Re: TPAC topic: Extending XInclude

"Grosso, Paul" <pgrosso@ptc.com> writes:

>> -----Original Message-----
>> From: Norman Walsh [mailto:ndw@nwalsh.com]
>> Sent: Sunday, 2011 October 30 21:18
>> To: public-xml-core-wg@w3.org
>> Subject: Re: TPAC topic: Extending XInclude
>> 
>> "Grosso, Paul" <pgrosso@ptc.com> writes:
>> > Paul said we could say that all attributes on the xinclude element
>> > that are in the xinclude namespace should be copied to the "root"
>> > included node(s).  That one thing would allow a follow up process
>> > such as Norm's XSLT to do the necessary fixup.  If one did xinclude
>> > processing after validation, one could give those extra attributes
>> > defaults in the schema so that they didn't need to be authored.
>> 
>> I still feel like it might be better to say that all attributes in
>> *any* namespace get copied over. That avoids any contention over the
>> local names of the attributes. DocBook can have db:transclusion-policy
>> without walking on anyone else.
>
> Maybe, but then what do you do when the xinclude element
> has an attribute of the same name as the root included
> element.  Say, like id.

Well, "id" wouldn't qualify as in any namespace by the definition
I had in mind. Otherwise we'd be copying over href and parse too.

I guess I'd say all attributes are copied over *except* attributes
that aren't in a namespace, attributes in the XML namespace, and
attributes in the XInclude namespace (there aren't any, but if we added
some I don't suspect we'd want them to be copied).

[ Probably my last transmission for a while, departing CLE shortly. ]

                                        Be seeing you,
                                          norm

-- 
Norman Walsh
Lead Engineer
MarkLogic Corporation
Phone: +1 413 624 6676
www.marklogic.com