Re: transaction in postgres

> Hi averyone,
> I'm using a new function to include two query into an unique
> transaction. Is this the only way to do it? The following lines show
> you my function, the purpose is to return the id_utente field (serial)
> of the last insert done: ----------------------------------------
> CREATE FUNCTION "id_from_insert"("varchar", "varchar", "int4",
> "bpchar", "varchar", "int4", "int4", "int4", "varchar", "int4",
> "bpchar", "varchar", "varchar") RETURNS "int4" AS 'INSERT INTO
> g_registrazione_generale ( nome,
> cognome,
> data_di_nascita,
> genere,
> indirizzo,
> codice_postale,
> telefono,
> id_documentazione,
> email,
> data_inserimento,
> stato,
> velocita_connessione,
> istruzione
> ) VALUES (
> $1,
> $2,
> $3,
> $4,
> $5,
> $6,
> $7,
> $8,
> $9,
> $10,
> $11,
> $12,
> $13
> );
> SELECT max(g_registrazione_generale.id_utente) FROM
> g_registrazione_generale;
> ' LANGUAGE 'sql';


Use a query like this: select curval('sequence_name') as nextid;
Before you can issue this query, you *must* have started a transaction(
begin work ).Its mutch simpler and you dont have to mess with record locking( in your
solution there is no garantee that you would get the max value, because
another connection may have inserted a new record ).
Miguel