Global Deployment Guide > Planning Global Deployments > Integration Considerations >

Character Conversion Errors in External Systems

When Siebel eBusiness Applications send data to an external application using a traditional (non-Unicode) character set encoding, the external application may not be able to accept all of the data if the data does not belong to the character set encoding of the application. If this happens, the unrepresentable character will be converted to a substitution character. The substitution character can be a question mark or upside down question mark (indicating character loss in the database) or a black box (indicating a display issue), depending on the database. This is called unrepresentable character conversion and the following section shows how to avoid this situation through character conversion error handling techniques.

Character Conversion Error Handling

The transport and adaptor business services have two character conversion error handling modes:

• Generate an error without sending the message
• Ignore the error, substitute replacement characters wherever possible, and send the message

These error handling modes are specified in an argument to the business service call.

There is second case that occurs when you are using a Unicode communication stream, such as XML to communicate between Siebel eBusiness Applications and an external, non-Unicode application. In this case, the character conversion can be handled by the external application after it receives the message. This could be too late and the external application might not provide the desired error handling functionality.

A solution to this would be use the Transcode business service before sending any data to the external application. You can use the Validate method in the Transcode business service to detect whether the character conversion performed by the external application will fail. If the character conversion will fail, you can chose not to send the data. For more information, see Transcode Business Service.