Overview: Siebel eBusiness Application Integration Volume l > Overview of Siebel eAI >

About Unicode Support and Siebel eAI

In past releases, Siebel eBusiness Applications required a single Siebel Server for each code page that was used in the integration environment. For example, if you wanted to implement Siebel eBusiness Applications using code pages 1252, 936, and 952, you then had to have three instances of Siebel eBusiness Applications implemented. Each one only integrated with an external system with a compatible code page.

However, by enabling Unicode, a single instance of Siebel eBusiness Applications can be used for global deployment. This allows a global company, with divisions in different countries with different languages, to deploy only a single instance of a Siebel application release. This provides unique integration challenges since the other systems in their global IT landscape may or may not be Unicode enabled. This issue is addressed by providing two Unicode deployment modes:

  • Unicode Siebel Server with Unicode database.
  • Unicode Siebel Server with traditional code page database.

For both of these deployment modes, integration of Siebel eBusiness Applications with external applications may involve data replication or data sharing or both. If the code page database is used, then data replication should be allowed only if the code page of the external system matches that of the Siebel Database.

When data is exchanged between applications using different code pages, each character's code point (numeric representation) must be converted from the standard used by the source code page into the equivalent used by the destination code page. This is true when characters are converted from:

  • Unicode to a traditional code page, for example UTF-16 to Shift-JS.
  • A traditional code page to Unicode, for example Shift-JS to UTF-8.
  • One type of Unicode to another type of Unicode, for example UTF-8 to UCS-2.

When there is no equivalent character in the destination code page and the conversion cannot be performed, then a conversion error is generated. Depending on the configuration, the conversion errors may either terminate the current transaction or proceed with the transaction by substituting replacement characters as necessary and ignoring the conversion error.

For details on conversion error handling, see Transports and Interfaces: Siebel eBusiness Application Integration Volume III.

As illustrated in Figure 1, Siebel eAI supports different adapters such as COM, CORBA, and MQSeries to integrate with external applications. Each of these adapters has to have the ability to:

  • Be able to work with the character encoding as specified in the picklist.
  • Do conversion to or from the external encoding.
  • Handle conversion errors.

NOTE:  COM, CORBA, and Java Bean are Unicode protocols. The sender and the recipient are responsible for conversion of the character sets. For example, Siebel application interfaces automatically take care of the conversion so these protocols do not need to do any conversion. Error handling is also handled differently by these protocols compared to the other adapters.

Figure 1.  Unicode Conversion

Siebel eAI components and adapters have been enhanced to support code page conversion and to allow you to specify how to handle the conversion errors. The Character Set Conversion for Text Data argument is available in a number of business services such as EAI MQSeries Server Transport, EAI DLL Transport, EAI File Transport, EAI HTTP Transport, EAI MSMQ Transport, EAI XML Converter, EAI JMS Business Service or Transcode. When these business services are invoked from a workflow process the valid set of encodings is controlled by a picklist, but if they are invoked through scripting or similar mechanisms, the character set name is supplied textually. For example, if the Siebel application is communicating with an external system through EAI MQSeries Server Transport, the invocation of the EAI MQSeries Server Transport business service to send or receive a message would also specify the character set the external system sends or expects to receive.

Target character encoding choices are:

  • None
  • Unicode (UTF-8, UTF-16)
  • Traditional Code page

NOTE:  For more information on Unicode support and a list of Siebel Character Set Encoding Names, see Global Deployment Guide.

Overview: Siebel eBusiness Application Integration Volume l