SQL Server using wildcard within IN

Question

Since I believe this should be a basic question I know this question has probably been asked, but I am unable to find it. I'm probably about to earn my Peer Pressure badge, but I'll ask anyway:

Is there a way in SQL Server that I am not aware of for using the wildcard character % when using IN.

I realize that I can use OR's like:

select *
from jobdetails
where job_no like '0711%' or job_no like '0712%'

and in some cases I can use a subquery like:

select *
from jobdetails
where job_no in (select job_no from jobs where job_id = 39)

but I'm looking to do something like the following:

select *
from jobdetails
where job_no in ('0711%', '0712%')

In this case it uses the percent sign as a character instead of a wildcard character so no rows are returned. I currently just use a bunch of OR's when I have to do this, but I know there has to be a better way. What method do you use for this?

Answer 1

How about:

WHERE LEFT(job_no, 4) IN ('0711', '0712', ...)

Answer 2

How about something like this?

declare @search table
(
    searchString varchar(10)
)

-- add whatever criteria you want...
insert into @search select '0711%' union select '0712%'

select j.*
from jobdetails j
    join @search s on j.job_no like s.searchString

Answer 3

I think I have a solution to what the originator of this inquiry wanted in simple form. It works for me and actually it is the reason I came on here to begin with. I believe just using parentheses around the column like '%text%' in combination with ORs will do it.

select * from tableName
where (sameColumnName like '%findThis%' or sameColumnName like '%andThis%' or 
sameColumnName like '%thisToo%' or sameColumnName like '%andOneMore%') 

Answer 4

You could try something like this:

select *
from jobdetails
where job_no like '071[12]%'

Not exactly what you're asking, but it has the same effect, and is flexible in other ways too :)

Answer 5

SELECT c.* FROM(
SELECT '071235' AS token UNION ALL SELECT '07113' 
 UNION ALL SELECT '071343'
UNION ALL SELECT '0713SA'
UNION ALL SELECT '071443') AS c
JOIN (
SELECT '0712%' AS pattern UNION ALL SELECT '0711%' 
 UNION ALL SELECT '071343') AS d
ON c.token LIKE d.pattern

071235
07113
071343

Answer 6

I had a similar goal - and came to this solution:

select *
from jobdetails as JD
where not exists ( select code from table_of_codes as TC 
                      where JD.job_no like TC.code ) 

I'm assuming that your various codes ('0711%', '0712%', etc), including the %, are stored in a table, which I'm calling *table_of_codes*, with field code.

If the % is not stored in the table of codes, just concatenate the '%'. For example:

select *
from jobdetails as JD
where not exists ( select code from table_of_codes as TC 
                      where JD.job_no like concat(TC.code, '%') ) 

The concat() function may vary depending on the particular database, as far as I know.

I hope that it helps. I adapted it from:

http://us.generation-nt.com/answer/subquery-wildcards-help-199505721.html

Answer 7

1. I firstly added one off static table with ALL possibilities of my wildcard results (this company has a 4 character nvarchar code as their localities and they wildcard their locals) i.e. they may have 456? which would give them 456[1] to 456[Z] i.e 0-9 & a-z

2. I had to write a script to pull the current user (declare them) and pull the masks for the declared user.

3. Create some temporary tables just basic ones to rank the row numbers for this current user

4. loop through each result (YOUR Or this Or that etc...)

5. Insert into the test Table.

Here is the script I used:

Drop Table #UserMasks 
Drop Table #TESTUserMasks 

Create Table #TESTUserMasks (
    [User] [Int] NOT NULL,
    [Mask] [Nvarchar](10) NOT NULL)

Create Table #UserMasks (
    [RN] [Int] NOT NULL,
    [Mask] [Nvarchar](10) NOT NULL)

DECLARE @User INT
SET @User = 74054

Insert Into #UserMasks 
select ROW_NUMBER() OVER ( PARTITION BY ProntoUserID ORDER BY Id DESC) AS RN,
       REPLACE(mask,'?','') Mask
from dbo.Access_Masks 
where prontouserid = @User

DECLARE @TopFlag INT
SET @TopFlag = 1

WHILE (@TopFlag <=(select COUNT(*) from #UserMasks))
BEGIN
    Insert Into #TestUserMasks 
    select (@User),Code from dbo.MaskArrayLookupTable 
    where code like (select Mask + '%' from #UserMasks Where RN = @TopFlag)

    SET @TopFlag = @TopFlag + 1
END
GO

select * from #TESTUserMasks

Answer 8

As Jeremy Smith posted it, i'll recap, since I couldn't answer to that particular question of his.

select *
from jobdetails
where job_no like '071[1-2]%'

If you just need 0711% and 0712% you can also place a ranges within the brackets. For the NOT keyword you could also use [^1-2]%

Answer 9

The IN operator is nothing but a fancy OR of '=' comparisons. In fact it is so 'nothing but' that in SQL 2000 there was a stack overflow bug due to expansion of the IN into ORs when the list contained about 10k entries (yes, there are people writing 10k IN entries...). So you can't use any wildcard matching in it.

Answer 10

In Access SQL, I would use this. I'd imagine that SQLserver has the same syntax.

select * from jobdetails where job_no like "0711*" or job_no like "0712*"

Answer 11

You have the answer right there in your question. You cannot directly pass wildcard when using IN. However, you can use a sub-query.

Try this:

select *
from jobdetails
where job_no in (
select job_no
from jobdetails
where job_no like '0711%' or job_no like '0712%')
)

I know that this looks crazy, as you can just stick to using OR in your WHERE clause. why the subquery? How ever, the subquery approach will be useful when you have to match details from a different source.

Raj

Answer 12

Try this

select * 
from jobdetails 
where job_no between '0711' and '0713'

the only problem is that job '0713' is going to be returned as well so can use '07299999999999' or just add and job_no <> '0713'

Dan zamir

Answer 13

This might me the most simple solution use like any

select *
from jobdetails
where job_no like any ('0711%', '0712%')

In Teradata this works fine.