Input please. (Dup. Check script or Separate files?)

Thread view:

Enable EMail Alerts

Start New Thread

Thread rating:

K&V P - 31 Mar 2004 00:01 GMT

FMP 6.0(4?) Windows XP

I have to load inventory from three different stores (approximately 200,000
records per store) then eliminate any duplicates to compile a list of what is
available at one or more of the stores for placement on our web site. The whole
thing is automated from a start-up script. At the moment, I am loading all three
original text files into one database, sorting by part number and Store, then
running a script using a self join to eliminate duplicates (around 225,000 final
unique records). I'm wondering if it would be more efficient to load each store
into its own DB, relate the three and select records that don't have a match.

If I were to use a relationship, it would be Store A gets priority of listing,
so relate B -> A, then C -> B and C -> A, to get all of A's records, Any of B's
that aren't in A, and Any of C's that aren't in either.

Anyone have any experience to tell me which will run faster? Or is there a third
choice I'm missing?

Kent

Reply to this Message

Glenn Schwandt - 31 Mar 2004 15:34 GMT

If you don't have one already, create an auto-enter serial number field in
your database. Import all three stores items into one database. Create a
self-relationship based upon the item number on both sides of the
relationship. Create a new calculation field as follows:

First (calculation, number) =
SerialNumber = Min( Self::SerialNumber)

Find for 1 in the SerialNumber field. This will give you just the first
record for each item number. If you import Store A, Store B and then Store
C, they will have that priority (lowest SerialNumber is kept). If you want
to keep only those records in the database, Find Omitted and Delete All
Records.

> FMP 6.0(4?) Windows XP
>
[quoted text clipped - 15 lines]
>
> Kent

Reply to this Message

Glenn Schwandt - 31 Mar 2004 15:44 GMT

> Find for 1 in the SerialNumber field.

That should say

Find for 1 in the *First* field.

Reply to this Message