So currently the only way to specify a row type is by using a table,
right? E.g.:

    CREATE TABLE t2_retval (
        value1 int NOT NULL DEFAULT -1,
        value2 int NOT NULL,
        value3 int

Are there plans to add another way of declaring this sort of thing so
that I don't have to add a new table to my schema for every function
that returns a rowtype?

Second, it seems that there's still a problem with NULLs here:

        RETURNS t2_retval
    AS '
        retval t2_retval%ROWTYPE;
        SELECT INTO retval null, null, null;
        RETURN retval;
    ' LANGUAGE 'plpgsql';

This is returning a row that (to my mind) doesn't match the type of the
table above, because it's returning null for non-nullable columns:

cjs=> select coalesce(value1, -999), coalesce(value2, -999),
cjs->     coalesce(value3, -999) from t2();
 case | case | case
 -999 | -999 | -999
(1 row)

(You get the same result if you delete the SELECT INTO line above.)

Am I misunderstanding something here, or is this a bug?

Curt Sampson  <[EMAIL PROTECTED]>   +81 90 7737 2974
    Don't you know, in this new Dark Age, we're all light.  --XTC

---------------------------(end of broadcast)---------------------------
TIP 4: Don't 'kill -9' the postmaster

Reply via email to