Skip to Content.
Sympa Menu

xom-interest - Re: [XOM-interest] XPath Namespace Nodes rethink

xom-interest AT lists.ibiblio.org

Subject: XOM API for Processing XML with Java

List archive

Chronological Thread  
  • From: Elliotte Harold <elharo AT metalab.unc.edu>
  • To: Steve Loughran <steve.loughran AT gmail.com>
  • Cc: xom-interest <xom-interest AT lists.ibiblio.org>
  • Subject: Re: [XOM-interest] XPath Namespace Nodes rethink
  • Date: Mon, 31 Jan 2005 16:35:01 -0500

Steve Loughran wrote:

I hear bad things about the performance impact of integrating
namespace support into things, things like DOM and XPath. Has anyone
some quantified numbers or should I just make them up?

Make them up of course. :-)

My concern was not with speed. Memory was an issue, and a quantifiable one, but it was probably solvable. The nasty problems were conceptual. For instance, if we model namespaces in scope, what happens when we disconnect an element from its parent? Does it retain inherited namespaces or not? What happens when you insert it in a new tree?

We addressed a lot of similar issues in handling xml:base, and that was extremely painful. We did the best we could, but I'm still not sure any answer we came up with was always correct. Namespaces have the same problem times 10. XPath-style namespaces have the same problem times a hundred.

The whole namespace declaration attribute syntax is nice and unambiguous as long as you're only talking about complete, well-formed XML documents that never change. It completely blows up as soon as you begin slicing and dicing documents and removing nodes from and inserting nodes into a document. Static documents are fine. Dynamic ones just aren't. :-(

--
Elliotte Rusty Harold elharo AT metalab.unc.edu
XML in a Nutshell 3rd Edition Just Published!
http://www.cafeconleche.org/books/xian3/
http://www.amazon.com/exec/obidos/ISBN=0596007647/cafeaulaitA/ref=nosim




Archive powered by MHonArc 2.6.24.

Top of Page