I am setting a session variable inside a postgres function and the values are not getting set.
Any help is most appreciated. Thanks in advance.
I am using "PostgreSQL 10.6, compiled by Visual C++ build 1800, 64-bit"
My code is as follows:
The function:
CREATE FUNCTION set_rp_vals(iv_rp_company varchar, iv_rp_portfolio varchar)
RETURNS integer
LANGUAGE plpgsql
SECURITY DEFINER
AS $function$
DECLARE
l_retval integer;
BEGIN
l_retval := 1;
RAISE NOTICE '1.iv_rp_company: >>> %', iv_rp_company;
RAISE NOTICE '2.iv_rp_portfolio: >>> %', iv_rp_portfolio;
--set the session variable
set rp.company = iv_rp_company;
set rp.portfolio = iv_rp_portfolio;
RETURN l_retval;
EXCEPTION
WHEN OTHERS THEN
RETURN 9;
END;
$function$
;
The function call:
SELECT set_rp_vals(iv_rp_company := 'COMPAN',iv_rp_portfolio := 'PORTOF');
--Retrieving the session variables:
select
current_setting('rp.company') as company,
current_setting('rp.portfolio') as portfolio;
The value returned by the above query:
I would use the set_config() function for this:
CREATE OR REPLACE FUNCTION set_rp_vals(iv_rp_company varchar, iv_rp_portfolio varchar)
RETURNS integer
LANGUAGE plpgsql
SECURITY DEFINER
AS $function$
DECLARE
l_retval integer;
BEGIN
l_retval := 1;
RAISE NOTICE '1.iv_rp_company: >>> %', iv_rp_company;
RAISE NOTICE '2.iv_rp_portfolio: >>> %', iv_rp_portfolio;
--set the session variable
perform set_config('rp.company', iv_rp_company, false);
perform set_config('rp.portfolio', iv_rp_portfolio, false);
RETURN l_retval;
EXCEPTION
WHEN OTHERS THEN
RETURN 9;
END;
$function$
;
CREATE FUNCTION
Execute with your values:
SELECT set_rp_vals(iv_rp_company := 'COMPAN',iv_rp_portfolio := 'PORTOF');
NOTICE: 1.iv_rp_company: >>> COMPAN
NOTICE: 2.iv_rp_portfolio: >>> PORTOF
┌─────────────┐
│ set_rp_vals │
├─────────────┤
│ 1 │
└─────────────┘
(1 row)
select
current_setting('rp.company') as company,
current_setting('rp.portfolio') as portfolio;
┌─────────┬───────────┐
│ company │ portfolio │
├─────────┼───────────┤
│ COMPAN │ PORTOF │
└─────────┴───────────┘
(1 row)
Related
https://www.postgresql.org/docs/14/plpgsql-declarations.html#PLPGSQL-DECLARATION-PARAMETERS
CREATE FUNCTION concat_selected_fields(in_t sometablename) RETURNS text AS $$
BEGIN
RETURN in_t.f1 || in_t.f3 || in_t.f5 || in_t.f7;
END;
$$ LANGUAGE plpgsql;
is in_t alias sometable alias name? I try to replicate it by following:
create function conca_selected_col_emp( in_t public.emp) returns text as $$
begin
return in_t.name || in_t.department;
end;
$$ LANGUAGE plpgsql;
Obviously, this will not work...
It should to work:
create table foo(a varchar, b varchar);
insert into foo values('hello', 'world');
create or replace function fx(arg foo)
returns varchar as $$
begin
return arg.a || ', ' || arg.b;
end;
$$ language plpgsql;
postgres=# select fx(foo) from foo;
┌──────────────┐
│ fx │
╞══════════════╡
│ hello, world │
└──────────────┘
(1 row)
or alternative syntax:
postgres=# select foo.fx from foo;
┌──────────────┐
│ fx │
╞══════════════╡
│ hello, world │
└──────────────┘
(1 row)
I need to pass dynamic number of parameters to a function as well as their data types and then return a table having those parameters as fields.
Is that possible to do that in postgres?
