- From: Felix Sasaki <fsasaki@w3.org>
- Date: Thu, 02 Feb 2006 22:48:54 +0900
- To: "Yves Savourel" <yves@opentag.com>, "'Sebastian Rahtz'" <Sebastian.Rahtz@oucs.ox.ac.uk>
- Cc: public-i18n-its@w3.org
On Thu, 02 Feb 2006 22:08:18 +0900, Yves Savourel <yves@opentag.com> wrote: > OK, I'm convinced <ns> is the way to go then :) o.k., if everybody agrees (or nobody disagrees), I will add a small explanation to the draft with an example. - Felix > -ys > > -----Original Message----- > From: Felix Sasaki [mailto:fsasaki@w3.org] > Sent: Thursday, February 02, 2006 5:40 AM > To: Sebastian Rahtz > Cc: Yves Savourel; public-i18n-its@w3.org > Subject: Re: using <ns> to specify namespaces > > On Thu, 02 Feb 2006 18:27:20 +0900, Sebastian Rahtz > <Sebastian.Rahtz@oucs.ox.ac.uk> wrote: > >> >> >> Felix Sasaki wrote: >>> >>> Talking from the XSLT / XQuery perspective: Both Sebastians and my >>> implementations need to copy the string values of the selector >>> attributes into generated XSLT stylesheet / XQuery files. So we need an >>> extra step anyway to generate zero or more xmlns elements in these >>> files, so that the XSLT / XQuery processor understands these values as >>> "namespace using" XPath expressions. >>> This task seems to me easier to do with the <its:ns> element, rather >>> than analyzing xmlns attributes at "its:documentRules". Although the >>> task would be not impossible for XQuery, I think. Sebastian, what do >>> you think about XSLT? >> >> as it stands in XSLT 1.0, I don't see any way to get the necessary >> information across from the (otherwise reasonable) notation Yves >> proposes. Can you see a way, Felix? >> > Ah, sorry, I just thought of XSLT 2.0 ... With XSLT 1.0, you are really > lost, I think. > > Felix >
Received on Thursday, 2 February 2006 13:49:03 UTC