Hi Rainer. I built the recursive walker outside of castor. I did use the ClassDescriptor/FieldDescriptor to walk the graph.
I'm not sure of the performance implications yet. I believe that in many cases the database would only be updated if the object has been modified or the dirty checking that castor does determines that an update is necessary. -- Don -----Original Message----- From: Rainer Mueller [mailto:[EMAIL PROTECTED]] Sent: Thursday, June 06, 2002 9:44 AM To: [EMAIL PROTECTED] Subject: Re: [castor-dev] Any luck with many-to-many queries? Hello Don! > For 1) > I solved the problem with storing many-to-many relationship by making sure > all objects in the relationship were visible to the current db transaction. > I use long transactions and hence this was causing a problem with them > persisting. The solution was to make a recursive walker that walked across > the parent object. When it came to non-dependent collections I would make > sure to call db.update on those object(s) directly. After this I would then > call db.commit() and the relationships would be saved. Update the database its data every time when you are calling db.update? I think it would be getting very slow - am I right? I have the same problem with long transactions. Is your code in the CVS or is just a patch for your application? Kind regards Rainer ----------------------------------------------------------- If you wish to unsubscribe from this mailing, send mail to [EMAIL PROTECTED] with a subject of: unsubscribe castor-dev ----------------------------------------------------------- If you wish to unsubscribe from this mailing, send mail to [EMAIL PROTECTED] with a subject of: unsubscribe castor-dev
