• To: Paul Prescod <paul@p...>
• Subject: Re: Revised Internet-Draft: Media Feature - xmlns
• From: "Simon St.Laurent" <simonstl@s...>
• Date: 25 Jan 2002 00:46:53 -0500
• Cc: xml-dev@l...
• In-reply-to: <3C50D752.84974050@p...>
• References: <200201250244.VAA27345@m...> <3C50D752.84974050@p...>

On Thu, 2002-01-24 at 22:56, Paul Prescod wrote:
> How can you process XML from anywhere other than the beginning? Not only
> would you have missed entity references and potential namespace
> declarations, you wouldn't know whether you were in the middle of a
> CDATA section.

It's plausible - though not especially bright - in SAX processing. 
Using the NamespaceSupport class can avoid most of the namespace mayhem,
but not everyone knows to do that or does so.

Grabbing nodes out of the middle of a DOM tree could have similar
consequences, but at least the tree is there for inspection.

I can imagine scenarios where an application would grab text out of the
middle of an XML document based on some indexing scheme, but I'd hope it
was unlikely.

-- 
Simon St.Laurent
Ring around the content, a pocket full of brackets
Errors, errors, all fall down!
http://simonstl.com

• References:
  • Re: Revised Internet-Draft: Media Feature - xmlns
    • From: Mark Baker <distobj@a...>
  • Re: Revised Internet-Draft: Media Feature - xmlns
    • From: Paul Prescod <paul@p...>

• Prev by Date: Re: Revised Internet-Draft: Media Feature - xmlns
• Next by Date: Re: Instance-centric Semantics (WAS RE: Being "precise" vs being "human")
• Previous by thread: Re: Revised Internet-Draft: Media Feature - xmlns
• Next by thread: Re: Revised Internet-Draft: Media Feature - xmlns
• Index(es):
  • Date
  • Thread