From: | "Lee Harr" <missive(at)hotmail(dot)com> |
---|---|
To: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: sequence in schema -- broken default |
Date: | 2004-01-23 20:14:53 |
Message-ID: | BAY2-F172yAeYQsXMqf0001a39e@hotmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
>>By the way... should the way you wrote it work?
>>
>No. I did not complete the syntax.
>
>
>># INSERT INTO one.foo VALUES ();
>>ERROR: syntax error at or near ")" at character 29
>>
>>Seems like maybe it should work with the default, but I don't know.
>>
>No, the reason the below works is version() is a function where values
>() is not.
Yes. I was just including that in case something had changed in
recent versions. I guess it just seemed strange to me that this
works ...
# create table x (a int, b int);
CREATE TABLE
# insert into x values(5);
INSERT 18518 1
but this does not ...
# insert into x values();
ERROR: syntax error at or near ")" at character 22
>I don't think you are going to have any choice but to hardcode the
>sequence value
>unless you want to bounce in between search paths based on who is
>connecting.
All I really want is some way to ensure that the DEFAULT is hooked
up to the right sequence.
I guess maybe it could be considered a feature that you can code
the nextval as a relative name and have the value pulled from different
sequences depending on your search_path, but I think a more useful
pattern is to always pull from the same sequence.
Of course, I guess that is why there is a serial type ;o)
Thanks for your time.
_________________________________________________________________
Add photos to your e-mail with MSN 8. Get 2 months FREE*.
http://join.msn.com/?page=features/featuredemail
From | Date | Subject | |
---|---|---|---|
Next Message | Martín Marqués | 2004-01-23 20:15:56 | Re: If table A value IS NULL then table B |
Previous Message | Holger Marzen | 2004-01-23 19:44:57 | Article in German iX magazine |