Thread: [ADMIN] unique constraints
PostgreSQL 9.5
I have a table:
CREATE TABLE test
(
id integer NOT NULL,
tag_id integer,
CONSTRAINT test_id_tag_id_key UNIQUE (id, tag_id)
)
WITH (
OIDS=FALSE
);
I noticed I can insert this into the database:
Insert into test (id, tag_id) values(1,null);
Insert into test (id, tag_id) values(1,null);
But this does not work:
Insert into test (id, tag_id) values(2,1);
Insert into test (id, tag_id) values(2,1); (error occurs)
Select * from test;
1, null
1, null
2, 1
I understand why in the second set of inserts I get an error when trying to insert the same set of values. But why does the first example work? I would have thought that it would have only allowed the first insert to work.
Thanks,
Lance
Because NULL != NULL, therefore you can’t have a unique value for NULL.
From: pgsql-admin-owner@postgresql.org [mailto:pgsql-admin-owner@postgresql.org] On Behalf Of Campbell, Lance
Sent: Friday, December 16, 2016 8:30 AM
To: pgsql-admin@postgresql.org
Subject: [ADMIN] unique constraints
PostgreSQL 9.5
I have a table:
CREATE TABLE test
(
id integer NOT NULL,
tag_id integer,
CONSTRAINT test_id_tag_id_key UNIQUE (id, tag_id)
)
WITH (
OIDS=FALSE
);
I noticed I can insert this into the database:
Insert into test (id, tag_id) values(1,null);
Insert into test (id, tag_id) values(1,null);
But this does not work:
Insert into test (id, tag_id) values(2,1);
Insert into test (id, tag_id) values(2,1); (error occurs)
Select * from test;
1, null
1, null
2, 1
I understand why in the second set of inserts I get an error when trying to insert the same set of values. But why does the first example work? I would have thought that it would have only allowed the first insert to work.
Thanks,
Lance
Because null is not a normal value you can compare against. Null basically is equivalent to "do not know". I haven't caught up wit the latest SQL flavors of pgsql, but old school you would do this by adding this :PostgreSQL 9.5
I have a table:
CREATE TABLE test
(
id integer NOT NULL,
tag_id integer,
CONSTRAINT test_id_tag_id_key UNIQUE (id, tag_id)
)
WITH (
OIDS=FALSE
);
I noticed I can insert this into the database:
Insert into test (id, tag_id) values(1,null);
Insert into test (id, tag_id) values(1,null);
But this does not work:
Insert into test (id, tag_id) values(2,1);
Insert into test (id, tag_id) values(2,1); (error occurs)
Select * from test;
1, null
1, null
2, 1
I understand why in the second set of inserts I get an error when trying to insert the same set of values. But why does the first example work? I would have thought that it would have only allowed the first insert to work.
CREATE UNIQUE INDEX test_id_uk ON test(id) WHERE (tags_id IS NULL);
Thanks,
Lance
-- Achilleas Mantzios IT DEV Lead IT DEPT Dynacom Tankers Mgmt