[sql] How to add a Try/Catch to SQL Stored Procedure

CREATE PROCEDURE [dbo].[PL_GEN_PROVN_NO1]        
@GAD_COMP_CODE  VARCHAR(2) =NULL, 
@@voucher_no numeric =null output 
AS         
BEGIN  
    DECLARE @NUM NUMERIC 
    DECLARE @PNO  NUMERIC                               
    SET @PNO = 0 
    DECLARE @PNO1 NUMERIC
    SET @PNO1=0 

--  begin transaction 

    IF NOT EXISTS (select GLDC_NEXT_PRV_NO
               FROM   GLAS_FINANCIAL_DOCUMENTS          
                   WHERE  GLDC_COMP_CODE  = @GAD_COMP_CODE        
                   AND GLDC_DOC_CODE  = 'JV' )
    BEGIN
               RAISERROR ('Error in generating provision number..',16,1) 
               -- ROLLBACK TRANSACTION
    END
ELSE
SELECT @PNO=ISNULL(GLDC_NEXT_PRV_NO,0)+1
FROM   GLAS_FINANCIAL_DOCUMENTS          
WHERE  GLDC_COMP_CODE  = @GAD_COMP_CODE        
AND GLDC_DOC_CODE  = 'JV' 

UPDATE  GLAS_FINANCIAL_DOCUMENTS        
SET GLDC_NEXT_PRV_NO = @PNO         
WHERE  GLDC_COMP_CODE  = @GAD_COMP_CODE        
AND GLDC_DOC_CODE  = 'JV' 

set @@VOUCHER_NO=@PNO    
--commit transaction 
END

In this proc how can I handle try catch for exception?

Create Proc[usp_mquestions]  
( 
 @title  nvarchar(500),   --0
 @tags  nvarchar(max),   --1
 @category  nvarchar(200),   --2
 @ispoll  char(1),   --3
 @descriptions  nvarchar(max),   --4
)              
 AS  
 BEGIN TRY




BEGIN
DECLARE @message varchar(1000); 
DECLARE @tempid bigint; 

IF((SELECT count(id) from  [xyz] WHERE title=@title)>0)
BEGIN
SELECT 'record already existed.';
END
ELSE
BEGIN               


if @id=0 
begin 
select @tempid =id from [xyz] where id=@id;

if @tempid is null 
BEGIN 
        INSERT INTO xyz
        (entrydate,updatedate)
        VALUES
        (GETDATE(),GETDATE())

        SET @tempid=@@IDENTITY;
 END 
END 
ELSE 
BEGIN 
set @tempid=@id 
END 
if @tempid>0 
BEGIN 

    -- Updation of table begin--


UPDATE  tab_questions
set title=@title, --0 
 tags=@tags, --1 
 category=@category, --2 
 ispoll=@ispoll, --3 
 descriptions=@descriptions, --4 
 status=@status, --5

WHERE id=@tempid ; --9 ;


IF @id=0 
BEGIN 
SET @message= 'success:Record added successfully:'+ convert(varchar(10), @tempid)
END 
ELSE 
BEGIN 
SET @message= 'success:Record updated successfully.:'+ convert(varchar(10), @tempid)

END 
END 
ELSE 
BEGIN 
SET @message= 'failed:invalid request:'+convert(varchar(10), @tempid)
END 

END
END

END TRY
BEGIN CATCH
    SET @message='failed:'+ ERROR_MESSAGE();
END CATCH
SELECT @message;

Error-Handling with SQL Stored Procedures

TRY/CATCH error handling can take place either within or outside of a procedure (or both). The examples below demonstrate error handling in both cases.

If you want to experiment further, you can fork the query on Stack Exchange Data Explorer.

_{(This uses a temporary stored procedure... we can't create regular SP's on SEDE, but the functionality is the same.)}

--our Stored Procedure
create procedure #myProc as --we can only create #temporary stored procedures on SEDE. 
  begin
    BEGIN TRY
      print 'This is our Stored Procedure.'
      print 1/0                          --<-- generate a "Divide By Zero" error.
      print 'We are not going to make it to this line.'
    END TRY

    BEGIN CATCH
      print 'This is the CATCH block within our Stored Procedure:'
          + ' Error Line #'+convert(varchar,ERROR_LINE())
          + ' of procedure '+isnull(ERROR_PROCEDURE(),'(Main)')
      --print 1/0                        --<-- generate another "Divide By Zero" error.
        -- uncomment the line above to cause error within the CATCH ¹ 
    END CATCH
  end
go

--our MAIN code block:
BEGIN TRY
  print 'This is our MAIN Procedure.'
  execute #myProc  --execute the Stored Procedure
      --print 1/0                        --<-- generate another "Divide By Zero" error.
        -- uncomment the line above to cause error within the MAIN Procedure ²
  print 'Now our MAIN sql code block continues.'
END TRY

BEGIN CATCH
  print 'This is the CATCH block for our MAIN sql code block:'
          + ' Error Line #'+convert(varchar,ERROR_LINE())
          + ' of procedure '+isnull(ERROR_PROCEDURE(),'(Main)')
END CATCH

Here's the result of running the above sql as-is:

This is our MAIN Procedure.
This is our Stored Procedure.
This is the CATCH block within our Stored Procedure: Error Line #5 of procedure #myProc
Now our MAIN sql code block continues.

