I have a SQL script that has to be run every time a client executes the "database management" functionality. The script includes creating stored procedures on the client database. Some of these clients might already have the stored procedure upon running the script, and some may not. I need to have the missing stored procedures added to the client database, but it doesn't matter how much I try to bend T-SQL syntax, I get
CREATE/ALTER PROCEDURE' must be the first statement in a query batch
I've read that dropping before creating works, but I don't like doing it that way.
IF EXISTS (SELECT * FROM sys.objects WHERE type = 'P' AND name = 'MyProc')
DROP PROCEDURE MyProc
GO
CREATE PROCEDURE MyProc
...
How can I add check for the existence of a stored procedure and create it if it doesn't exist but alter it if it does exist?
I realize this has already been marked as answered, but we used to do it like this:
IF NOT EXISTS (SELECT * FROM sys.objects WHERE type = 'P' AND OBJECT_ID = OBJECT_ID('dbo.MyProc'))
exec('CREATE PROCEDURE [dbo].[MyProc] AS BEGIN SET NOCOUNT ON; END')
GO
ALTER PROCEDURE [dbo].[MyProc]
AS
....
Just to avoid dropping the procedure.
You can run procedural code anywhere you are able to run a query.
Just copy everything after AS
:
BEGIN
DECLARE @myvar INT
SELECT *
FROM mytable
WHERE @myvar ...
END
This code does exactly same things a stored proc would do, but is not stored on the database side.
That's much like what is called anonymous procedure in PL/SQL
.
Update:
Your question title is a little bit confusing.
If you only need to create a procedure if it not exists, then your code is just fine.
Here's what SSMS
outputs in the create script:
IF EXISTS ( SELECT *
FROM sys.objects
WHERE object_id = OBJECT_ID(N'myproc')
AND type IN ( N'P', N'PC' ) )
DROP …
CREATE …
Update:
Example of how to do it when including the schema:
IF EXISTS ( SELECT *
FROM sysobjects
WHERE id = object_id(N'[dbo].[MyProc]')
and OBJECTPROPERTY(id, N'IsProcedure') = 1 )
BEGIN
DROP PROCEDURE [dbo].[MyProc]
END
In the example above, dbo is the schema.
Update:
In SQL Server 2016+, you can just do
CREATE OR ALTER PROCEDURE dbo.MyProc
@astander
: you can call anonymous code from the stored procedures as well. To use their output in an INSERT
, you'll need to use OPENROWSET
or OPENQUERY
which works with the anonymous code as well. Of course there are drawbacks in the anonymous code: for instance, it only runs under the caller's privileges. My point is that it is possible, not preferred way of doing things :)
If you're looking for the simplest way to check for a database object's existence before removing it, here's one way (example uses a SPROC, just like your example above but could be modified for tables, indexes, etc...):
IF (OBJECT_ID('MyProcedure') IS NOT NULL)
DROP PROCEDURE MyProcedure
GO
This is quick and elegant, but you need to make sure you have unique object names across all object types since it does not take that into account.
I Hope this helps!
I know you want to "ALTER a procedure if it exists and create it if it does not exist", but I believe it is simpler to:
Drop the procedure (if it already exists) and then Re-create it.
Like this:
IF OBJECT_ID('MyProcedure', 'P') IS NOT NULL
DROP PROCEDURE MyProcedure
GO
CREATE PROCEDURE MyProcedure AS
BEGIN
/* ..... */
END
GO
The second parameter tells OBJECT_ID
to only look for objects with object_type = 'P'
, which are stored procedures:
AF = Aggregate function (CLR) C = CHECK constraint D = DEFAULT (constraint or stand-alone) F = FOREIGN KEY constraint FN = SQL scalar function FS = Assembly (CLR) scalar-function FT = Assembly (CLR) table-valued function IF = SQL inline table-valued function IT = Internal table P = SQL Stored Procedure PC = Assembly (CLR) stored-procedure PG = Plan guide PK = PRIMARY KEY constraint R = Rule (old-style, stand-alone) RF = Replication-filter-procedure S = System base table SN = Synonym SO = Sequence object TF = SQL table-valued-function TR = Trigger
You can get the full list of options via:
SELECT name
FROM master..spt_values
WHERE type = 'O9T'
As of SQL SERVER 2016 you can use the new DROP PROCEDURE IF EXISTS
.
DROP { PROC | PROCEDURE } [ IF EXISTS ] { [ schema_name. ] procedure } [ ,...n ]
Reference : https://msdn.microsoft.com/en-us/library/ms174969.aspx
I know it is a very old post, but since this appears in the top search results hence adding the latest update for those using SQL Server 2016 SP1 -
create or alter procedure procTest
as
begin
print (1)
end;
go
This creates a Stored Procedure if does not already exist, but alters it if exists.
DROP IF EXISTS is a new feature of SQL Server 2016
DROP PROCEDURE IF EXISTS dbo.[procname]
I had the same error. I know this thread is pretty much dead already but I want to set another option besides "anonymous procedure".
