RFC: "even simpler" C++ XML parser for object hierarchies

[Paul Miller]

> Is holding the document in memory not an option?  Because if you can hold
> it all in memory, it's simple enough to build a tree representation of the
> thing, then walk the tree with your handlers, then throw away the tree.

I thought about that at first, and I could do it easily over expat (or
SAX). In fact, I could even keep the same API I came up with, because
then I'd just be dealing with a DOM-like representation.

However, for "large" data-files (several megabytes worth), this could be
a problem, because the in-memory representation could be many megabytes
larger (consider a 3D model with 10,000 vertices, each one with a
<Vertex></Vertex> pair). I suppose I could minimize the amount of extra
memory usage by using hash tables, but I tend to prefer to streaming
solution.

I was going under the assumption that for this type of use, namespaces
and validation probably aren't necessary, so there aren't that many
advantages to layering over expat.

> If that isn't an option, then I agree that you have a combination of:
> 1) A nice interface paradigm (you could even generate a stub of the program
> from a DTD!); and,
> 2) Quite a challenge getting an event-based parser to work with it because
> of control-flow issues.

Thanks, and yeah, it would be/is a real effort to match this idiom to
expat.

> Here's a nutty idea -- Try threads.  Run the parser and expat as two

Ouch! :-)  You get an 'A' for creativity!

-Paul

--
Paul Miller - stele at fxtech.com

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/ and on CD-ROM/ISBN 981-02-3594-1
To unsubscribe, mailto:majordomo at ic.ac.uk the following message;
unsubscribe 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)