Step 3: Set the character encoding

Voice Platform uses the browser.characterEncoding service property to determine what character encoding to use to parse language-specific application data. If this service property is not set, the default is the system encoding of the machine, which is ISO 8859-1 in most Latin-based language countries.

You set this service property on the Advanced tab for the Voice Browser service. The value must be supported by the Nuance language. If you are using multiple languages, then the chosen encoding must apply to all of them.

For example, to specify an encoding for a non-Latin language like Japanese, you could set browser.characterEncoding to utf-8. To specify an encoding for the Greek language, you would specify ISO8859-7. You must ensure that the locale you specify has actually been installed on your host. Refer to the documentation for your operating system for information on installing locales.

Please note:

• The service property is set at initialization only.
• The Voice Browser service supports all encodings supported by the Nuance Speech Server and Java.
• The service property affects inline grammars, prompts, and scripts, that is, only the content specified within the VoiceXML document. It doesn’t apply to externally referenced grammars or SSML files.
• If the encoding isn’t recognized, the Voice Browser service will not start and an alarm will be generated indicating that the encoding isn’t supported.