If you set your schema up correctly you don't need the id.
Here is what you should have:
<table name="DETAILS">
...
<foreign-key foreignTable="REQUEST">
<reference local="REQUEST" foreign="ID"/>
</foreign-key>
</table>
When torque generates the OM layer it will generate a method:
Request.addDetails(Details)
[BTW: It is better to name your tables in the singular - i.e. DETAIL
rather than DETAILS. I also find it better to have more meaningful
column names - i.e. DETAIL_ID, REQUEST_ID, etc. and to use
the full key name where appropriate - i.e. DETAILS would use
REQUEST_ID rather than REQUEST. But these may just be my
preferences.]
So this means that you don't need the id. You go like this:
Request r = new Request();
r.setSrc(3);
...
Details d = new Details();
d.setCode(5);
...
... and here is the magic line:
r.addDetails(d);
Now when it comes to saving the records all you need to do is:
r.save();
... and this will save all of the Details records you have added to the
Request as well as the Request itself. I am pretty sure it also wraps
them in the appropriate transaction logic (which is I guess ignored for
MySQL - can anyone tell me it works when using one of the table
drivers that does support transactions - i.e. BerkeleyDB, InnoDB
Gemini, etc.)
Whilst the code generated by torque doesn't do everything, it does
do a hell of a lot. Set your schema up correctly and then take the
time to examine the methods generated in the Base~ and Base~Peer
classes - there are plenty of goodies to make your life easier.
HTH
Scott
----- Original Message -----
From: "Andrea Papotti" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Friday, August 31, 2001 9:00 PM
Subject: autoincrement and key retriving
> I have two tables like:
>
> <table name="REQUEST">
> <!-- substitution requests (header) -->
> <column name="ID" type="INTEGER" required="true"
> autoIncrement="true" primaryKey="true"/>
> <column name="SRC" type="INTEGER" required="true"/>
> <column name="DST" type="INTEGER" required="true"/>
> <column name="DATE" size="8" type="VARCHAR" required="true"/>
> <column name="NOTE" size="80" type="VARCHAR" required="true"/>
> </table>
>
> <table name="DETAILS">
> <!-- substitution requests (detail) -->
> <column name="ID" type="INTEGER" required="true"
> autoIncrement="true" primaryKey="true"/>
> <column name="REQUEST" type="INTEGER" required="true"/>
> <column name="CODE" type="INTEGER" required="true"/>
> <column name="QTY" type="INTEGER" required="true"/>
> </table>
>
> which are related via REQUEST.ID = DETAILS.REQUEST
>
>
> I need to write 1 record in "REQUEST" fallowed by many records in "DETAILS",
>
> try
> {
> Richieste header = new Request();
> header.setSrc( src_code );
> header.setDst( dst_code);
> header.setDatel( today() );
> header.setNote( note );
> header.save();
>
> header_id = header.getId().getBigDecimal().intValue();
> }
> catch (Exception e)
> {
> context.put("error", "header: " + e.toString());
> return;
> }
>
>
> fallowed by code to write records in "DETAILS".
>
> BUT
>
> header_id = header.getId().getBigDecimal().intValue();
>
> fails bacause header.getId() returns null.
>
> the question is:
>
> I must re-read the record just saved (with the immediatly previous
> statment), or there is a method to obtain the value of the primary-key just
> afer the .save() call?
>
> Re-read the record could be a problem, because the only unique field is
> right the ID.
>
> tia, Andrea =8-)
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
>
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
