Hi Sam, I know you posted this DTS query a while back. There is however
a relatively simple cludge I use that works and I'd be very interested
to know how you resolved the issue yourself.



I would agree with Darrens' assessment simply handling imports in pure
DTS. Adding the file system object in your import routines will
increase your options considerably though.



If, for example your DTS destination is say a 255 column table made up
of nvarchar(255)'s you can always ensure that a text file (say a csv) of
any column length up to 255 columns (or the 8096 byte row limit) can be
imported into this table - irrespective of either data or column
variations.



Use the FileSystemObject (Allen Mitchel has some good examples -
http://www.sqldts.com) and add 255 column delimiters to the end of each
row, save the file and envoke the datapump. The datapump will ignore
rows in excess of the destination table limit.



This is not an elegant solution but it does enable me to provide a data
import mechanism for users when I have no idea what their data import
files looks like.



Cheers







In your DTS script use



I'd be very interested in how you resolved the issue since you posted
this message.


--
Posted via http://dbforums.com