>Hi Dorin
>
>Thanks for responding.
>
>>You can do another way
>
>I like the idea but will not creating "( primary keys )" list reach a length-of-string kind of limit?

You can limit the IN ( ) predicate to a number of keys

#define IN_KEYS    20
cSql = [SELECT ... FROM ... WHERE pk IN ( ]
iCount = 0 
SCAN
    iCount = iCount + 1
    cSql = cSql + transform( cursor.pk ) + [,]
    IF iCount = IN_KEYS
        remove the last [,]
        cSql = cSql + [ ) ]
        SQLEXEC()
        insert spt to result
        iCount = 0
        cSql = [SELECT ... FROM ... WHERE pk IN ( ]
   ENDIF
ENDSCAN  

for last keys < IN_KEYS, you should execute again after SCAN loop exit