Am Dienstag, den 12.07.2005, 12:47 -0300 schrieb Adam O'Toole:
> I am trying to INSERT multiple rows to a table using a stored procedure
> something like this:
>
> CREATE FUNCTION test(varchar) RETURNS int2 AS '
> DECLARE
> id_list ALIAS FOR $1;
> BEGIN
> INSERT INTO history (media_id, media_type) SELECT media.media_id,
> media.media_type WHERE media.media_id IN (id_list);
> .
> .
>
> So I would call this function by passing desired media_id's to be put in the
> history table like this:
> SELECT test( '24,25,26,27' );
> In the INSERT statement, I am taking values from a table called Media, and
> adding a row to a table called History. In this example, the only rows
> copied would be rows where the media_ID was 24,25,26 or 27. This function
> is working for me, but it only works if the varChar being passed has only
> one value, like this:
> SELECT test('24'); This works.
> But when I try to pass more then one value in the list ( '24,25'), the
> function runs with no errors but does not add the rows, it does nothing.
>
> What am I doing wrong? Do I have the syntax wrong for using INSERT with a
> SELECT-IN statement?
>
No, you only do the IN statement wrong. This is not related to
the INSERT. IN wants SQL literal, not a text argument or anything.
All you can do is to either use a function from contrib to
ilterate thru an array instead or build up the SQL from text
string and use EXECUTE and friends - see pl/pgsql reference and
examples.
Regards
Tino
---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
choose an index scan if your joining column's datatypes do not
match
