Re: [xml-dev] CDATA headache

[michael odling-smee <mike.odlingsmee@gmail.com>]

All,

Thanks very much for all the helpful responses. I thought I needed to clarify that we are not trying change our application to meet the (arguably dubious) requirements of our customer regarding CDATA. We are just trying to show them that our application is working to specification and hence should not require changes. On this front I think we have the answer in the aforementioned XPath 1.0 specification: http://www.w3.org/TR/xpath#section-Text-Nodes.

Kind regards,

Michael

On Thu, Sep 10, 2009 at 10:03 AM, Mukul Gandhi <gandhi.mukul@gmail.com> wrote:

On Wed, Sep 9, 2009 at 4:54 PM, michael odling-smee

<mike.odlingsmee@gmail.com> wrote:

>  about 5% of the elements in the source
> document are <value/> elements not all of which contain CDATA. If we added
> new CDATA sections to the output this would probably not satisfy the
> customer either.

Just to add another point. If some of input <value /> elements,
contain CDATA markers, then it means to me, that your application
design has thought, that CDATA was necessary on <value /> elements.
Some of <value /> elements in your input XML don't have CDATA markers,
and this I think, is the problem of the process, which created this
input.

But leaving the problem aside, of incompletely specified CDATA markers
in your input, you can fix this problem completely, entirely within
your stylesheet. You can put cdata-section-elements="value" in your
XSLT stylesheet, and this will generate CDATA markers in all the
<value /> elements, in the output. And this is what your application
design thought, in the first place. So this to me, is what should be
done.

As an additional work, and if you want, you can work to improve the
process, which is producing your input XML, so that it put's CDATA
markers in all the <value /> elements.


--
Regards,
Mukul Gandhi