On Dec 9, 2007, at 11:05 , Alain Roger wrote:
> Hi,
>
> i would like to understand why the following INSERT INTO statement
> works :
>
> INSERT INTO mytable
> SELECT nextval('my_sequence'),
> 'myname',
> 'myfirstname'
> ;
>
> whereas usually we should do :
>
> INSERT INTO mytable
> VALUES
> (
> SELECT nextval('my_sequence'),
> 'myname',
> 'myfirstname'
> );
>
Well, imho, if the sequence was set up via serial (or otherwise is
set as the default for the first column), I think the easiest way is :
INSERT INTO mytable (name, firstname)
VALUES ('myname', 'myfirstname');
No need to include the nextval call at all.
If you look at the INSERT synoposis:
http://www.postgresql.org/docs/8.2/static/sql-insert.html
INSERT INTO table [ ( column [, ...] ) ]
{ DEFAULT VALUES | VALUES ( { expression | DEFAULT } [, ...] )
[, ...] | query }
[ RETURNING * | output_expression [ AS output_name ] [, ...] ]
you can see that a VALUES expression or a query are legitimate forms
for INSERT. The query form is particularly useful if you'd like to
insert a number of rows that are the result of a SELECT. For example,
when loading data from a temp table.
INSERT INTO mytable (name, firstname)
SELECT name, firstname
FROM temp_table;
Michael Glaesemannn
grzm seespotcode net