Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Data Loading/SQLLDR

Data Loading/SQLLDR

From: grape_lady <member26043_at_dbforums.com>
Date: Wed, 05 Mar 2003 23:17:39 +0000
Message-ID: <2607467.1046906259@dbforums.com>

When loading seismic navigation data, a record is generated in a header table. The unique combination of survey and linename yeild a lineid from a sequence table.

I have raw data files (ascii) with other attributes to be loaded for thousands of seismic lines. Unfortunately all of the misc. tables in the model that will be populated are keyed off of the lineid which I don't have in my raw files because it was generated when the navigation was loaded.

I use an application called Finder to populate other oracle databases. They have tweeked sqlldr and some of the functionality is not really part of native sqlldr. I don't know what all of the tweeks are. I can load table2 with records from a raw file without the key by looking it up in table1. That syntax I use is as follows:

line_id xtlu '<seis_line_hdr, line_id, survey=position(1:6), line_name=position(12:27)> 'saveas line_id

This database is not connected to Finder in any way. I am looking for the best way to load the following tables given the fact that the header table has the columns (lineid, survey, linename):

raw file           aliastable       sourcetable      locationtable
survey            lineid (key)     lineid (key)      lineid (key)
linename         linealias          procsource       county
linealias                                                       state
procsource
county
state

Any advise will be appreciated. I am hoping not to have to create and populate a temp table and join the tables.

Thank you.

--
Nicole


Posted via http://dbforums.com
Received on Wed Mar 05 2003 - 17:17:39 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US