Any idea or example is appreciated
Here is an example that could probably be improved.
Beware of SQL injection!
CREATE OR REPLACE FUNCTION create_table(
tabname text,
VARIADIC coldef text[]
) RETURNS void
LANGUAGE plpgsql STRICT AS
$$DECLARE
l integer;
i integer;
sql text;
sep text := '';
BEGIN
l := array_upper(coldef, 1);
IF l % 2 <> 0 THEN
RAISE EXCEPTION 'Number of arguments must be odd';
END IF;
sql := 'CREATE TABLE ' || quote_ident(tabname) || '(';
FOR i IN 1 .. l/2 LOOP
sql := sql || sep || quote_ident(coldef[2*i-1]) || ' ' || quote_ident(coldef[2*i]);
sep := ', ';
END LOOP;
sql := sql || ')';
EXECUTE sql;
END;$$;
It can be used like this:
test=> SELECT create_table('tabname', 'col1', 'int4', 'col2', 'text');
test=> \d tabname
Table "laurenz.tabname"
┌────────┬─────────┬───────────┐
│ Column │ Type │ Modifiers │
├────────┼─────────┼───────────┤
│ col1 │ integer │ │
│ col2 │ text │ │
└────────┴─────────┴───────────┘
CREATE OR REPLACE FUNCTION fn_SplitArrayStr( anyelement , anyelement )
RETURNS anyarray
LANGUAGE SQL
AS $$
DECLARE f1 text , f2 text ;
BEGIN
f1 := $1::text ;
f2 := $2::text ;
SELECT * FROM UNNEST( string_to_array(f1, f2) ) as c1 ;
END ;
$$;
ERROR : Syntax error at or near "text"
LINE 2 : DECLARE f1 text , f2 text ;
How do I change?
I see two issues:
wrong language specification - PostgreSQL native procedural language is plpgsql.
DECLARE statement uses semicolon for separation between individual variable declarations.
postgres=# create or replace function foo()
returns int as $$
declare a int; b int;
begin
a := 10; b := 20;
return a + b;
end;
$$ language plpgsql;
CREATE FUNCTION
postgres=# select foo();
┌─────┐
│ foo │
╞═════╡
│ 30 │
└─────┘
(1 row)
You can try to specify the language at the end of the procedure
$$
LANGUAGE plpgsql;
You've written PL/PgSQL code but marked it LANGUAGE SQL. Use LANGUAGE plpgsql if you're writing pl/pgsql functions.
I'm busy trying to rewrite an Informix stored procedure for a PostgreSQL
database and I am stuck on something that is probably quite obvious to
everyone who know PostgreSQL.
I have my sql script as follows
-- ensure type and function get created
drop type if exists tp_users cascade;
drop function if exists sp_cmplist();
-- create type
create type tp_users as (
us_id char(30),
us_status char(1)
);
create function sp_cmplist()
returns tp_users as $$
declare
lr_users tp_users;
begin
for lr_users in
select users.us_id, users.us_status
from users
loop
return lr_users;
end loop;
end
$$ language 'plpgsql';
select sp_cmplist();
this is just a dummy script to select from an imaginary users table but how would I use this script with a cursor or loop to make sure all results are returned?
