Drop table if it exists with DB2/400 SQL - sql

My goal is pretty straightforward - if table has rows, drop it.
Despite the fact that currently there are several similar answers none of them worked for me.
DB2 Drop table if exists equivalent
Suggested solution:
IF EXISTS (SELECT name FROM sysibm.systables WHERE name = 'mylib.mytable') THEN
DROP TABLE mylib.mytable;END IF;
Result:
SQL State: 42601 Vendor Code: -199 Message: [SQL0199] Keyword IF not expected.
Valid tokens: ( CL END GET SET CALL DROP FREE HOLD LOCK OPEN WITH ALTER BEGIN
Drop DB2 table if exists
Suggested solution:
--#SET TERMINATOR #
begin
declare statement varchar(128);
declare continue handle for sqlstate '42710' BEGIN END;
SET STATEMENT = 'DROP TABLE MYLIB.MYTABLE';
EXECUTE IMMEDIATE STATEMENT;
end #
Result:
Message: [SQL0104] Token HANDLE was not invalid. Valid tokens: HANDLER or, if replace handle with handler:
Message: [SQL0199] Keyword STATEMENT not expected. Valid tokens: SQL PATH RESULT SCHEMA CURRENT CONNECTION DESCRIPTOR.
From answer about views
Suggested solution:
DROP TABLE MY_TABLE ONLY IF EXISTS source.
Result:
Message: [SQL0104] Token ONLY was not invalid. Valid tokens: RESTRICT CASCADE
So, I wonder if an alternate solution exists. CL solution is also interesting.

I'm assuming you may want to do this more than once, so a procedure might be in order.
CREATE or replace PROCEDURE DROP_LIVE_TABLE
(in #table varchar(10)
,in #library varchar(10)
)
BEGIN
declare #stmt varchar(100);
declare #cnt int;
IF exists( select * from systables
where sys_dname = #library
and sys_tname = #table
and table_type in ('P','T')
) THEN
SELECT int(sum(number_rows))
INTO #cnt
from SYSTABLESTAT
where sys_dname = #library
and sys_tname = #table
;
IF #cnt > 0 THEN
set #stmt = 'DROP TABLE '||#library||'.'||#table||' CASCADE';
execute immediate #stmt;
END IF;
END IF;
RETURN;
END;
The CASCADE keyword causes any dependent objects such as indexes, logical files, views, or such to be deleted as well.

Here is a CL answer to this question:
PGM PARM(&FILENAME)
DCL VAR(&FILENAME) TYPE(*CHAR) LEN(10)
DCL VAR(&NUMRECS) TYPE(*DEC) LEN(10 0)
RTVMBRD FILE(&FILENAME) NBRCURRCD(&NUMRECS)
IF COND(&NUMRECS > 0) THEN(DLTF +
FILE(&FILENAME))
OUT: ENDPGM
This solution would have trouble if the physical file has dependencies such as indexes or logical files. Those dependencies would have to be deleted first.
The solution by #danny117 on the other hand does not work in all environments. For example I was unable to coerce it to work in SQuirreL client. But it does work in i Navigator. It also works in RUNSQLSTM, but I was unable to determine how to make it work with unqualified table references. If the tables are unqualified, RUNSQLSTM uses the default collection from DFTRDBCOL. The CURRENT_SCHEMA special register does not return the value from DFTRDBCOL.
Here is the if table has rows drop it solution using a compound statement:
begin
if( exists(
select 1 from qsys2.systables
where table_schema = 'MYLIB'
and table_name = 'MYTABLE'
)) then
if( exists(
select 1 from mylib.mytable
)) then
drop table mylib.mytable;
end if;
end if;
end;
I am guessing at the reason you would want to do this, but if it is to allow creation of a new table, then best way may be with a CREATE OR REPLACE TABLE if you are at IBM i v7.2 or greater.
If all you want to do is make sure you have an empty table, TRUNCATE (v7.2+) or DELETE may be better options.

Drop table if exists using atomic statement.
BEGIN ATOMIC
IF( EXISTS(
SELECT 1 FROM TABLES
WHERE TABLE_SCHEMA = 'MYLIB'
AND TABLE_NAME = 'MYTABLE'
)) THEN
DROP TABLE MYLIB/MYTABLE;
END IF;
END;

try this:
BEGIN
IF EXISTS (SELECT NAME FROM QSYS2.SYSTABLES WHERE TABLE_SCHEMA = 'YOURLIBINUPPER' AND TABLE_NAME = 'YOURTABLEINUPPER') THEN
DROP TABLE YOURLIB.YOURTABLE;
END IF;
END ;

