Date: Tue, 21 Nov 2006 11:13:26 -0800
To: pgsql-sql@postgresql.org
From: Steve Midgley <public@misuse.org>
Subject: "Copy from" omit source columns?
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"; format=flowed
Message-Id: <20061121221911.C10D09FB26E@postgresql.org>

Hello,

(I already fear the answer to my question is "no..") :)

I am using the 'COPY FROM' SQL command (not psql \copy) to move data 
into a table from a CSV file. The source data are suitable for import 
EXCEPT there are extraneous columns of data in the source file. In 
essence, my table schema is:

id, col1, col2, col4

My source file schema is:

col1,col2,col3,col4

I'm curious if there is any way to tell "COPY" to ignore the third 
column of data? Basically I'd like to write something along the lines 
of:

copy import (col1,col2,\n,col4)
from 'c:/dev/import/source/test.csv'
with CSV HEADER

Note "\n" would indicate there there is a column in the source that 
should simply be ignored during the import. Of course, I tried this 
syntax and it didn't work. Perhaps there is another way of indicating 
that there is a "null column" in the source file that should be skipped 
during import?

Any assistance in solving this problem (without having to change the 
source file) would be greatly appreciated. If I have to change the 
source file, I will, but it would greatly simplify things if I didn't 
have to..

Best regards and thanks for any ideas,

Steve