Re: [xsl] Copy results

[Steven Ericsson-Zenith <steven@xxxxxxxxxxxxx>]

On some occasions the result is processed for preview purposes, on
others it is "published" (determined by the state of attributes in the
processed file) Versions that are published are copied as permanent
records for application purposes.

I am processing a tree of multiple collections in the same transform.
Your suggestion then that I simply translate the document again is a
cost multiplied by the number of actual documents processed (which can
be quite large). Do you still want to do it that way?

Generating a published file list and then running a Python script, or
some such, on it is an obvious solution - but I prefer to minimize the
points of failure and have to maintain certain application guarantees.

With respect,
Steven

--
	Dr. Steven Ericsson-Zenith
	Institute for Advanced Science & Engineering
	http://iase.info
	http://senses.info


On Jul 7, 2008, at 11:47 PM, Michael M|ller-Hillebrand wrote:

> Am 08.07.2008 um 01:29 schrieb Steven Ericsson-Zenith:
>
> > Dear list,
> >
> > What is the optimal way to copy a results document? So, what one
> > would like perhaps is:
> >
> > <xsl:result-document
> >               href="file1.xhtml,file2.xhtml">
> >
> > </xsl:result-document>
> >
> > Is the optimal way to do this to assign the page output to a
> > variable and then dump it in two consecutive result-documents?
>
> I would not use a variable, but a named template which is called
> twice. In many situations the XSLT performance is so good, you would
> not notice any delay.
>
> I would say the general way is to avoid redundancies like this. And
> I guess this request has nothing to do with the XML content, but
> rather some obscure requirements down the line. So, do not burden
> your XSLT with such issues. Just use a command line command to
> duplicate the file when needed.
>
> - Michael