I am used to creating sequences separately and using them for default
values instead of using the serial type. I am not sure why... but that's the
way I have been doing it. Maybe I don't trust the generated names.
I should probably just get over it, but anyhow...
Let's say I am writing a file for creating a database. Say ...
-- schematest
CREATE SEQUENCE foo_id_seq;
CREATE TABLE foo(
id integer
PRIMARY KEY
DEFAULT nextval('foo_id_seq'),
xx text
);
I want to be able to ...
lee=# drop schema beta cascade;
DROP SCHEMA
lee=# create schema beta;
CREATE SCHEMA
lee=# set search_path to beta;
SET
lee=# \i schematest
CREATE SEQUENCE
CREATE TABLE
The problem is that unlike a comparable serial column, my default value
does not qualify the sequence with the right schema name.
Is there a way to do that, such that my schematest file does not need to
know which schema I am loading the file into?
Or should I just get over my aversion to serial?
_________________________________________________________________
Working moms: Find helpful tips here on managing kids, home, work � and
yourself. http://special.msn.com/msnbc/workingmom.armx