¹ Uncommenting the "additional error line" from the Stored Procedure's CATCH block will produce:

This is our MAIN procedure.
This is our Stored Procedure.
This is the CATCH block within our Stored Procedure: Error Line #5 of procedure #myProc
This is the CATCH block for our MAIN sql code block: Error Line #13 of procedure #myProc

² Uncommenting the "additional error line" from the MAIN procedure will produce:

This is our MAIN Procedure.
This is our Stored Pprocedure.
This is the CATCH block within our Stored Procedure: Error Line #5 of procedure #myProc
This is the CATCH block for our MAIN sql code block: Error Line #4 of procedure (Main)

Use a single procedure for error handling

On topic of stored procedures and error handling, it can be helpful (and tidier) to use a single, dynamic, stored procedure to handle errors for multiple other procedures or code sections.

Here's an example:

--our error handling procedure
create procedure #myErrorHandling as
  begin
    print ' Error #'+convert(varchar,ERROR_NUMBER())+': '+ERROR_MESSAGE()  
    print ' occurred on line #'+convert(varchar,ERROR_LINE())
         +' of procedure '+isnull(ERROR_PROCEDURE(),'(Main)')
    if ERROR_PROCEDURE() is null       --check if error was in MAIN Procedure
      print '*Execution cannot continue after an error in the MAIN Procedure.'
  end
go

create procedure #myProc as     --our test Stored Procedure
  begin
    BEGIN TRY
      print 'This is our Stored Procedure.'
      print 1/0                       --generate a "Divide By Zero" error.
      print 'We will not make it to this line.'
    END TRY
    BEGIN CATCH
     execute #myErrorHandling
    END CATCH
  end
go

BEGIN TRY                       --our MAIN Procedure
  print 'This is our MAIN Procedure.'
  execute #myProc                     --execute the Stored Procedure
  print '*The error halted the procedure, but our MAIN code can continue.'
  print 1/0                           --generate another "Divide By Zero" error.
  print 'We will not make it to this line.'
END TRY
BEGIN CATCH
  execute #myErrorHandling
END CATCH

Example Output: _{^{(This query can be forked on SEDE here.)}}

This is our MAIN procedure.
This is our stored procedure.
 Error #8134: Divide by zero error encountered.
 occurred on line #5 of procedure #myProc
*The error halted the procedure, but our MAIN code can continue.
 Error #8134: Divide by zero error encountered.
 occurred on line #5 of procedure (Main)
*Execution cannot continue after an error in the MAIN procedure.

Documentation:

In the scope of a TRY/CATCH block, the following system functions can be used to obtain information about the error that caused the CATCH block to be executed:

• ERROR_NUMBER() returns the number of the error.
• ERROR_SEVERITY() returns the severity.
• ERROR_STATE() returns the error state number.
• ERROR_PROCEDURE() returns the name of the stored procedure or trigger where the error occurred.
• ERROR_LINE() returns the line number inside the routine that caused the error.
• ERROR_MESSAGE() returns the complete text of the error message. The text includes the values supplied for any substitutable parameters, such as lengths, object names, or times.

^(Source)

Note that there are two types of SQL errors: Terminal and Catchable. TRY/CATCH will [obviously] only catch the "Catchable" errors. This is one of a number of ways of learning more about your SQL errors, but it probably the most useful.

It's "better to fail now" (during development) compared to later because, as Homer says . . .

Transact-SQL is a bit more tricky that C# or C++ try/catch blocks, because of the added complexity of transactions. A CATCH block has to check the xact_state() function and decide whether it can commit or has to rollback. I have covered the topic in my blog and I have an article that shows how to correctly handle transactions in with a try catch block, including possible nested transactions: Exception handling and nested transactions.

create procedure [usp_my_procedure_name]
as
begin
    set nocount on;
    declare @trancount int;
    set @trancount = @@trancount;
    begin try
        if @trancount = 0
            begin transaction
        else
            save transaction usp_my_procedure_name;

        -- Do the actual work here

lbexit:
        if @trancount = 0   
            commit;
    end try
    begin catch
        declare @error int, @message varchar(4000), @xstate int;
        select @error = ERROR_NUMBER(),
                 @message = ERROR_MESSAGE(), @xstate = XACT_STATE();
        if @xstate = -1
            rollback;
        if @xstate = 1 and @trancount = 0
            rollback
        if @xstate = 1 and @trancount > 0
            rollback transaction usp_my_procedure_name;

        raiserror ('usp_my_procedure_name: %d: %s', 16, 1, @error, @message) ;
        return;
    end catch   
end

yep - you can even nest the try catch statements as:

BEGIN TRY
SET @myFixDte = CONVERT(datetime, @myFixDteStr,101)
END TRY
BEGIN CATCH
    BEGIN TRY
        SET @myFixDte = CONVERT(datetime, @myFixDteStr,103)
END TRY
BEGIN CATCH
    BEGIN TRY
        SET @myFixDte = CONVERT(datetime, @myFixDteStr,104)
    END TRY
    BEGIN CATCH
        SET @myFixDte = CONVERT(datetime, @myFixDteStr,105)
    END CATCH
END CATCH END CATCH