Ankit,
by chance I've been having performance problems in a similar area. I've got
something that takes 120 seconds to complete, and it is parsing an xml file
with lots of entities and nesting.
Your speedup for this entity processing is 0.8/0.02 or forty times faster,
which would cut my time down to 3 seconds. Of course I'm being highly
optimistic here.
I'd be happy to try out Xerces with that patch and let you know what it did
for my situation.
What is the Xerces JIRA issue key?
----------
Brian Minchau
XSLT Development, IBM Toronto
e-mail: [EMAIL PROTECTED]
Ankit
Pasricha/Toronto/
[EMAIL PROTECTED] To
[EMAIL PROTECTED]
10/20/2004 01:58 cc
PM
Subject
performance numbers for the entity
Please respond to perf patch
xerces-j-dev
Here are some informal numbers from xerces after the performance patch
submitted earlier is applied.
The comparison was done taking Xerces 2.6.2 binary as the base.
The XML file used contained multiple entity references with nesting depth
of upto 3 levels.
XMLEntityManager$ScannedEntity creation:
Xerces 2.6.2: 0.8 sec for 674 invocations (avg)
New : 0.02 sec for 674 invocations (avg)
This translates to a 7-9% improvement in the overall speed of parsing.
This, of course, is highly
dependent on the content of the input.
Ankit Pasricha
XML Parser Development
IBM Toronto Lab
8200 Warden Avenue, Ontario L6G 1C7
Phone: (905) 413 4941
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