This code works:
CREATE TABLE foo(a int);
INSERT INTO foo VALUES(10),(20);
CREATE OR REPLACE FUNCTION retfoo()
RETURNS SETOF foo AS $$
BEGIN
RETURN QUERY SELECT * FROM foo;
RETURN;
END;
$$ LANGUAGE plpgsql;
postgres=# SELECT * FROM retfoo();
┌────┐
│ a │
├────┤
│ 10 │
│ 20 │
└────┘
(2 rows)
Time: 1.143 ms
I may have answered my own question with the following
drop type if exists tp_users cascade;
drop function if exists sp_cmplist();
create type tp_users as (
us_id text,
us_status text,
lv_nothing text,
lv_cnt int
);
create function sp_cmplist()
returns setof tp_users as $$
declare
lr_users tp_users;
lv_cnt int;
begin
lv_cnt := 0;
for lr_users in
select users.us_id, users.us_status
from users
loop
-- increment this counter for testing purposes
lv_cnt := lv_cnt + 1;
lr_users.lv_nothing := 'yupy';
lr_users.lv_cnt := lv_cnt;
return next lr_users;
end loop;
return;
end
$$ language 'plpgsql';
select * from sp_cmplist();
this seems to work perfectly
My function's job is to extract XML nodes. The code is as follows:
CREATE TYPE xml_node_looper_record AS (
allomany xml,
i integer,
actual_node text,
nodi_parts text[]
);
CREATE OR REPLACE FUNCTION ds.xml_node_looper_rec(rec xml_node_looper_record)
RETURNS SETOF xml_node_looper_record AS
$BODY$
DECLARE
nodes text[];
field_val text;
r xml_node_looper_record;
n integer;
BEGIN
nodes = xpath(rec.actual_node, rec.allomany);
IF nodes[1] IS NOT NULL THEN
rec.i = rec.i + 1;
FOR n IN 1..array_upper(nodes, 1) LOOP
IF rec.i = array_upper(rec.nodi_parts, 1) THEN
field_val = trim(ARRAY[xpath(rec.actual_node || '/text()', rec.allomany)]::text, ' {}"');
IF field_val IS NOT NULL AND field_val != '' THEN
RAISE NOTICE '% % % %', n, rec.actual_node, rec.i, field_val;
RETURN NEXT (NULL::xml, rec.i, rec.actual_node, ARRAY[field_val]::text[]);
END IF;
END IF;
SELECT ds.xml_node_looper_rec((rec.allomany, rec.i, rec.actual_node || '[' || n::text || ']' || rec.nodi_parts[rec.i + 1], rec.nodi_parts)) INTO r;
END LOOP;
END IF;
END;
$BODY$
LANGUAGE plpgsql VOLATILE COST 100;
As you can see the function is recursive and the goal is to collect field values from multiple XML nodes where we have no information, how many nodes we have under a tag. (I have a version with return table, but that method is too slow.) I use my own defined custom type, and when I check the return values with RAISE NOTICE, I can see the result in pgAdmin on the Messages tab, but the RETURN NEXT command returns nothing, only an empty table.
The parameters of my type:
allomany: an XML data
i: actual depth of nodi_parts
actual_node: XML node, I would like to extract. Can have multiple
nodes, I mark them with []. For example:
/tagone/tagtwo[]/tagthree[]/fieldname
nodi_parts: coming from actual_node, splitting it with []. For
example ARRAY["/tagone/tagtwo", "/tagthree", "/fieldname"]
What is the problem?
You don't propagate the result of nested calls. RETURN NEXT pushes a result to the stack related to the function call. But this stack is private - if caller doesn't fetch this stack, then the result is cleaned. Anyway - any function instance (the called function) has its own result stack. This stack is not shared.
The recursive table function in PL/pgSQL should to look like:
postgres=# CREATE OR REPLACE FUNCTION foo(level int) RETURNS SETOF int AS $$
BEGIN
IF level > 5 THEN RETURN; END IF;
RETURN NEXT level;
--!! must to take result of nested call
RETURN QUERY SELECT * FROM foo(level + 1);
RETURN;
END;
$$ LANGUAGE plpgsql;
postgres=# SELECT * FROM foo(1);
┌─────┐
│ foo │
╞═════╡
│ 1 │
│ 2 │
│ 3 │
│ 4 │
│ 5 │
└─────┘
(5 rows)
Your code is a equivalent of code:
postgres=# CREATE OR REPLACE FUNCTION foo(level int) RETURNS SETOF int AS $$
BEGIN
IF level > 5 THEN RETURN; END IF;
RETURN NEXT level;
-- error, only call of nested function, but returned table is lost
PERFORM foo(level + 1);
RETURN;
END;
$$ LANGUAGE plpgsql;
postgres=# SELECT * FROM foo(1);
┌─────┐
│ foo │
╞═════╡
│ 1 │
└─────┘
(1 row)