Solved

# Get differences

Posted on 2013-10-23
250 Views
Hi,

Select Col1,Col2,Col3,
From T1
Except

Select Col1,Col2,Col3,
From T2

I have like above query but with more than 3 columns and I got some rows , How can I get a specific column's name that has a difference value?.
0
Question by:SanPrg
• 2
• 2

LVL 48

Accepted Solution

PortletPaul earned 500 total points
If I have interpreted your question properly, you want to not only identify the rows that are different, but also which columns of those rows are the cause of the difference.

I'm not personally familiar with any prior solutions to this, but I think the following may give you some ideas. However you will note that that this approach requires running the except query twice, then generating a row_number() on each, then transposing both from row orientation to column orientation - and in the transposition everything gets converted to nvarchar(max).

So: in a nutshell, don't expect this to be quick.

Perhaps others will know of a better approach, but here goes:
``````**MS SQL Server 2012 Schema Setup**:

CREATE TABLE Table1
([ID] int, [Col1] varchar(50), [Col2] varchar(30), [Col3] varchar(100), [Col4] varchar(60))
;

INSERT INTO Table1
([ID], [Col1], [Col2], [Col3], [Col4])
VALUES
(444111, 'Flintstone', 'Fred', '1 Stone St', 'Bedrock')
;

CREATE TABLE Table2
([ID] int, [Col1] varchar(50), [Col2] varchar(30), [Col3] varchar(100), [Col4] varchar(60))
;

INSERT INTO Table2
([ID], [Col1], [Col2], [Col3], [Col4])
VALUES
(555111, 'Flintstone', 'Fred', '5 Pebble St', 'Bedroc')

**Query 1**:

;WITH
cte1 AS (
SELECT
rn
, ca1.col_name
, ca1.col_value
FROM (
SELECT
row_number() over (ORDER BY Col1,Col2,Col3,Col4) AS rn
, Col1,Col2,Col3,Col4
FROM (
SELECT Col1,Col2,Col3,Col4
FROM Table1                 --<< nb table order is important
EXCEPT

SELECT Col1,Col2,Col3,Col4
FROM Table2                 --<< nb table order is important
) AS x
) AS y
CROSS APPLY (
VALUES
('Col1',convert(nvarchar(MAX),col1))
,('Col2',convert(nvarchar(MAX),col2))
,('Col3',convert(nvarchar(MAX),col3))
,('Col4',convert(nvarchar(MAX),col4))
) AS ca1 (col_name, col_value)
)
,cte2 AS (
SELECT
rn
, ca1.col_name
, ca1.col_value
FROM (
SELECT
row_number() over (ORDER BY Col1,Col2,Col3,Col4) AS rn
, Col1,Col2,Col3,Col4
FROM (
SELECT Col1,Col2,Col3,Col4
FROM Table2                 --<< nb table order is important
EXCEPT

SELECT Col1,Col2,Col3,Col4
FROM Table1                 --<< nb table order is important
) AS x
) AS y
CROSS APPLY (
VALUES
('Col1',convert(nvarchar(MAX),col1))
,('Col2',convert(nvarchar(MAX),col2))
,('Col3',convert(nvarchar(MAX),col3))
,('Col4',convert(nvarchar(MAX),col4))
) AS ca1 (col_name, col_value)
)
SELECT
cte1.rn
, cte1.col_name  as "Column Name"
, cte1.col_value as "Table1 Value"
, cte2.col_value as "Table2 Value"
FROM cte1
INNER JOIN cte2 ON cte1.rn=cte2.rn AND cte1.col_name = cte2.col_name
WHERE cte1.col_value <> cte2.col_value

**[Results][2]**:

| RN | COLUMN NAME | TABLE1 VALUE | TABLE2 VALUE |
|----|-------------|--------------|--------------|
|  1 |        Col3 |   1 Stone St |  5 Pebble St |
|  1 |        Col4 |      Bedrock |       Bedroc |

[1]: http://sqlfiddle.com/#!6/6b8a1/1
``````
{+ edit, sorry - typos}
0

Author Comment

Great,It's so quick.
Just I need to understand it, why  CROSS APPLY  ?
0

LVL 48

Expert Comment

'Unpivoting' using CROSS APPLY
There are several methods for transposing columns to rows, this is one of them.

Personally I find the syntax very simple to understand, plus it is also very fast.
(In some testing done there is a suggestion it is faster than using unpivot, e.g:
http://www.sqlservercentral.com/articles/CROSS+APPLY+VALUES+UNPIVOT/91234/ )

So, it's a personal preference but one based on simplicity and speed.

------
I should note while I'm here that there might be occasions where my suggested solution might not work reliably. But I think this is is unavoidable.

The matching of the 2 results is achieved by row_number(), which rely on the order by that is used. e.g. order by col1,col2,col3,col4

If col1 is different between the 2 lists then row numbers may not align, and in fact this is true of any column:

so to mitigate use the columns with least difference at the start of that order by

The difficulty is that we can't include any reliably unique id in the except query (else everything is different) and because we don't have this reliable id we have to generate something, but because we have listed only differences that generated id may be faulty.

- a classic catch 22

{+ edit, corrected the URL }
0

Author Closing Comment

Excellent Work.
Thanks
0

## Featured Post

Let's review the features of new SQL Server 2012 (Denali CTP3). It listed as below: PERCENT_RANK(): PERCENT_RANK() function will returns the percentage value of rank of the values among its group. PERCENT_RANK() function value always in be…
For both online and offline retail, the cross-channel business is the most recent pattern in the B2C trade space.
Via a live example, show how to setup several different housekeeping processes for a SQL Server.
Viewers will learn how to use the SELECT statement in SQL and will be exposed to the many uses the SELECT statement has.