Walking the DOM (was: XML APIs)

[Paul Prescod]

Sean Mc Grath wrote:
> 
> In my experience, there is a strong "principle of locality" in
> XML/SGML processing. I find I can get by quite happily with
> mini-tree structures harvested at suitable points from
> a larger document processed event-style. In my
> Python toolkit for SGML/XML processing I added support
> for sparse tree building some time ago and I find myself
> using it more and more.

As long as you "get by", more power to you. But what happens when you hit
a document where the first paragraph makes a cross reference to the last
paragraph and the last paragraph makes a reference to somewhere in the
middle? You can hack around it (after all, some people get away with using
Omnimark!), but you will be hacking. You could also hack around local tree
access. Given the choice of hacking around one or the other, I would
rather hack around local references, because those are more predictable.

> This is certainly far easier to do that implement virtual
> tree access with swapping to disk etc. You don't need
> no object database either:-)

That's true, but it doesn't scale to the full generality of problems. As
long as you can get away with it, do so, but I know that I have problems
that require the Full Monty.

 Paul Prescod  - http://itrc.uwaterloo.ca/~papresco

"I always wanted to be somebody, but I should have been more
specific." --Lily Tomlin

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