Related

Oracle SQL only drop a column if a table exists

For Microsoft SQL Server I have following statement to only drop a column if the table exist.
IF EXISTS(SELECT 1
FROM INFORMATION_SCHEMA.COLUMNS
WHERE TABLE_NAME = 'TEST_TABLE')
ALTER TABLE TEST_TABLE DROP COLUMN LEGACY_VALUE
GO
I was wondering if there was a related IF-EXISTS mechanism is present in Oracle.
All the metadata about the columns in Oracle Database is accessible using one of the following views.
user_tab_cols; -- For all tables owned by the user
all_tab_cols ; -- For all tables accessible to the user
dba_tab_cols; -- For all tables in the Database.
So, if you are looking for a column that exists and want to drop it, your code may look something like this ( see below).
Since this appears to be a one time task, is the effort really worth it?
DECLARE
v_column_exists number := 0;
BEGIN
Select count(*) into v_column_exists
from user_tab_cols
where upper(column_name) = 'LEGACY_VALUE''
and upper(table_name) = 'TEST_TABLE';
--and owner = 'SCOTT --*might be required if you are using all/dba views
if (v_column_exists = 1) then
execute immediate 'alter table test_table drop column legacy_value';
end if;
end;
/

Redshift Alter table if not exists

I'm trying add a new field in a redshift table. But I want to add only if this field doesn't exists.
I tried wrapping it with IF NOT EXISTS. But I got following error:
Amazon](500310) Invalid operation: syntax error at or near "IF" Position: 5;
BEGIN
IF NOT EXISTS (SELECT 1 FROM information_schema.columns WHERE table_schema = 'schema_name' and table_name='table_name' and column_name='new_field') THEN
ALTER TABLE schema_name.table_name
ADD new_field INT;
END IF;
COMMIT;
I'm not sure if I'm correctly using "IF NOT EXISTS" statement inside the BEGIN block.
Can someone please help me?
Thanks in advance!
It could be better to handle it using EXCEPTION
BEGIN
ALTER TABLE
<table_name> ADD COLUMN <column_name> <column_type>;
EXCEPTION
WHEN duplicate_column
THEN RAISE NOTICE 'column <column_name> already exists in <table_name>.';
END;
The Answer by Yauheni Khvainitski is not completely wrong. But you do have to use a SP and the only option Redshit has (at this point is to have "EXCEPTION WHEN OTHER"). An example:
CREATE OR REPLACE PROCEDURE change_column_to_big_int_TABLE_NAME_X(column_name varchar(200)) AS
$$
DECLARE
new_column_name VARCHAR;
BEGIN
SELECT INTO new_column_name (table_name)||'_new';
-- RAISE INFO 'new_table_name = % table_name = %',new_column_name, table_name;
ALTER TABLE TABLE_NAME_X ADD COLUMN "(new_column_name)" bigint;
EXCEPTION WHEN OTHERS
THEN RAISE NOTICE 'column already exists on table';
END;
$$
LANGUAGE plpgsql;
CALL change_column_to_big_int_TABLE_NAME_X('COLUMN_Y');
Some links from AWS on:
trapping errors: https://docs.aws.amazon.com/redshift/latest/dg/stored-procedure-trapping-errors.html
CREATE PROCEDURE: https://docs.aws.amazon.com/redshift/latest/dg/r_CREATE_PROCEDURE.html
Also please notice that this is valid at this point in time. Redshift seems to be always evolving.
The issue I think is that AWS Redshift does not support the IF statement, but instead uses CASE statements. The CASE statements are very similar to IF the way they implement them. But I admit, I prefer the IF statements.

Executing a PostgreSQL query with an EXCEPTION results in two different ERROR messages

I have an PostgreSQL query that includes a transaction and an exception if a column is duplicated:
BEGIN;
ALTER TABLE "public"."cars"
ADD COLUMN "top_speed" text;
EXCEPTION WHEN duplicate_column THEN NOTHING;
ROLLBACK;
In this query I am trying to add a column that already exists (playing a little bit with exceptions) and if it does then the query shall just ignore it. At the moment I am not really sure if the exception-code I am using is the right (couldn't find a site where they are described; only found this)
My Problem is if I execute this query I get the error-message:
ERROR: column "top_speed" of relation "cars" already exists
And if I execute it a second time the error-message changes to:
ERROR: current transaction is aborted, commands ignored until end of transaction block
Try an anonymous code block. As Laurenz mentioned, you were mixing PL/pgSQL and SQL commands.
Sample table
CREATE TABLE t (f1 TEXT);
Anonymous code block
DO $$
BEGIN
IF (SELECT count(column_name) FROM information_schema.columns
WHERE table_schema = 'public' AND
table_name = 't' AND
column_name = 'f2') = 0 THEN
ALTER TABLE public.t ADD COLUMN "f2" text;
END IF;
END$$;
After execution you have your new column. If the column already exists, it will do nothing.
SELECT * FROM t;
f1 | f2
----+----
0 Zeilen
In PostgreSQL 9.6+ you can use IF NOT EXISTS to check if a given column already exists in the table before creating it:
ALTER TABLE t ADD COLUMN IF NOT EXISTS f2 TEXT;
Code at db<>fiddle

