Hi,
The memory consumption should not grow if you release the documents.
Cheers,
Gareth
Tomas Telecky wrote:
Hallo,
I want to use the Xerces parser in a multithreaded program running as
DLL. It will run virtualy infinitely. Threads will be started and will
use Xerces. I will call Initialize a
On Tue, 21 Aug 2001 09:54:41 -0400, Mark Northcott wrote:
>I thought that SAX avoided this type of memory consumption since it is
>event driven and simply notifies my custom ContentHandler as each type
>of XML tag is encountered.
It is. Replace your handler with an empty one, and measure mem u
alled PlatformUtils::Terminate()). Any advice on
this topic would be greatly appreciated.
Mark
-Original Message-
From: Tinny Ng [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, August 21, 2001 10:30 AM
To: [EMAIL PROTECTED]
Subject: Re: Memory usage of SAX2XMLReader
This sounds similar to bugzilla bug
This sounds similar to bugzilla bug 1329 which is now fixed in the nightly
build http://xml.apache.org/dist/xerces-c/nightly/.
Tinny
Mark Northcott wrote:
> I am working on a project that is making use of the SAX2XMLReader to
> parse XML documents, and I am noticing that the parser is consuming
Ehmm, Jeremy, you're creating a new XmlErrorReporter() in every iteration;
maybe you should try deleting that?
> -Original Message-
> From: Jeremy Ford [mailto:[EMAIL PROTECTED]]
> Sent: Wednesday, July 18, 2001 16:58 PM
> To: [EMAIL PROTECTED]
> Subject: memory usage/performance for xer
There is a known memory leak in Xerces 1.5. A bug-fix release Xerces 1.5.1 is
going to be released soon (within these two days). You can try our latest
nightly build http://xml.apache.org/dist/xerces-c/nightly/2001-07-17/ which is
the candidate for Xerces 1.5.1 and see if this solves your proble
I believe that I resolved this problem. The DTD that I use has multiple
ATTLIST entries for each ELEMENT. This is a documented leak reported by :
Erik Rydgren
Mandarinen Systems AB, Sweden
Having just one ATTLIST for each attribute per ELEMENT seems to eliminate
the leak.
Sean Radcliffe wro
Like all object oriented systems, this is very difficult. The whole point of
OO is to encapsulate data, and there is often a lot of pointers to
encapsulated data, making it impossible to use any kind of 'sizeof' type of
tricks.
--
Dean Roddey
Software Geek Extraordinaire
Portal, Inc
[
