I like these changes. I posted a message a while ago about some problems
with this task but got no responses. The changes I like is
1) specify schema in call DatabaseMetaData.getTables/getColumns method
instead of null. In Oracle, the username can be used as the schema name
but must be in upper case. I don't know about other dbs, so an attribute
for schema should be fine.
2) In DatabaseMetaData.getTables method, the last parameter can be
specified as {TABLE} so no need to check for table.
Howard Lin
> -----Original Message-----
> From: Dan Bachelder [mailto:[EMAIL PROTECTED]]
> Sent: Tuesday, October 09, 2001 11:42 PM
> To: [EMAIL PROTECTED]
> Subject: Re: TorqueJDBCTransformTask - additions
>
>
> on a related note.. specifying the database schema for operations on
> DatabaseMetaData is important or we can end up with multiple
> definitions of
> the same tables.. that may exist in different schemas... I
> added a dbSchema
> attribute to the task as well...
>
> ----- Original Message -----
> From: "Dan Bachelder" <[EMAIL PROTECTED]>
> To: <[EMAIL PROTECTED]>
> Sent: Tuesday, October 09, 2001 9:54 PM
> Subject: TorqueJDBCTransformTask - additions
>
>
> > I was having some problems with TorqueJDBCTransformTask
> with a very large
> > oracle db schema.
> >
> > I added some .close()s to get rid of an out of cursors exception...
> >
> > I was then getting out of memory exceptions which I wasn't
> as sure how to
> > clean up.. I assume the 300+ tables in memory as XML
> elements was a bit
> much
> > and the schema would need to be written to disk periodically.
> >
> > I added the includeTables and excludeTables attributes to
> the task... if
> > these attributes contain CSV of tables to include or
> exclude respectively,
> > they will be included or excluded... empty values on both of these
> > attributes makes the task behave as always. The
> includeTables attribute
> > forces the task to add only the tables in the list so
> > includeTables="MY_TABLE,MY_SECOND_TABLE" would generate an
> XML schema with
> > just those two tables.
> >
> > this makes getting an XML schema for part of a large legacy DB much
> quicker
> > and more efficient. it is also handy for DBs like oracle where you
> probably
> > don't want XML generated for DUAL or *_PRIVILEGE_MAP.
> >
> > I also added a bit of logging so one can see what is
> happening... over VPN
> > this process can take quite a long time so I find it
> helpful to see what
> > step the process is on.
> >
> > Is there any interest for making any of this a patch? there
> are obviously
> > some things that could be added like regex... but it's a start...
> >
>
>
> ---------------------------------------------------------------------
> 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]
