Yes and no. Xerces is damned optimized you can look at the code and you
won't be able to make a much faster parser. However, it does support
features that eat up CPU/memory.
Validation for one. There's a tradeoff between using validation (more
robust applications) and not using it (faster).
DOM is another one. DOM eases application development but eats up memory
like crazy. IF you're concerned about performance stick to SAX.
Namespace handling, and so on.
By the large XML parsing and serialization are on the same scale of
performance. Serialization of objects is faster than Java-XML binding,
but serialization of data can be slower than its XML counterpart.
For those curious what serialization of data is, consider the case where
one application stores a PO from a myapp.PurchaseOrder object and
another reads it into a yourapp.Order object. The two objects are
totally different, thus object serialization cannot work, but the data
is generic enough to be mappable to both objects.
This is where XML excel. And even if it does sound wacky, more and more
companies are using XML internally between different applications that
talk the same data but do not use the same objects.
arkin
James Cook wrote:
>
> Tim, we have been generally disappointed at the speed of parsing an XML file
> of any significance. I like the ideas behind it, but I think most of the
> more widely-used XML parsers suffer from some speed issues.
>
> I have recently become interested in less-robust XML parsers. I am assuming
> that less is more (faster). Of course, that assumption could be horribly
> askew. Specifically, I want to look into:
>
> (NanoXML) http://nanoxml.sourceforge.net/
> (MinimalXML) http://www.docuverse.com/smldev/minxmlspec.html
>
> This wouldn't satisfy your XSLT needs, but as far as shuttling data around,
> I want a lean/mean, small and fast XML parser.
>
> jim
>
> ----- Original Message -----
> From: Tim Endres <[EMAIL PROTECTED]>
>
> > I realize that I am going to incur a performance hit using XML, but I
> > believe that we can afford it.
> >
> > I just ran some tests with Sun's JDK1.1.8, a Pentium II 450 W2K box, and
> > using Apache's Xalan and Xerces packages. Very preliminary tests show that
> > I can parse very simple XML documents to the tune of 1000 per 6 seconds.
> > And for XML/XSLT, I can able to translate the simplest of documents to the
> > tune of 1000 per 12 seconds. The numbers fall within our requirements, but
> > I doubt they will be sufficient for all projects. And I still do not have
> > any numbers of memory consumption.
>
> ===========================================================================
> To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
> of the message "signoff EJB-INTEREST". For general help, send email to
> [EMAIL PROTECTED] and include in the body of the message "help".
--
----------------------------------------------------------------------
Assaf Arkin www.exoffice.com
CTO, Exoffice Technologies, Inc. www.exolab.org
===========================================================================
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff EJB-INTEREST". For general help, send email to
[EMAIL PROTECTED] and include in the body of the message "help".
