RE: Proposed change of transformation input

> From: Ian Davis [mailto:Ian.Davis@talis.com] 
> I don't understand why this allows "the GRDDL transformation to 
> control *all* XML document processing"
> 
> Can you explain in detail?

Yes.  In essence, by receiving the original representation (i.e., the
actual character sequence, received from the information resource,
representing an XML document) the transformation could parse the
document however it needs to do so, perhaps creating an infoset, perhaps
not.  If the transformation author does not care about having some
ambiguity in the results, or if the transformation author knows that for
this kind of document the potential ambiguity will not be a problem,
then the transformation author can just use XSLT (which is ambiguous
about the parsing/preprocessing) and the result would be the same as if
the transformation had instead received the XPath node tree.  (In other
words, the ambiguity would merely have shifted into the transformation.)

But the point is that if the transformation author *does* care about the
ambiguity, this change is crucial because it permits the transformation
to be unambiguous if the transformation author chooses to write it in a
language -- such as XProc or perl -- that allows the pre-processing to
be fully controlled.  If this change were not made, then even if the
GRDDL transformation were written in XProc or perl (for example), if the
wrong pre-processing were *already* applied by the GRDDL-aware agent
before passing control to the transformation then the transformation
would be unable to do anything about it.

I hope this answers your question.  Let me know if you want more detail
or an example.

Thanks

David Booth, Ph.D.
HP Software
+1 617 629 8881 office  |  dbooth@hp.com
http://www.hp.com/go/software

Opinions expressed herein are those of the author and do not represent
the official views of HP unless explicitly stated otherwise.

Received on Tuesday, 19 June 2007 02:08:06 UTC