Prevent Duplicates from Inserting from File in Stored Procedure

Question

CREATE TABLE dbo.Logins
[ID] numeric(10, 0) NOT NULL Primary Key,
[Dr_FName] varchar(50) NOT NULL,
[Dr_LName] varchar(50) NOT NULL,
[Login] varchar(50) NOT NULL
GO
CREATE TABLE [dbo].[CRIS_USER] (
[id] numeric(10, 0) NOT NULL Primary Key,
[login_id] varchar(20) NOT NULL
GO
CREATE TABLE [dbo].[CRIS_SYSTEM_USER] (
[id] numeric(10, 0) NOT NULL Primary Key,
[cris_user_id] numeric(10, 0) NOT NULL,
INSERT INTO Logins
(ID, Dr_FName, Dr_LName,Login)
VALUES(1,'Lisa','Mars','lmars')
INSERT INTO Logins
(ID, Dr_FName, Dr_LName,Login)
VALUES(2,'Becky','Saturn','bsaturn')
INSERT INTO Logins
(ID, Dr_FName, Dr_LName,Login)
VALUES(3,'Mary','Venus','mvenus')
INSERT INTO CRIS_USER
(ID,login_id)
VALUES(10, 'lmars')
INSERT INTO CRIS_USER
(ID,login_id)
VALUES(20, 'bsaturn')
INSERT INTO CRIS_USER
(ID,login_id)
VALUES(30, 'mvenus')
INSERT INTO CRIS_SYSTEM_USER
(ID,cris_user_id)
VALUES(110, 10)
INSERT INTO CRIS_SYSTEM_USER
(ID,cris_user_id)
VALUES(120,20)
INSERT INTO CRIS_SYSTEM_USER
(ID,cris_user_id)
VALUES(130, 30)
I'm aware that "ID" is a bad column name and that it should not be numeric. The ID columns are incremented by a program. They are not auto incremented. I didn't design it.
I have a stored procedure that does a bulk insert from a tab delimited file into the three tables:
CREATE PROCEDURE [dbo].[InsertUserText]
WITH EXEC AS CALLER
AS
IF OBJECT_ID('TEMPDB..#LoginTemp') IS NULL
BEGIN
CREATE TABLE #LoginTemp(Login varchar(50),Dr_FName varchar(50),Dr_LName varchar(50))
BULK INSERT #LoginTemp
FROM 'C:\loginstest\InsertUserText.txt'
WITH (ROWTERMINATOR ='\n'
-- New Line Feed (\n) automatically adds Carrige Return (\r)
,FIELDTERMINATOR = '\t'
--delimiter
,FIRSTROW=4)
PRINT 'File data copied to Temp table'
END
DECLARE @maxid NUMERIC(10,0)
DECLARE @maxid2 NUMERIC(10,0)
DECLARE @maxid3 NUMERIC(10,0)
BEGIN TRANSACTION
SELECT @maxid = coalesce(MAX(ID), 0)
FROM dbo.LOGINS WITH (UPDLOCK)
INSERT INTO dbo.LOGINS(ID,Dr_FName,Dr_LName,Login)
SELECT row_number() OVER(ORDER BY (SELECT NULL)) + @maxid,
Dr_FName,Dr_LName,Login
FROM #LoginTemp
WHERE #LoginTemp.Dr_FName is not Null;
SELECT @maxid3 = coalesce(MAX(id), 0)
FROM dbo.CRIS_USER WITH (UPDLOCK)
INSERT INTO dbo.CRIS_USER(id,login_id)
SELECT row_number() OVER(ORDER BY (SELECT NULL)) + @maxid3,
Login
FROM #LoginTemp
WHERE #LoginTemp.Dr_FName is not Null;
SELECT @maxid2 = coalesce(MAX(id), 0)
FROM dbo.CRIS_SYSTEM_USER WITH (UPDLOCK)
INSERT INTO dbo.CRIS_SYSTEM_USER(id,cris_user_id)
SELECT row_number() OVER(ORDER BY (SELECT NULL)) + @maxid2,
+ row_number() OVER(ORDER BY (SELECT NULL)) + @maxid3
FROM #LoginTemp
WHERE #LoginTemp.Dr_FName is not Null;
PRINT 'Copied from Temp table to CRIS_USER'
COMMIT TRANSACTION
GO
What suggestions do you have to prevent a duplicate Logins.Login? None of the inserts for all three tables should occur if a login already exists in the Logins table. There should be a message to indicate which Login failed. I haven't yet decided if I want
all of the logins in the text file to fail if there is a duplicate, or just the one with the duplicate. I'm open to suggestions on that. So far, the duplicates only occur when someone forgets to update the tabbed delimited file and accidently runs the procedure
on an old one. I'm sure I can come up with an if statement that will accomplish this. I could maybe use WHERE EXISTS or WHERE NOT EXISTS. But I know I can get a good solution here.
I'm also aware that duplicates could be prevented in the table design. Again, I didn't design it.
I have a tab delimited file created but don't see a way to attach it.
Thanks for any help.

Answer 1

Thanks to all three that replied. I meant to mark the question as answered sooner. I've tried all the suggestions on a test system. All will work with maybe some slight variations. Below was my temporary quick fix. I'm working on switching to a permanent
solution based on the replies.
This is not the real solution and is not the answer to my question. It's just temporary.
IF OBJECT_ID('TEMPDB..#LoginTemp') IS NULL
BEGIN
CREATE TABLE #LoginTemp(Login varchar(50),Dr_FName varchar(50),Dr_LName varchar(50))
BULK INSERT #LoginTemp
FROM 'C:\loginstest\InsertUserText.txt'
WITH (ROWTERMINATOR ='\n'
Return (\r)
,FIELDTERMINATOR = '\t'
,FIRSTROW=4)
PRINT 'File data copied to Temp table'
END
DECLARE @maxid NUMERIC(10,0)
DECLARE @maxid2 NUMERIC(10,0)
DECLARE @maxid3 NUMERIC(10,0)
IF EXISTS(SELECT 'True' FROM Logins L INNER JOIN #LoginTemp LT on L.Login = LT.Login
BEGIN
SELECT 'Duplicate row!'
END
ELSE
BEGIN
BEGIN TRANSACTION
SELECT @maxid = coalesce(MAX(ID), 0)
FROM dbo.LOGINS WITH (UPDLOCK)
INSERT INTO dbo.LOGINS(ID,Dr_FName,Dr_LName,Login)
SELECT row_number() OVER(ORDER BY (SELECT NULL)) + @maxid,
Dr_FName,Dr_LName,Login
FROM #LoginTemp
WHERE #LoginTemp.Dr_FName is not Null;
SELECT @maxid3 = coalesce(MAX(id), 0)
FROM dbo.CRIS_USER WITH (UPDLOCK)
INSERT INTO dbo.CRIS_USER(id,login_id)
SELECT row_number() OVER(ORDER BY (SELECT NULL)) + @maxid3,
Login
FROM #LoginTemp
WHERE #LoginTemp.Dr_FName is not Null;
SELECT @maxid2 = coalesce(MAX(id), 0)
FROM dbo.CRIS_SYSTEM_USER WITH (UPDLOCK)
INSERT INTO dbo.CRIS_SYSTEM_USER(id,cris_user_id)
SELECT row_number() OVER(ORDER BY (SELECT NULL)) + @maxid2,
+ row_number() OVER(ORDER BY (SELECT NULL)) + @maxid3
FROM #LoginTemp
WHERE #LoginTemp.Dr_FName is not Null;
COMMIT TRANSACTION
END
GO