Re: General algorithm for finding nodes between PIs

Good question, but these xml documents come from Adobe FrameMaker, which has
a "conditional text" feature. Since this feature is part of both
unstructured and structured (sgml- or xml-based) FrameMaker, it uses
processing-instructions to mark the bounds of each condition format. The
purpose of the transform is to move the information from the PIs to
attributes.

I ended up doing this inside of FrameMaker itself with its scripting tools.
But thanks to this list, I ended up learning some good techniques for doing
it with XSLT. Thanks to all for a great list.

Rick

-----Original Message-----
From: Flynn, Peter pflynn@xxxxxx
[mailto:xsl-list-service@xxxxxxxxxxxxxxxxxxxxxx] 
Sent: Monday, January 30, 2017 6:41 AM
To: xsl-list@xxxxxxxxxxxxxxxxxxxxxx
Subject: Re:  General algorithm for finding nodes between PIs

On 27/01/17 18:21, Rick Quatro rick@xxxxxxxxxxxxxx wrote:
> Thank you all for the help.

Unrelated to the solution in this case, but from an encoding standpoint, I'm
curious to know if there was any opportunity to choose another schema/dtd
for this document or application; specifically to use one which implements a
different way to encode this kind of effectivity, perhaps using attributes
or elements.

///Peter