Re: Extract Oracle data into data file

From: ProNews/2 User <kiyoinc_at_ATTGlobal.XOUT.net>
Date: 2000/09/21
Message-ID: <hdwX38NCLQJC-pn2-drgMdirmX6qq_at_localhost>#1/1


On Wed, 20 Sep 2000 13:42:32, Basile Tohoundjona <btohoundjo_at_sinfor.fr> wrote:

> I have to extract data from Oracle database into text file using
> user defined stored procedures.
> Can somebody tell me if there is any problem proceding this way?
> Thank,
> Best regards.
>

There are several solutions with different trade offs. We're working on one. Our product has been used by a couple Alpha, Internal Beta testers with promising results.

At this time, we do most Oracle 7 and Oracle 8 field types but we don't do multi-table exports yet. This is under development and might be in the next Beta.

see: http://www.kiyoinc.com/nxtract.html

There are other solutions, SQL can do a lot. You might not need our utility.

  ==

I have a question. Our product produces a generic tab or comma delimited dump file. We have a couple customers who have managed to insert tabs, commas, and every other bit pattern into their data.

This makes it hard to extract variable length fields.

There are two general solutions to this.

  1. produce an import file that includes a field length prefix.
  2. use an escaped character as the delimiter. If the escaped character appears in the data, double it. If it appears twice, put it in the output four times, etc. Compress the character on import, if it occurs once, it is the delimiter and not data.

My questions are, which of the solutions is most acceptable to Sybase, Informix, Access, Excel, DB2, IMS, IDMS, Nomad, Ingres, Adabas, etc?

Is there a generic "transparent" import format? Should we distribute an importer for each of the possible databases?

While we're exporting and formating data, what kind of statistics would you like to see? Bytes per attribute? Longest row? Bar graphs of field lengths? We're thinking that this might be useful for sizing and verifying that updates ran.

Finally, if we have to write import utilities for the binary (not tab or comma delimitable) data, which database should we support first? DB2? Sybase?

Thanks for your time.

--
cory hamasaki 
Received on Thu Sep 21 2000 - 00:00:00 CEST

Original text of this message