Comments on: A ScalaQuery Update

By: philip andrew

philip andrew — Fri, 23 Oct 2009 09:50:25 +0000

Just came across this via DZone http://www.sts.tu-harburg.de/people/mi.garcia/ScalaQL/

By: Stefan Zeiger

Stefan Zeiger — Wed, 21 Oct 2009 19:09:48 +0000

@JL Huynh: Agreed, there needs to be an option for specifying the sequence name. If it’s missing, a name should be generated automatically from the table and column names.

@philip andrew: I read the lift tutorial and experimented with the code, but I haven’t gotten around to building a real application with it. I’d like to get started with that after getting ScalaQuery 1.0 out of the door, and try to build the database connectivity both with Lift’s ORM and with ScalaQuery. What is your experience with this combination so far?

By: philip andrew

philip andrew — Wed, 21 Oct 2009 07:39:52 +0000

Hi,

I use this for two projects now, one is a straight Scala project application the other is using Liftweb.
Liftweb http://liftweb.net/ has some database mapping classes some simular but not as good as this one database mapping features. Do you think that integration of this with liftweb would be useful?

Thanks, Philip

By: JL Huynh

JL Huynh — Sun, 18 Oct 2009 20:10:23 +0000

May be a new option O AutoIncSeq “sequenceName” would work well. I usually need to interact with existing tables that already have particular sequence names assigned to them.

By: Stefan Zeiger

Stefan Zeiger — Sun, 18 Oct 2009 15:24:16 +0000

Ah, I didn’t think of such cases with computed columns. I suppose they could naturally fall out of an implementation for INSERT…FROM statements which is planned for the future.

The special case of simulating an auto_increment column in Oracle should ideally be handled by ScalaQuery directly. It offers an AutoInc flag among the basic column options, so it should be supported for all drivers. We could automatically create a sequence (together with the table) and get a value from it when an INSERT is done without the AutoInc column or with a NULL value for it.

By: JL Huynh

JL Huynh — Sun, 18 Oct 2009 14:16:09 +0000

Trying to insert rows into an oracle table using the syntax:
val ins1 = (Users.id ~ Users.first ~ Users.last).insert((userSequence.next, “Homer”, Some(”Simpson”)))

Users.id is defined as: column[Int](”userID”, O PrimaryKey, O NotNull)
And userSequence as: val userSequence = Sequence[Int](”user_seq”) start 1 inc 1

The does not compile because userSequence.next is not an Int.

How do I use the combinator syntax to get it to translate to:
insert into users (id, first, last) values (user_seq.nextval, “Homer”, “Simpson”);

By: JL Huynh

JL Huynh — Sun, 18 Oct 2009 01:12:14 +0000

Thank you. Breaking out in individual statements did solve the problem.

Batches inserts would be useful for the kinds of things I do.

By: Stefan Zeiger

Stefan Zeiger — Sat, 17 Oct 2009 10:07:16 +0000

@JL Huynh: Your comments got eaten by the spam filter. I’ve restored them now.

I don’t think Oracle likes multiple statements being concatenated to one statement “s1; s2″ as in populateUsers. If you want a portable solution, you’ll have to send the statements separately. In the future, ScalaQuery may support statement batches (if the JDBC driver does) to avoid the overhead of sending multiple requests in a portable way.

For an Oracle-only solution, “BEGIN s1; s2; END” should work IIRC.

By: JL Huynh

JL Huynh — Sat, 17 Oct 2009 06:18:16 +0000

Guess it’s not moderated. Reposting:

I’m getting a “ORA-00933: SQL command not properly ended” when running the SimpleTest class against an oracle DB. The error occurs on the for(i <- populateUsers) println(" "+i) statement.

The USERS table is getting created, so I know I'm getting a connection and at least the first statement works.

What is the best way to troubleshoot this?

By: JL Huynh

JL Huynh — Sat, 17 Oct 2009 06:16:12 +0000

Are comments moderated? Just posted something and it did not show up…