Question about Xerces performance

[Brad O'Hearne <ca...@megapathdsl.net>]

Hey guys,

I wanted to ask a question about something I have heard mentioned in design
discussions and perf-related discussion, and that pertains to the claim that
"Xerces is optimized for larger documents", and performs relatively slower
on smaller documents.

I ran some informal benchmarks on a Xerces non-validating SAX parser and
found this to be the case, but as far as initialization and startup code,
isn't this done when the parser is created, not when the parse() method is
invoked?  I am curious as to what is being meant when we are saying that it
is "optimized for larger documents".  If the only thing we are referring to
here is startup code (like initial creation of char set arrays, etc.) unless
this code is massively intrusive and causes some significant wait, then
isn't this optimization claim a bit of a misnomer?  It isn't so much that
Xerces is optimized for larger docs, just that the initial startup costs are
relatively small compared with the processing time of the file.

Is this all there is to it, or is there something being done specifically
for the purposes of larger doc performance which really is of little to no
benefit for smaller docs?

Thanks in advance,

BradO


---------------------------------------------------------------------
To unsubscribe, e-mail: xerces-j-dev-unsubscribe@xml.apache.org
For additional commands, e-mail: xerces-j-dev-help@xml.apache.org