From: itsme213 <itsme213@...>
Date: 2005-10-10T03:26:51+09:00
Subject: Re: REXML Question

If you expect to be doing a variety of transformations, I would recommend
not doing mutations on the source xml tree, but instead creating a
transformed copy of that tree. Straight Ruby works fine for that (much
better than xslt, imho).


"Phlip" <phlipcpp@yahoo.com> wrote in message
news:sv4Ze.1011$rx5.601@newssvr22.news.prodigy.net...
> mike leonard wrote:
>
> > I'm looking at REXML as a possible alternative to some things I've been
> > doing with XSLT, but I haven't the foggiest notion of how to accomplish
> > even the simplest transformations. Using REXML, how would I go about,
> > say, translating all <em>'s into <emph>'s, or all <chapter>'s into <div
> > type="chapter">'s?
> >
> > Thank you kindly in advance,
>
> REXML is a DOM-style parser, not a transformer.
>
> To do what you said, off the top of my head I would ...
>
>   doc = Document.new(File.read('some.xml'))
>   XPath.each(doc, '//em') do
>     |node|
>     node.name = 'emph'  #  or similar; check the members first
>     end
>  doc.write(File.open(''another.xml')) # or similar to write
>
> There; a transformation in a bottle. Because XPath.each() returns the
actual
> DOM nodes inside the parsed tree, and because writing on these nodes
changes
> that tree, you can loop thru the nodes and manipulate them using clever
> XPath expressions. The system will accomodate most of the XML rules for
you.
>
> -- 
>   Phlip
>   http://www.greencheese.org/ZeekLand  <-- NOT a blog!!!
>
>