how to process xml without any namespace

rjonk · July 31st, 2009, 09:04 AM

I have two xml files. One has the correct namespace and one xml doesn't have a namespace at all. My xslt expect a namespace on the xml. What is the best way to process the different xml documents.
I think I have to detect the namespace of the document and when it doesn't have the right namespace i have to add a namespace to all the xml nodes and then process the new copied nodes again.. Or is there a better way? If there was a possibility to have one prefix for a specific namespace and an empty namespace...
Can anybody help me what would be the easiest way.

I dont want to change all my rules like:
<xsl:template match="art:myNodeName | myNodeName">

Martin Honnen · July 31st, 2009, 09:09 AM

With XSLT 2.0 you could use
<xsl:template match="*:foo">
to match elements named 'foo' in any namespace (including no namespace). I don't think there is anything shorter.

samjudson · July 31st, 2009, 09:10 AM

Unfortunately the same 'named' element in a different namespace is a completely different element, so there is no easy way to match two same named nodes other than the way you illustrate.

It would certainly be very easy to write a stylesheet that copied the entire XML document from one namespace to another (i.e. from the 'empty' namespace to your specific one) and pipeline that into your workflow.

If you are using XSLT 2.0 this could be done in one XSLT document, but with XSLT 1.0 I don't think it is possible.

mhkay · July 31st, 2009, 09:45 AM

My personal preference in this situation is to preprocess the un-namespaced documents to add the namespace, then use the same transformation on both. This is an example of the general approach called "pipeline processing", and its major benefits are that it maximizes reuse and minimizes complexity. Your stylesheet that adds the namespace is a simple module that can be used in many different pipelines; your main transformation is simpler because it isn't trying to do two jobs at once.