Browse Prior Art Database

Method to optimize data type mapping on federated database

IP.com Disclosure Number: IPCOM000033106D
Original Publication Date: 2004-Nov-26
Included in the Prior Art Database: 2004-Nov-26
Document File: 2 page(s) / 24K

Publishing Venue

IBM

Abstract

In a federated system(*), the data types at the data source must map to corresponding data types of the federated database so that the federated server can retrieve data from data sources. This invention makes it possible to create easily a virtual table with suitable data type mapping even when the data type information cannot be acquired from the data source, such as a text file, XML data and so on. It is realized by the procedure below. - Create the conversion candidates of data type mapping automatically by reading data from a data source. - Prioritize the candidates by using the column information of other tables which are already on the federated server. - Set the candidate with the highest priority to a default data type Note: A federated system is a special type of distributed database management system (DBMS) and allows you to query and manipulate data located on other servers. In a federated system, users and applications interface with a virtual table on a federated server to access a remote data source.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 74% of the total text.

Page 1 of 2

Method to optimize data type mapping on federated database

This invention makes it possible to create easily a virtual table with suitable data type mapping even when the data type information cannot be acquired from the data source, such as a text file, XML data and so on. It is realized by the procedure below.
1. Read data from a data source
2. Check each value of the data and see what data types it can be cast to by using the conversion definition table

The conversion definition table contains the following information * the data types which are supported by a federated server * the definition of the casts supported by a federated server

A value with a given data type can be cast to a different data type or to the same data type with a different length, precision or scale.

  For example, "103103" can be CHAR, VARCHAR, INTEGER, LONG, DATE, and so on. * locale

The locale information is also used here because "311003" can be DATE type in Europe, but not in the U.S.
3. Get the list of the data types which the data in each column of the data source can be cast to.

The data types in this list are the candidates of the target data types of the data type mapping.
4. Get a column name of each column of the data source

For example, if the XML file is a data source, a XML tag could be used as a column name.

If there is no information which can be used as a column name, you could ask a user to provide it.
5. Collect the column information, including the name and the data type, o...