dts with 36mb csv probem

[howard edidin]

I have a performance problem where I need to import a 36mb csv into a new csv file.

The csv has 6 fields; isbn(char), whse(char), stock(int), backorder(int), discount(char), price(float)

There are four possible values for the whse. I need to merge the data into an unique isbn, stock, backorder, discount, price

What I have been trying to do is parse the rows thru the fso and set the whse and stock fields into arrays. Then write the row to the new file.

Does anyone know of a better solution?

Thanks,

Howard

[Paul]

How about, in DTS:
1. bulk inserting your csv, you'll need to use a format file
2. use T-SQL to load the data into an output table in the order you wish
3. use ADO & FSO to create your new csv

------------
howard edidin at 11/19/2001 3:29:04 PM

I have a performance problem where I need to import a 36mb csv into a new csv file.

The csv has 6 fields; isbn(char), whse(char), stock(int), backorder(int), discount(char), price(float)

There are four possible values for the whse. I need to merge the data into an unique isbn, stock, backorder, discount, price

What I have been trying to do is parse the rows thru the fso and set the whse and stock fields into arrays. Then write the row to the new file.

Does anyone know of a better solution?

Thanks,

Howard