New question: What's the right thing for tables with integer primary keys?

[James Turner]

Ok, rephrased second question after some thought.

Consider the following:
<database name="stocktrack"  defaultIdMethod="native">

   <table name="STOCK">
     <column name="STOCK_ID" primaryKey="true" required="true" type="INTEGER"/>
     <column name="STOCK_SYMBOL" required="true" type="VARCHAR" size="10"/>
     <column name="STOCK_LONG_NAME" required="true" type="VARCHAR" size="30"/>
     <column name="STOCK_TYPE_ID" required="true" type="CHAR" size="1"/>
   </table>
</database>

If I instantiate this under MySQL, I get the following SQL:

CREATE TABLE STOCK
(
     STOCK_ID INTEGER NOT NULL,
     STOCK_SYMBOL VARCHAR (10) NOT NULL,
     STOCK_LONG_NAME VARCHAR (30) NOT NULL,
     STOCK_TYPE_ID CHAR (1) NOT NULL,
     PRIMARY KEY(STOCK_ID),
   );

Here's the question:  By using defaultIdMethod=native, I'd think that I'm 
implying that integer primary key tables should autoincrement, since if I 
used defaultIdMethod=idBroker, that's the behavior I'd see.  This (to me) 
means that this should generate this instead:

CREATE TABLE STOCK
(
     STOCK_ID INTEGER AUTO_INCREMENT NOT NULL,
     STOCK_SYMBOL VARCHAR (10) NOT NULL,
     STOCK_LONG_NAME VARCHAR (30) NOT NULL,
     STOCK_TYPE_ID CHAR (1) NOT NULL,
     PRIMARY KEY(STOCK_ID),
   );

Can anyone think of a case where this would not be the right thing to 
do?  Or should I just by manually specifying autoincrement in the XML?

James


--
To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>