>>>>>>
>>>>>>select * from myTable where nNumber IN (Fld1, Fld2, Fld3).
>>>>>>
>>>>>>
>>>>>>They will behave the same as your original code, just a shorter code.
>>>>>>
>>>>>>There is no faster approach for this problem.
>>>>>
>>>>>Thank you.
>>>>
>>>>You're welcome.
>>>
>>>Is it optimizable?
>>
>>It is exactly the same as original Fld1 = @nNumber or Fld2 = @nNumber or ...
>>
>>So, it is optimizable if we have indexes on each of these columns.
>
>Not exacyly.
>Can't test it in SQL Server right now but in VFP:
>
>CREATE CURSOR crsTest (Fld1 I, Fld2 I)
>FOR lnFor = 1 TO 20
> INSERT INTO crsTest VALUES(lnFor,20-lnfor)
>NEXT
>INDEX ON Fld1 TAG Fld1
>INDEX ON Fld2 TAG Fld2
>lnNumber = 12
>
>SYS(3054,12,[tst])
>SELECT * FROM crsTest WHERE m.lnNumber IN(Fld1, Fld2)
>MESSAGEBOX(tst)
>
>SELECT * FROM crsTest WHERE Fld1 = m.lnNumber OR Fld2 = m.lnNumber
>
>MESSAGEBOX(tst)
>SYS(3054,0)
>
Borislav,
When I ran your sample code it shows that using WHERE In (Fld1, Fld2) is not optimized but the WHERE Fld1 OR Fld2 is optimized. Is this what you wanted to show?
"The creative process is nothing but a series of crises." Isaac Bashevis Singer
"My experience is that as soon as people are old enough to know better, they don't know anything at all." Oscar Wilde
"If a nation values anything more than freedom, it will lose its freedom; and the irony of it is that if it is comfort or money that it values more, it will lose that too." W.Somerset Maugham
