How to cope with tables that may have different numbers of columns?

Submitted by jvevfvf on Wed, 2012-04-25 21:07.Troubleshooting


I have some oracle views that i want to index.
The views can have different numbers of columns for each customer - basically custom attributes that can be configured differently for each customer.

I have my query as select * from .
However, all the columns need to be populated, and are filled in into the -dataset-config.xml file.

We want to be able to have a customer independent -dataset-config.xml file, as our deployment method consists of shipping the xml files, and then triggering remote index creation.

Is there anyway to support this?

e.g. leave out the columns from the xml file (i'll try this)



Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
Submitted by will on Fri, 2012-04-27 07:27.

Can you please avoid using "select *", but select the exact columns?

Those XML columns, can you convert them into a virtual column? For example:

   SELECT e.poDoc.getClobval() AS poXML
   FROM po_xml_tab e;
Submitted by jvevfvf on Wed, 2012-04-25 22:59.

Didn't really expect it to :-)

I tried doing a wget on this:


but it prompts me for a username/password.
Can i pass those in?



Submitted by will on Sat, 2012-04-28 07:19.

This would create security hole. And this is not the right approach either.

You can pre-generate the <index-name>-dataset-config.xml file in multiple versions. And choose one version during deploy time.

If this still does not fit your need, please email dbsight at gmail dot com to discuss in more details.

Submitted by jvevfvf on Sun, 2012-04-29 16:15.

Hi Will,
thanks, yes, that's how i'm doing it at the moment - just wondering if there was an easier way :-)

I'll stick with this.