It seems im not clear enough.
Here I go again.
Lets suppose I have a mapping like the following
(you may skip the mapping and see the small chart
below):
<class-descriptor class="B" table="B">
<field-descriptor
name="id"
column="B_id"
jdbc-type="INTEGER"
primarykey="true"
sequence-name="B_id_seq"
autoincrement="true"/>
<field-descriptor
name="c"
column="c"
jdbc-type="CHAR"/>
</class-descriptor>
<class-descriptor class="D" table="D">
<field-descriptor
name="id"
column="D_id"
jdbc-type="INTEGER"
primarykey="true"
sequence-name="D_id_seq"
autoincrement="true"/>
<field-descriptor
name="e"
column="e"
jdbc-type="CHAR"/>
<field-descriptor
name="f"
column="f"
jdbc-type="CHAR"/>
</class-descriptor>
<class-descriptor class="C" table="C">
<field-descriptor
name="id"
column="C_id"
jdbc-type="INTEGER"
primarykey="true"
sequence-name="C_id_seq"
autoincrement="true"/>
<field-descriptor
name="D_id"
column="D_id"
jdbc-type="INTEGER"/>
<reference-descriptor
name="D"
class-ref="D"
auto-update="false"
auto-delete="false"
auto-retrieve="true">
<foreignkey field-ref="D_id"/>
</reference-descriptor>
<field-descriptor
name="d"
column="d"
jdbc-type="CHAR"/>
</class-descriptor>
<class-descriptor class="A" table="A">
<field-descriptor
name="id"
column="A_id"
jdbc-type="INTEGER"
primarykey="true"
sequence-name="A_id_seq"
autoincrement="true"/>
<field-descriptor
name="B_id"
column="B_id"
jdbc-type="INTEGER"/>
<reference-descriptor
name="B"
class-ref="B"
auto-update="false"
auto-delete="false"
auto-retrieve="true">
<foreignkey field-ref="B_id"/>
</reference-descriptor>
<field-descriptor
name="C_id"
column="C_id"
jdbc-type="INTEGER"/>
<reference-descriptor
name="C"
class-ref="C"
auto-update="false"
auto-delete="false"
auto-retrieve="true">
<foreignkey field-ref="C_id"/>
</reference-descriptor>
<field-descriptor
name="a"
column="a"
jdbc-type="CHAR"/>
<field-descriptor
name="b"
column="b"
jdbc-type="CHAR"/>
</class-descriptor>
------------
Which I really think is easier to understand if you
look this chart
B(c) This means: A has a FK to B and C and A has properties a,b.
/ B has property c.
A(a,b) C has a FK to D and has property d.
\ D has properties e and f.
C(d)
\
D(e,f)
I am talking about the cases when you dont want the complete object.
I know that "select x from A" would bring me a DList with A Objects,
thats clear. My question is:
If I submit a query like this one:
"select x.C.D.e, x.B.c from A;"
Im not generating any already-defined objects so I was expecting to get
a DList of Object[] or something like that and I get a
org.apache.ojb.broker.util.collections.ManageableArrayList
not a DList anymore, why is that?
On Fri, 2004-07-02 at 19:49, Sukesh Garg wrote:
> check out the following link for the OJB ODMG API
>
> http://db.apache.org/ojb/docu/tutorials/odmg-tutorial.html
>
> The container is defined in the repository.xml file.
>
> http://db.apache.org/ojb/docu/tutorials/mapping-tutorial.html
>
>
> -A
>
> On Fri, 2004-07-02 at 07:01, Martin I. Levi wrote:
> > Aren't you supposed to do something like:
> >
> > select A.a, A.b, A.B.c, A.C.d, A.C.D.e, A.C.D.f from A,B,C,D;
> >
> > ???
> >
> > And THE BIG question is... when you perform an OQLQuery.execute() this
> > returns and Object what critery should I follow to choose the container
> > to store the resulting objects?
> >
> >
> >
> > On Thu, 2004-07-01 at 19:49, Auckyboy wrote:
> > > I'll take a stab at it.
> > >
> > > when performing OQL queries using OJB, you get back the complete object
> > > A with references to B, C and further objects if they exist otherwise
> > > they are null.
> > >
> > > Assuming you have the following objects..
> > >
> > > A{
> > > B b;
> > > int bId;
> > > C c;
> > > int cId;
> > > }
> > >
> > > and
> > >
> > > C{
> > > D d;
> > > int dId;
> > > }
> > >
> > >
> > > D{
> > > id;
> > > name;
> > > }
> > > Now you want A where B, C and D are defined...
> > >
> > > select a from A where bID != 0 and cId!=0 and c.dId!=0
> > >
> > > this in effect would give you A where B,C and D are defined.
> > >
> > > However, as a suggestion, if you need to perform queries on A, B, C
> > > etc..its better to use...
> > >
> > > select a from A where c.d.name = "somename"
> > >
> > >
> > > HTH
> > > -A
> > >
> > > On Thu, 2004-07-01 at 10:19, Martin I. Levi wrote:
> > > > Hi!
> > > >
> > > > Well, here I am with the ODMG 3.0 book trying to perform OQL querys
> > > > through OJB but I really think I just dont get it.
> > > > Lets suppose I have a table A with FK to B and C where C has a FK to D
> > > > Something like this:
> > > > B
> > > > /
> > > > A
> > > > \
> > > > C
> > > > \
> > > > D
> > > >
> > > > Which are mapped to the objects A, B, C and D.
> > > > Now I want to perform a query which in SQL would be:
> > > >
> > > > SELECT A.a, A.b, B.c, C.d, D.e, D.f
> > > > FROM A,B,C,D
> > > > WHERE A.B_key = B.B_key
> > > > AND A.C_key = C.C_key
> > > > AND C.D_key = D.D_key
> > > > ;
> > > >
> > > > My questions are...
> > > > 1) how could I do this on an OQL query through OJB?
> > > > 2) which kind of container should I use to store the result? An
> > > > ArrayList does always do the trick?, I suppose it doesn't...
> > > >
> > > > Any help will be useful!
> > > >
> > > > --
> > > > Saludos,
> > > >
> > > > Martin I. Levi
> > > >
> > > > Centre Tecnol�gic de Transferenci�ncia de Calor
> > > > Universitat Polit�cnica de Catalunya
> > > > www.cttc.upc.edu
> > > >
> > > >
--
Saludos,
Martin I. Levi
Centre Tecnol�gic de Transferenci�ncia de Calor
Universitat Polit�cnica de Catalunya
www.cttc.upc.edu
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
