RE: Copying most of an XML document

The classic design pattern for this a stylesheet that contains

(a) an identity template

<xsl:template match="*">
<xsl:copy><xsl:copy-of select="@*"/><xsl:apply-templates/></xsl:copy>
</xsl:template>

(b) an extra template for the elements you want to remove or modify

<xsl:template match="caterpillar"/>

Michael Kay 
http://www.saxonica.com/

> -----Original Message-----
> From: Trevor Nicholls [mailto:trevor@xxxxxxxxxxxxxxxxxx] 
> Sent: 23 August 2005 02:20
> To: xsl-list@xxxxxxxxxxxxxxxxxxxxxx
> Subject:  Copying most of an XML document
> 
> Hello
> 
> Given an original XML document that I need to process in 
> several different
> ways I am finding the job much more difficult than it needs 
> to be, I think,
> because the file contains a lot of "noise". I want to use XSL 
> to clean the
> file up in a first pass so that it will be simpler to process 
> later on.
> 
> Here is my magnolia tree - I need it copied exactly but I 
> want to remove
> every caterpillar. Unfortunately it is infested with caterpillars
> throughout.
> 
> Basically I want to do something like copy every single node, 
> children and
> attributes and all, but whenever I find (e.g.) an <a> node I 
> want to ignore
> it. If I could somehow tell 'xsl:copy-of' what to ignore that 
> would be all I
> needed (but I can't). If I knew the full structure of the file I could
> create templates to match and replicate all the nodes I wanted to keep
> without losing any of their attributes, but sadly this XML has been
> generated by Framemaker (from unstructured original 
> documents) and there is
> no document definition to work from.
> 
> I'm hoping there is an XSL solution to this.
> 
> Cheers
> Trevor