To Find and Eliminate the Duplicates - sql

Question

I use this code to update one of my table by calling a function which generates a random ID each item. I started with around 1000 rows but now the size is growing and i find that there are duplicate ID's in the table. Is there any way to can modify the code i am using, so that it look for ID's that are already generated in the table so that it will generate a new code if there is a similar one. I also noticed

Answer 1

Your code shows you setting the field password, but the results show that UniqueID is the duplicated field. (Maybe it's password renamed?)
Assuming userId is unique: (if not, ADD an actual identity column NOW, "ALTER TABLE dbo.Users ADD ID INT NOT NULL IDENTITY(1, 1)" should do the trick) and assuming password is the field to change, use the following:
DECLARE #FN VARCHAR(20);
DECLARE #LN VARCHAR(20);
DECLARE #PW VARCHAR(20);
DECLARE #ID INT;
SELECT TOP 1
#FN = FirstName,
#LN = LastName,
#ID = userID
FROM dbo.Users
WHERE Password IS NULL;
WHILE ##ROWCOUNT = 1
BEGIN
SET #PW = dbo.GenerateID(FirstName, LastName);
WHILE EXIST (SELECT TOP 1 Password FROM dbo.Users WHERE Password = #PW)
SET #PW = dbo.GenerateID(FirstName, LastName);
UPDATE dbo.Users SET Password = #PW WHERE userId = #ID;
SELECT TOP 1
#FN = FirstName,
#LN = LastName,
#ID = userID
FROM dbo.Users
WHERE Password IS NULL;
END
This should look for a blank password. If none is found the outer loop is skipped. If one is found, we generate passwords until we find one not in the table. Next we look for another row with a blank password before the end of the outer loop.

Answer 2

Sounds like your new to this. Don't worry, TSQL is pretty easy to learn. First thing first, I suggest that you create a unique non-clustered index on the UniqueID column--this will prevent duplicates values from being inserted into your table. If someone does try to insert a duplicate value into the table it will throw an exception. Before you can use this though you'll need to remove all the duplicate 'UniqueID' values from your table.
CREATE UNIQUE NONCLUSTERED INDEX [IDX_UniqueID] ON [dbo].[Users]
(
[UniqueID] ASC
) ON [PRIMARY]
You can learn more about non-clustered indexes here: https://learn.microsoft.com/en-us/sql/relational-databases/indexes/clustered-and-nonclustered-indexes-described
I also suggest that you consider changing the underlying type of your UniqueID field to a 'uniqueidentifier.' Here's an example of a table schema that uses a 'uniqueidentifier' column type for the UniqueID column:
CREATE TABLE [dbo].[Users](
[personId] [int] IDENTITY(1,1) NOT NULL,
[firstName] [nvarchar](50) NOT NULL,
[lastName] [nvarchar](50) NOT NULL,
[UniqueID] [uniqueidentifier] NOT NULL,
CONSTRAINT [PK_Users] PRIMARY KEY CLUSTERED
(
[personId] ASC
) ON [PRIMARY]
) ON [PRIMARY]
A 'uniqueidentifier' column type in SQL Serever holds a Global Unique Identifier (aka a GUID or UUID). It's easy to generate a GUID in most languages. To generate a GUID in TSQL you just new to invoke the NEWID() function.
SELECT NEWID() -- output: D100FC00-B482-4580-A161-199BE264C1D1
You can learn more about GUIDs here: https://en.wikipedia.org/wiki/Universally_unique_identifier
Hope this helps. Best of luck on your project. :)