BUG #18116: This is definitiv a BUG in INOUT parameter in stored procedures in Postgres 12.16

Started by PG Bug reporting formover 2 years ago2 messagesbugs
Jump to latest
#1PG Bug reporting form
noreply@postgresql.org

The following bug has been logged on the website:

Bug reference: 18116
Logged by: Wolfgang Launhardt
Email address: wlaunhardt@web.de
PostgreSQL version: 12.16
Operating system: Linux Mint 20.3 Cinnamon
Description:

Please don't send me that again - it is a BUG ! :
"
BUG #18106: Stored Procedure (Text-)Array as INOUT parameter

has been rejected by a moderator and will not be posted.
The reason given for rejection was:

This does not appear to be a bug report."

DO
$$
DECLARE
test text[];
BEGIN
test := '{"x","y","z"}';
CALL ptest( test );
RAISE NOTICE '%', test;
END
$$;

PRODUCES:
ERROR: type with OID 0 does not exist
CONTEXT: SQL statement "CALL ptest( test )" PL/pgSQL function
inline_code_block line 6 at CALL

IN A FUNCTION IT WORKS:
create or replace FUNCTION ptest(INOUT text[] ) returns text[] as 'ptest.so'
language 'c';

C-Source:

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "postgres.h"
#include "fmgr.h"
#include "funcapi.h"
#include "array.h"

PG_MODULE_MAGIC;

Datum ptest(PG_FUNCTION_ARGS);

PG_FUNCTION_INFO_V1(ptest);

Datum
ptest ( PG_FUNCTION_ARGS )
{
ArrayType * arg_arr;

arg_arr = PG_GETARG_ARRAYTYPE_P( 0 );

PG_RETURN_ARRAYTYPE_P( arg_arr );

// TEXTOID = 25 the following code produces the same error:
// PG_RETURN_ARRAYTYPE_P( construct_empty_array( 25 ));
}

#2Tom Lane
tgl@sss.pgh.pa.us
In reply to: PG Bug reporting form (#1)
Re: BUG #18116: This is definitiv a BUG in INOUT parameter in stored procedures in Postgres 12.16

PG Bug reporting form <noreply@postgresql.org> writes:

Please don't send me that again - it is a BUG ! :

Maybe, but you haven't shown us a reproducible example.
What I get from the fragments you gave here is

ERROR: ptest(text[]) is not a procedure
LINE 1: CALL ptest( test )
^
HINT: To call a function, use SELECT.
QUERY: CALL ptest( test )
CONTEXT: PL/pgSQL function inline_code_block line 6 at CALL

I suspect that the underlying problem may involve a SQL declaration
for "ptest" that doesn't match what the C code does, so I'm not going
to guess at what you actually used as the SQL declaration. But
clearly it's not what you wrote here.

regards, tom lane