Introduction to Character Sets

Conversion requirements among components

For a single user session, character sets are used in several different areas of the software:

• Instance character set

• Client character set

• Analytic workspace character set

Moreover, you will need to import data from the RDBMS or from a text file. When the user makes a query, data is transferred among these different parts of the system. In the best scenario, the same character set is used throughout these various parts. In other scenarios, the use of different character sets might have no effect, or might render the data partially or fully undecipherable, or might slow I/O time.

Instance character set

The instance character set is the one used internally by OLAP Services to manipulate and process text. Because OLAP Services performs whatever character conversions are required to communicate with its clients and its data sources, you must choose the instance character set with care. This information is provided in "Choosing an Instance Character Set".

Analytic workspace character set

The analytic workspace character set is used to encode the data stored on disk. When the workspace is created, it is assigned the same character set as the instance of OLAP Services.

If you are creating or using a persistent analytic workspace, note that the workspace character set cannot be changed. All of the text data in a workspace must be in the workspace character set.

To attach a workspace, the workspace character set must be the same as the instance character set or US7ASCII, which has the same encoding as all available instance character sets. A US7ASCII workspace must be attached read-only unless the instance character set is also US7ASCII.

To update a workspace, the workspace character set must be the same as the instance character set.

Client character set

The client character set is used by a client application to represent text. Applications developed using the OLAP API or the BI Beans are Java applications and always use UTF-16 Unicode, as described in "Use of UTF-8 encoding". These applications identify the appropriate language and territory settings for their client sessions.