Subjects
Home
VOTE Move XML Commons to Xerces
Commented: (XERCESJ 589) Bug with pattern restriction on long strings
: Xerces J 2 8 1 Release on Wednesday, September 13th
: Xerces J 2 9 0 Release on Wednesday, November 22nd
Commented: (XERCESJ 1066) Restriction+choice+substitutionGroup error
Commented: (XERCESJ 1178) Error getting prefix for an attribute with no n
Updated: (XERCESJ 1244) XMLSchemaValidator does not contribute element 's
Some consideration about the xerces DOM implementation
Updated: (XERCESJ 1066) Restriction+choice+substitutionGroup error
Commented: (XERCESJ 1227) Poor performance / OutOfMemoryError for sequenc
retain exception stack traces
Updated: (XERCESJ 1193) NPE or hang when parsing using the "continue afte
Future of NekoHTML
Commented: (XERCESJ 1203) NPE in XMLDTDProcessor
DOM Level 3 APIs for Xalan J and a new Xalan release (2 7 1)
: xml commons external 1 3 04 Release on Wednesday, November 22nd
Commented: (XERCESJ 1247) Incorrect location information on SAX when usin
XInclude exceptions how to mirror Xerces J functionality into Xerces C++?
First proposal on SoC project "Add support for the StAX (JSR 173) cursor API
: xml commons resolver 1 2 Release on Wednesday, November 22nd
Typo in RangeToken java Please check
Validator features
java lang ClassCastException when adopting Node
using the org apache xerces impl xs identity package
Updated: (XERCESJ 1257) buffer overflow in UTF8Reader for characters out
Problem with ref attributes and schema validation
Updated: (XERCESJ 122) XMLSchemaValidator does not contribute element 's d
Performance problem under load Xerces with Weblogic 9 x
remove ignored memory allocation
Commented: (XERCESJ 1177) SAXXMLStreamReader doesn 't always report namesp
Commented: (XERCESJ 977) Null pointer exception during DOM parsing
Commented: (XERCESJ 1197) Code cleanup for org apache xml serialize
Commented: (XERCESJ 1201) Initial contribution for StAX Event API
Updated: (XERCESJ 1061) Regex "$ " and "^ " characters treated as special c
Commented: (XERCESJ 1199) SAXXMLStreamReader should attempt to register a
Commented: (XERCESJ 1061) Regex "$ " and "^ " characters treated as special
Updated: (XERCESJ 589) Bug with pattern restriction on long strings
StackOverflow
xerces Range unnecessarily not garbage collectable if not detached
Updated: (XERCESJ 1178) Error getting prefix for an attribute with no nam
Bug in xs:redefine
Commented: (XERCESJ 1204) Can not set XMLEntityResolver for LSParser
Updated: (XERCESJ 1253) Prototype for SoC2007 project "Add support for th
Updated: (XERCESJ 1259) Add SteamFilter Function to SoC2007 project "Add
Assigned: (XERCESJ 444) SAXException thrown by EntityResolver is reported
Google Summer of Code 2007
Xerces J and XInclude relative path issue
Assigned: (XERCESJ 206) Stack overflow when using a schema validation
Commented: (XERCESJ 1215) Restrictions involving two levels of substituti
Closed: (XERCESJ 1203) NPE in XMLDTDProcessor
non overriding equals methoda
Resolved: (XERCESJ 1079) invalid value returned for TOTALDIGITS facet in
Xerces AS3 port
Updated: (XERCESJ 325) Regular Expression; Pattern "| " clause order de
Updated: (XERCESJ 1196) Javadoc generation fails on Java SE 5 0
Closed: (XERCESJ 1202) DTD validation on XIncluded documents when the sch
Created: (XERCESJ 1124) Nonspecific schema error message
a bug in xerces
Updated: (XERCESJ 1201) Initial contribution for StAX Event API
Closed: (XERCESJ 1254) Empty uris in targetNamespace attribute not report
Links
Home
Oracle database error code
 
Search:  
Power your search with and, or, +, -, or "some phrase" operators.
xerces Range unnecessarily not garbage-collectable if not detached...

xerces Range unnecessarily not garbage-collectable if not detached...

