RULE vs. SEQUENCE - Mailing list pgsql-hackers

From Karel Zak
Subject RULE vs. SEQUENCE
Date
Msg-id Pine.LNX.3.96.1000904115840.219D-100000@ara.zf.jcu.cz
Whole thread Raw
Responses Re: RULE vs. SEQUENCE
List pgsql-hackers
Hi,
I have a question... why RULE call nexval() and data in RULE statement are
differend than data in original stmt.
An example:

create sequence a;
create table aa (id int DEFAULT nextval('a'), data text);

insert into aa (data) values ('xxxx');
insert into aa (data) values ('yyyy');

select * from aa;
id|data
--+----1|xxxx2|yyyy
(2 rows)

... all is right.

create table log (aid int, act text);

create rule a_ins as on insert to aa do insert into log (aid, act) values (NEW.id, 'INSERT');

insert into aa (data) values ('zzzz');
insert into aa (data) values ('qqqq');

test=> select * from aa;
id|data
--+----1|xxxx2|yyyy4|zzzz          <----------6|qqqq
(4 rows)

select * from log;
aid|act
---+------ 3|INSERT       <---------- 5|INSERT
(2 rows)
But I expect in 'log' table as 'aid' same number as numbers for 'zzzz' and
'qqqq'...It's interesting feature (6.5, 7.0, 7.1...). How is a possible in RULE
obtain same data as in 'aa' table for a default data from the sequence.
                Karel



pgsql-hackers by date:

Previous
From: "Hiroshi Inoue"
Date:
Subject: RE: func() & select func()
Next
From: Jan Wieck
Date:
Subject: Re: RULE vs. SEQUENCE