[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: #cdata?

• From: "Eve L. Maler" <elm@a...>
• To: xml-dev@i...
• Date: Thu, 07 May 1998 12:49:33 -0400

At 04:01 PM 5/6/98 -0400, Lars Marius Garshol wrote:
>(I post this here since it's XML-relevant and I know at least some of
>the XPointer/XLink WG members read this list.)
>
>I'm currently writing a toy XPointer implementation in Python (which
>will quite possibly become serious later) and have to say I'm impressed
>with the XPointer specification. Very clear descriptions and a nice
>syntax.

Thank you. :-)  I hope you'll share future news of your implementation as
you go along.

>However, I'm rather surprised that the 19980303 WD allows pointers to
>distinguish nodes of type #cdata, that is, CDATA marked sections, so 
>that 
>
>root().child(3,#cdata)
>
>points to the third CDATA section that's a direct child of the document
>element.
>
>The reason I'm surprised that this is in the WD is that to support this
>the XPointer implementation will have to ride on top of a parser that
>reports CDATA sections as a distinct syntactic construct, instead of
>classing it as text, which is the usual solution. SAX does not do this,
>nor does the DOM level 1 (WD 19980416) AFAICS.
>
>I also have problems seeing the utility of this feature.
>
>Can anyone comment on on whether it will be an XPointers feature in
>future WDs and if so why?

It was decided to add this back in December.  However, several WG members
have since had a change of heart.  So I have a feeling that it may not
survive.

My own belief is that having the "string" location term is sufficient, and
sufficiently robust, to do any addressing you'd want to do into a CDATA
section.

If we do want to allow addressing into some construct that the DOM doesn't
support, it's incumbent on the XLink side to request new functionality from
the DOM folks.  In other words, we're not artificially going to restrict
ourselves to what the DOM has today.  This approach was actually first
suggested by Lauren Wood (the DOM chair) to ensure that the dependencies
flow in the right direction.

	Eve

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev@i...
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/
To (un)subscribe, mailto:majordomo@i... the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo@i... the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa@i...)

• Follow-Ups:
  • Re: #cdata?
    • From: Lars Marius Garshol <larsga@s...>

• Prev by Date: Re: ANY
• Next by Date: Re: parser for xml-data?
• Previous by thread: #cdata?
• Next by thread: Re: #cdata?
• Index(es):
  • Date
  • Thread