Create or replace table in Oracle pl/sql

I need a script which creates table or if it already exist drops it, and when recreates table. After some research I have found out that CREATE OR REPLACE TABLE in pl/sql doesn't exist. So I come up with this script :
DECLARE
does_not_exist EXCEPTION;
PRAGMA EXCEPTION_INIT (does_not_exist, -942);
BEGIN
EXECUTE IMMEDIATE 'DROP TABLE foobar';
EXCEPTION
WHEN does_not_exist
THEN
NULL;
END;
/
CREATE TABLE foobar (c1 INT);
Is there any proper way to achieve this functionality?
You really shouldn't be doing this in PL/SQL, tables created at runtime would be indicative of a flaw in your data model. If you're really convinced you absolutely have to do this then investigate temporary tables first. Personally, I'd reassess whether it's necessary at all.
You seem to be going for the EAFP as opposed to LBYL approach, which is described in a few answers to this question. I would argue that this is unnecessary. A table is a fairly static beast, you can use the system view USER_TABLES to determine whether it exists before dropping it.
declare
l_ct number;
begin
-- Determine if the table exists.
select count(*) into l_ct
from user_tables
where table_name = 'THE_TABLE';
-- Drop the table if it exists.
if l_ct = 1 then
execute immediate 'drop table the_table';
end if;
-- Create the new table it either didn-t exist or
-- has been dropped so any exceptions are exceptional.
execute immediate 'create table the_table ( ... )';
end;
/
Using a global temporary table would seem to be a better option. However, if you insist on dropping and re-adding tables at runtime you could query one of the _TABLES views (i.e. USER_TABLES, DBA_TABLES, ALL_TABLES) to determine if the table exists, drop it if it does, then create it:
SELECT COUNT(*)
INTO nCount
FROM USER_TABLES
WHERE TABLE_NAME = 'FOOBAR';
IF nCount <> 0 THEN
EXECUTE IMMEDIATE 'DROP TABLE FOOBAR';
END IF;
EXECUTE IMMEDIATE 'CREATE TABLE FOOBAR(...)';
Share and enjoy.

How do I check if a temporary table exists in SQL Anywhere?

I want to write a SQL IF statement that checks whether or not a local temporary table exists, but those kinds of tables are not recorded in the SQL Anywhere system catalog.
Note that you can do this in 11.0.1 and higher:
DROP TABLE IF EXISTS t;
If you're asking the question, "How do I drop a local temporary table without raising an error if it doesn't exist?" then the answer's simple: just DROP it and ignore any error:
BEGIN
DROP TABLE t;
EXCEPTION WHEN OTHERS THEN
END;
If you really need to know the answer to the question "Does table t exist?" you can query the table and analyze the resulting SQLSTATE. The following function makes use of several features:
ON EXCEPTION RESUME ignores any exception raised by the SELECT and passes control to the IF statement.
EXECUTE IMMEDIATE lets you write a query where the table name is in a string variable.
TOP 1 makes sure that only one row is selected even if the table contains a million rows.
ORDER BY 1 lets you meet the requirement that TOP can only be used when the result set is ordered.
SELECT 1 relieves you of the burden of specifying a column name.
INTO #dummy means the SELECT (and consequently the EXECUTE IMMEDIATE) doesn't return a result set.
If the SELECT works, it's either going to set SQLSTATE to '00000' for success or '02000' for row not found. Any other SQLSTATE means there's some serious problem with the table... like it doesn't exist.
CREATE FUNCTION f_table_is_ok
( IN #table_name VARCHAR ( 128 ) )
RETURNS INTEGER
ON EXCEPTION RESUME
BEGIN
DECLARE #dummy INTEGER;
EXECUTE IMMEDIATE STRING (
'SELECT TOP 1 1 INTO #dummy FROM ',
#table_name,
' ORDER BY 1' );
IF SQLSTATE IN ( '00000', '02000' ) THEN
RETURN 1
ELSE
RETURN 0
END IF;
END;
Here's some test code:
BEGIN
DECLARE LOCAL TEMPORARY TABLE tt ( c INTEGER );
DECLARE LOCAL TEMPORARY TABLE "t t" ( c INTEGER );
SELECT f_table_is_ok ( 'asdf' );
SELECT f_table_is_ok ( 'tt' );
SELECT f_table_is_ok ( '"t t"' );
SELECT f_table_is_ok ( '"SYS"."SYSTABLE"' );
END;
just try to drop it anyways and ignore the error...
BEGIN
DROP TABLE table;
EXCEPTION WHEN OTHERS THEN
END;