From: loktai@engineer.com (BadMan!)
Newsgroups: comp.databases.oracle
Subject: Re: Sqlloader: All database in one single plain text file
Date: 3 Mar 2004 09:02:18 -0800
Organization: http://groups.google.com
Lines: 28
Message-ID: <8e5d4aee.0403030902.3aaad9c3@posting.google.com>
References: <8e5d4aee.0402251004.5fbc6c7f@posting.google.com> <2687bb95.0402260739.57d4b603@posting.google.com>
NNTP-Posting-Host: 194.224.160.129
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
X-Trace: posting.google.com 1078333338 30316 127.0.0.1 (3 Mar 2004 17:02:18 GMT)
X-Complaints-To: groups-abuse@google.com
NNTP-Posting-Date: Wed, 3 Mar 2004 17:02:18 +0000 (UTC)


Mark.Powell@eds.com (Mark D Powell) wrote in message news:<2687bb95.0402260739.57d4b603@posting.google.com>...

> From looking at the notes and example in the 9.2 Utilities manual it
> does not appear that sqlldr is really designed to load different sets
> of data into multiple tables.  The only expample takes portions of the
> same input and places it into different tables, but the input data
> format does not change.

Yap.

> It should however be easy to write a C program or Shell script to
> split the input file into one file per table based on the first
> delimited parameter being the table name: fruits.dat, users.dat,
> etc.....

awk is the tool. 

But I wanted to know if there was a way to use Oracle-native
platform-independent tools...


> Perhaps a true sqlldr expert will see your post and have time to
> respond if there is any practical way to feed multiple layouts to
> multiple tables in a single job.

Thank you anyway!

Best regards from spain.