I have a Postgres database with many tables. I want to declare a function that reset all auto_increment values (next_value) of all the tables to max(id) of the table. For example, if max(id) of sample_table is 22:
postgres=# SELECT MAX(id) + 1 FROM sample_table;
?column?
----------
22
(1 row)
While next_value of the table is 7:
postgres=# SELECT nextval('"public"."sentry_project_id_seq"'::regclass);
nextval
---------
7
(1 row)
I want to set the next_value of all the tables to their max(id) by:
SELECT setval('auth_user_id_seq', (SELECT MAX(id) from "auth_user"));
To do this, I declare a function called reset_auto_increment by:
CREATE OR REPLACE FUNCTION reset_auto_increment(username IN VARCHAR) RETURNS void AS $$
DECLARE
statements CURSOR FOR
SELECT tablename FROM pg_tables
WHERE tableowner = username AND schemaname = 'public';
BEGIN
FOR stmt IN statements LOOP
EXECUTE 'SELECT setval(' || concat(stmt.tablename, '_id_seq') || ' , (SELECT MAX(id) from ' || quote_ident(stmt.tablename) || '));';
END LOOP;
END;
$$ LANGUAGE plpgsql;
But the function declared by the statement above has error below:
postgres=# select reset_auto_increment('postgres');
**ERROR: column "_id_seq" does not exist**
LINE 1: SELECT 'SELECT setval(' || concat(stmt.tablename, "_id_seq")...
^
QUERY: SELECT 'SELECT setval(' || concat(stmt.tablename, "_id_seq") || ' , (SELECT MAX(id) from ' || quote_ident(stmt.tablename) || '));'
CONTEXT: PL/pgSQL function reset_auto_increment(character varying) line 8 at EXECUTE
How to correct this function?