2007-03-08       - By Ludger Buenger
Reply:     1     2     3  

Hi Michael, thanks for your quick answer.

Ok, Here I send a first proposal for changes as a patch regarding Ranges and
weak references.
It should not be difficult to apply similar changes to NodeIterators.

A more optimized implementation would likely make use of reference queues
removing the need to actively check the weak reference's content upon each
range update by DocumentImpl.

Since I am not proficient regarding java.io.Serialization at all someone more
competent in this matter has to check whether something like this would break
serializability.


Internally we were also thinking about EventListeners to use WeakReferences
however I think that references to objects provided by an external application
should not be weak (which I think applies to EventListener and user data).

However classes xerces provides to the application where xerces stores an
reference in order to be able to update it's content can be weak and thus
garbage collectable (which I think applies to Ranges and NodeIterators).


The following is a use case where a weak reference to an EventListener could
cause undesired behaviour.

Imagine an EventListener implementation that solely logs all events to System
.out or a file:

node.addEventListener(MutationEventImpl.DOM_SUBTREE_MODIFIED,
           new EventListener() {
               public void handleEvent( Event evt ) {
                   System.out.println( "Subtree Modification Event occured" );
               }
           }, true );


The only reference to this EventListener resides inside the DOM implementation.
If I am not mistaken usage of WeakReferences inside the DOM implementation here
would result in the logging EventListener eventually being garbage-collected
and thus cease event logging, right?

Same applies to user data: It might happen, that an application provides user
data and relies upon the node implementation to hold on to it until the
application retrieves the user data from the node again.


Thanks,

Ludger Bünger
--
Dipl.Inf. Ludger Bünger
RealObjects GmbH

> -----Original Message-----
> From: Michael Glavassevich [mailto:mrglavas@(protected)]
> Sent: Wednesday, March 07, 2007 9:11 PM
> To: j-dev@(protected)
> Cc: Ludger Buenger
> Subject: Re: xerces Range unnecessarily not garbage-collectable if not
> detached...
>
>
> Hi Ludger,
>
> Oddly enough I was thinking about this (again) yesterday while I was
> fixing some bugs in the Range implementation.
>
> Definitely interested in moving to WeakReferences. NodeIterators,
> EventListeners, user data and probably a few other things in the DOM
> implementation which hold on to objects which could otherwise
> be garbage
> collected would also benefit from this. It's been on my TODO
> list since we
> dropped JDK 1.1 support, though I first wanted to check what
> impact it
> would have on performance and serialization compatibility (i.e.
> java.io.Serializable) if any before committing these
> improvements. Just
> haven't gotten around to it yet.
>
> Thanks.
>
> Michael Glavassevich
> XML Parser Development
> IBM Toronto Lab
> E-mail: mrglavas@(protected)
> E-mail: mrglavas@(protected)
>
> "Ludger Buenger" <ludger.buenger@(protected)> wrote on 03/07/2007
> 11:12:54 AM:
>
> > Hello Mr. Glavassevich,
> >
> > Since I am not actively registered to the xerces-mailinglist and
> > unsure whether this is an issue with regards to the xerces
> > philosophy, I write to you as the one most involved with
> this project.
> >
> > During development we discovered that DOM ranges always have to be
> > detached as defined by the DOM Range specification.
> >
> > This sure is a useful behaviour for a general API but since Java has
> > an inbuild-garbage-collector there is no need to have to rely upon
> > external developers allways actively detaching a range (however it
> > still might be desirable).
> >
> > Since I patched xerces for our own product to use garbage-
> > collectable ranges, I'd like to ask whether you or the community is
> > interested to add this 'feature'?
> >
> > Sincerely yours,
> >
> > Ludger Bünger
> >
> > --
> > Ludger Bünger
> > RealObjects GmbH
>

---------------------------------------------------------------------
To unsubscribe, e-mail: j-dev-unsubscribe@(protected)
For additional commands, e-mail: j-dev-help@(protected)