xml-dev - Re: Procedural vs Declarative XML transformation approaches

[Sean McGrath <sean@digitome.com>]

At 09:32 AM 11/6/00 +0000, Matt Sergeant wrote:
>On Mon, 6 Nov 2000, Sean McGrath wrote:
>
> > Using the Pyxie Python library the code would look like this:
> >
> > for f in T.Elements("foo"):
> >          T.Seek(f)
> >          print "\nfoo = "
> >          for b in T.Children("bar"):
> >                  T.Seek("bar")
> >                  print T.JoinData()
> >
> >
> > What is difficult about that? (Or have I misunderstood you?)
>
>You've misunderstood me.
>
>Now try using that same sort of procedural method to process a document,
>rather than data, with a mixed content DTD such as
>
>   <!ELEMENT p (#PCDATA|b|i)>
>   <!-- rest of DTD snipped for brevity -->
>
>For that sort of model you need an iterator, and to declare what you want
>to do with <p>, <b> and <i> tag, rather than trying to find the children
>directly.

Sorry, I still don't get it. The JoinData() method of Pyxie concatenates
the data nodes of the descendants. It works fine with mixed
content models. Maybe if you could post a use case we can
make faster progress on this.

regards,

Sean

Sean McGrath
CTO
Propylon - Enabling Universal Mobility
http://www.propylon.com
Tel: +353 1 6620482