You are absolutely right Michael.
In CRoom have overriden setSite method :
public void setSite(CSite site) {
if (getSite() != null) {
getSite().removeFromRooms(this);
}
super.setSite(site);
}
and everything works.
Does it mean that I have to override setter in every many to one
relationship ? Is there any reason why cayenne itself shouldn't do
that ?
It seems not right for me.
Regards
Marcin
On 02/05/2006, at 11:21 PM, Gentry, Michael ((Contractor)) wrote:
Marcin,
I'm not certain this is a Cayenne bug. I think it is more of a
misunderstanding. When you do:
logger.info("(here relationship is modified)");
temproom.setSite(tempsite2);
You point the room's site to site2, but you never unlinked the
relationship from site1->room. Therefore, when you then do:
logger.info("site 1 rooms");
rooms = tempsite1.getRooms();
for (int i=0; i<rooms.size();i++) {
logger.info("\troom "+i+" :
"+((CRoom)rooms.get(i)).getName());
}
tempsite1 will still print the room out because the room is still it
site1's rooms relationship. Try:
logger.info("(here relationship is modified)");
tempsite1.removeFromRooms(temproom);
temproom.setSite(tempsite2);
And see if it helps.
/dev/mrg
-----Original Message-----
From: Marcin Skladaniec (JIRA) [mailto:[EMAIL PROTECTED]
Sent: Monday, May 01, 2006 9:53 PM
To: [email protected]
Subject: [JIRA] Created: (CAY-538) Relationship not refreshed before
committing the context.
Relationship not refreshed before committing the context.
----------------------------------------------------------
Key: CAY-538
URL: http://issues.apache.org/cayenne/browse/CAY-538
Project: Cayenne
Type: Bug
Components: Cayenne Core Library
Versions: 1.2 [BETA]
Environment: I think any platform, but tested on os x and freebsd.
Cayenne 3t
Reporter: Marcin Skladaniec
I have two sites : site1 and site2 and a room linked to site1
If I change the relationship so the room links to site2, than before I
commit the changes both sites are linked to the room.
The code
{
Expression expression =
ExpressionFactory.matchExp(CSite.NAME_PROPERTY, "site1");
SelectQuery query = new SelectQuery(CSite.class,
expression);
List list = context2.performQuery(query);
CSite tempsite1 = (CSite)list.get(0);
expression =
ExpressionFactory.matchExp(CSite.NAME_PROPERTY, "site2");
query = new SelectQuery(CSite.class, expression);
list = context2.performQuery(query);
CSite tempsite2 = (CSite)list.get(0);
expression =
ExpressionFactory.matchExp(CRoom.NAME_PROPERTY, "room");
query = new SelectQuery(CRoom.class, expression);
list = context2.performQuery(query);
CRoom temproom = (CRoom)list.get(0);
logger.info("site 1 rooms");
List rooms = tempsite1.getRooms();
for (int i=0; i<rooms.size();i++) {
logger.info("\troom "+i+" :
"+((CRoom)rooms.get(i)).getName());
}
logger.info("site 2 rooms");
rooms = tempsite2.getRooms();
for (int i=0; i<rooms.size();i++) {
logger.info("\troom "+i+" :
"+((CRoom)rooms.get(i)).getName());
}
logger.info("(here relationship is modified)");
temproom.setSite(tempsite2);
logger.info("site 1 rooms");
rooms = tempsite1.getRooms();
for (int i=0; i<rooms.size();i++) {
logger.info("\troom "+i+" :
"+((CRoom)rooms.get(i)).getName());
}
logger.info("site 2 rooms");
rooms = tempsite2.getRooms();
for (int i=0; i<rooms.size();i++) {
logger.info("\troom "+i+" :
"+((CRoom)rooms.get(i)).getName());
}
}
produces output :
site 1 rooms
room 0 : room
site 2 rooms
(here relationship is modified)
site 1 rooms
room 0 : room
site 2 rooms
room 0 : room
I can publish a small project reproducing this bug.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the
administrators:
http://issues.apache.org/cayenne/secure/Administrators.jspa
-
For more information on JIRA, see:
http://www.atlassian.com/software/jira
-------------------------->
ish
http://www.ish.com.au
Level 1, 30 Wilson Street Newtown 2042 Australia
phone +61 2 9550 5001 fax +61 2 9550 4001
