W3C home > Mailing lists > Public > public-xml-core-wg@w3.org > October 2011

Re: Extending XInclude

From: Norman Walsh <ndw@nwalsh.com>
Date: Tue, 04 Oct 2011 12:30:18 -0400
To: public-xml-core-wg@w3.org
Message-ID: <m24nzo67ol.fsf@nwalsh.com>
"Grosso, Paul" <pgrosso@ptc.com> writes:
> I understand the bottom line of your experiments to be that,
> if one defines xinclude processing to include copying certain
> attributes on the xinclude element to the included root element,
> it is possible to define id-fixup.
>
> Is that a fair statement?

I think so.

Defining id-fixup along the lines suggested by the DocBook TC required
two things:

1. The ability to find where the XInclusions occurred.
2. The ability to add author-specified annotations to those roots.

On the whole, I think we could justify fixing point 1 in XInclude
without biting off the "id fixup" question ourselves in XInclude.

I'm much less confident that we can justify point 2 without having
the larger use case in mind.

What's more, the actual mechanism I chose, pushing attributes from the
xi:include element down into the included content seems like a total
hack.

> I didn't understand what cx:id="refid1" was doing, and (perhaps
> related) I didn't understand how idref fixup worked (no, I didn't
> study the XSLT to figure this out).

How it works depends on the fixup and linkscope settings. With respect
to cx:id, it's there to handle the case where you want to link to a
specific instance of the XIncluded content. Because in general fixup
is going to massage the ID values, if you don't have some way of
overriding the top-most ID, you can't link to it "from outside."

I'm not actually sure I thought that through all the way.

> Finally, if adding attributes to the xinclude element (and defining
> their processing) is sufficient to allow specification of id-fixup,
> is this something we as a WG might want to write up, perhaps as a
> WG Note?

Perhaps. I think the first question to answer is whether or not we
think my hack for migrating attributes from xi:include down into
content is something we could even come close to justifying.

                                        Be seeing you,
                                          norm

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

Received on Tuesday, 4 October 2011 16:30:50 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 4 October 2011 16:30:51 GMT