I have a problem with a duplicate key. The input to the datapump is a large
txt file, the output is a SQL table with primary keys defined. I'd like to
simply drop any rows which are duplicated on the input (i.e. I'll take the
first input, and discard any duplicates). I'm aware that the SQL table could
be defined without a PK, but the rest of the application depends on this
definition.
Thanks in advance.

OK the easiest way to do it would be to load the text file into a staging
SQL Server tble and then use TSQL to only load unique rows.

Have a look here also

http://www.mssqlserver.com/faq/development-duprows.asp


"George Morton" <gmorton1 (AT) optonline (DOT) net> wrote

George,

One way to do it would be to load your text file into a staging table set up
with a unique index and ignore duplicate keys option enabled.

Ilya

"George Morton" <gmorton1 (AT) optonline (DOT) net> wrote