I solved it like this:
Check if the stored procedure exist: IF NOT EXISTS (SELECT * FROM sysobjects WHERE name='my_procedure') BEGIN print 'exists' -- or watever you want END ELSE BEGIN print 'doesn''texists' -- or watever you want END However the "CREATE/ALTER PROCEDURE' must be the first statement in a query batch" is still there. I solved it like this: SET ANSI_NULLS ON GO SET QUOTED_IDENTIFIER ON GO CREATE -- view procedure function or anything you want ... I end up with this code: IF EXISTS (SELECT * FROM dbo.sysobjects WHERE id = OBJECT_ID('my_procedure')) BEGIN DROP PROCEDURE my_procedure END SET ANSI_NULLS ON GO SET QUOTED_IDENTIFIER ON GO CREATE PROCEDURE [dbo].my_procedure ...
Here's a method and some reasoning behind using it this way. It isn't as pretty to edit the stored proc but there are pros and cons...
UPDATE: You can also wrap this entire call in a TRANSACTION. Including many stored procedures in a single transaction which can all commit or all rollback. Another advantage of wrapping in a transaction is the stored procedure always exists for other SQL connections as long as they do not use the READ UNCOMMITTED transaction isolation level!
1) To avoid alters just as a process decision. Our processes are to always IF EXISTS DROP THEN CREATE. If you do the same pattern of assuming the new PROC is the desired proc, catering for alters is a bit harder because you would have an IF EXISTS ALTER ELSE CREATE.
2) You have to put CREATE/ALTER as the first call in a batch so you can't wrap a sequence of procedure updates in a transaction outside dynamic SQL. Basically if you want to run a whole stack of procedure updates or roll them all back without restoring a DB backup, this is a way to do everything in a single batch.
IF NOT EXISTS (select ss.name as SchemaName, sp.name as StoredProc
from sys.procedures sp
join sys.schemas ss on sp.schema_id = ss.schema_id
where ss.name = 'dbo' and sp.name = 'MyStoredProc')
BEGIN
DECLARE @sql NVARCHAR(MAX)
-- Not so aesthetically pleasing part. The actual proc definition is stored
-- in our variable and then executed.
SELECT @sql = 'CREATE PROCEDURE [dbo].[MyStoredProc]
(
@MyParam int
)
AS
SELECT @MyParam'
EXEC sp_executesql @sql
END
In Sql server 2008 onwards, you can use "INFORMATION_SCHEMA.ROUTINES
"
IF EXISTS (SELECT 1 FROM INFORMATION_SCHEMA.ROUTINES
WHERE ROUTINE_NAME = 'MySP'
AND ROUTINE_TYPE = 'PROCEDURE')
**The simplest way to drop and recreate a stored proc in T-Sql is **
Use DatabaseName
go
If Object_Id('schema.storedprocname') is not null
begin
drop procedure schema.storedprocname
end
go
create procedure schema.storedprocname
as
begin
end
Here is the script that I use. With it, I avoid unnecessarily dropping and recreating the stored procs.
IF NOT EXISTS (
SELECT *
FROM sys.objects
WHERE object_id = OBJECT_ID(N'[dbo].[uspMyProcedure]')
)
BEGIN
EXEC sp_executesql N'CREATE PROCEDURE [dbo].[uspMyProcedure] AS select 1'
END
GO
ALTER PROCEDURE [dbo].[uspMyProcedure]
@variable1 INTEGER
AS
BEGIN
-- Stored procedure logic
END
why don't you go the simple way like
IF EXISTS(SELECT * FROM sys.procedures WHERE NAME LIKE 'uspBlackListGetAll')
BEGIN
DROP PROCEDURE uspBlackListGetAll
END
GO
CREATE Procedure uspBlackListGetAll
..........
%
, so the LIKE
behaves as an =
In addition to the answer from @Geoff I've created a simple tool which generates a SQL-file which statements for Stored Procedures, Views, Functions and Triggers.
https://i.stack.imgur.com/7rInr.png
I wonder! Why i don't write the whole query like
GO
create procedure [dbo].[spAddNewClass] @ClassName varchar(20),@ClassFee int
as
begin
insert into tblClass values (@ClassName,@ClassFee)
end
GO
create procedure [dbo].[spAddNewSection] @SectionName varchar(20),@ClassID int
as
begin
insert into tblSection values(@SectionName,@ClassID)
end
Go
create procedure test
as
begin
select * from tblstudent
end
i already know that first two procedures are already exist sql will run the query will give the error of first two procedures but still it will create the last procedure SQl is itself taking care of what is already exist this is what i always do to all my clients!
CREATE Procedure IF NOT EXISTS 'Your proc-name' () BEGIN ... END
Success story sharing
GRANT
statements explicity in the script in case they change; so there is still justification to useDROP
instead ofALTER
.SET NOCOUNT ON
a good choice for the temporary stored procedure body? Asked here: stackoverflow.com/questions/40849374/…