Question

Possible Duplicate:
Column does not exist in the IN clause, but SQL runs

I stumbled upon this today while at work today and I was wondering, why is it that the following code does not generate and error?

CREATE TABLE #TableA (ColumnA VARCHAR(25))
CREATE TABLE #TableB (ColumnB VARCHAR(25))

INSERT INTO #TableA (ColumnA) VALUES('1')
INSERT INTO #TableA (ColumnA) VALUES('2')
INSERT INTO #TableA (ColumnA) VALUES('3')

INSERT INTO #TableB (ColumnB) VALUES('1')

SELECT *
FROM #TableA
WHERE ColumnA IN(SELECT ColumnA FROM #TableB)

Output:

ColumnA
1
2
3

ColumnA does not exists on #TableB, how come no error is generated?

@@VERSION tells me I’m running this in:

Microsoft SQL Server 2008 (RTM) - 10.0.1600.22 (X64)   Jul  9 2008 14:17:44   Copyright (c) 1988-2008 Microsoft Corporation  Enterprise Edition (64-bit) on Windows NT 6.1 <X64> (Build 7601: Service Pack 1) (VM) 

Answer 1

ColumnA in the subquery (SELECT ColumnA FROM #TableB)
refers to the ColumnA of #TableA which is a valid column in SELECT list.

Hence there’s no error and you get three rows since you are comparing the #TableA.ColumnA with #TableA.ColumnA

If you want to verify the above statement replace ColumnA with anything but other than a valid column (e.g ColumnAB) you will get an error.

If you try this:

CREATE TABLE #TableA (ColumnA VARCHAR(25))
CREATE TABLE #TableB (ColumnB VARCHAR(25))

INSERT INTO #TableA (ColumnA) VALUES('1')
INSERT INTO #TableA (ColumnA) VALUES('2')
INSERT INTO #TableA (ColumnA) VALUES('3')

INSERT INTO #TableB (ColumnB) VALUES('1')

SELECT *
FROM #TableA
WHERE ColumnA IN(SELECT ColumnAB FROM #TableB)

the output will be

Msg 207, Level 16, State 1, Line 14
Invalid column name ‘ColumnAB’.