596007647/cafeaulaitA/ref=nosim