Which is the fastest way to find duplicate numbers in a unsorted field in a
paradox database.

I have tried table.fieldbyname('Cust No').asstring - but it is extremely
slow. There are about 3500 records in the database. I'm not sure how to use
findnext, etc because I don't know how to set the filters.

Please Help
Kevin

news:ag2892$f2r$1@ctb-nnrp2.saix.net...

Doing lookups on field that isn't part of an index will always be quite slow
since, on average, 1/2 of all records must be examined. Since you are
looking for duplicates the entire table will have to be examined before
reporting no duplicate exists.

If you just want to find out if they exist:
See if you can create a unique index on the field.

To find out what they are use SQL like the following:
select cust_no, count(*)
from my_table
group by cust_no
having count(*)>1

The otherway is to sort the dataset by cust_no and then loop through
the whole
dataset checking that the current value does not equal the value in
the previous record.

Marc

me <fr...@telkomsa.net> wrote in <ag2892$f2...@ctb-nnrp2.saix.net>:

I don't know if the following works on a Paradox table or not.  I am using
it on a dbase Table to compare against 3 fields and it is FAST.

  // Duplicate Search using Locate Method:
 if LogTable.Locate ('CALL;FREQ;MODE',
    VarArrayOf([CallEntered.Text,FreqEntered.Text,Mode.Text]),[])
  then  // Duplicate