<say-as>

Specifies the type of text construct contained within the element when this is ambiguous.

For example, it can be used to indicate whether TTS should speak a number (“1012”) as a number (“one thousand twelve”) or as a string of digits (“one oh one two”).

Support for this element depends on the language used.

When you use this element, you must specify the text construct using the interpret-as attribute.

Attributes

This section describes the attributes you can set for the <say-as> element.

Name

Data type

 Default
detail

CDATA

optional
format

CDATA

 optional

interpret-as

CDATA

required

detail

Provides more specific instructions on how speech is to be rendered if the format attribute is ambiguous.

This attribute may only be used when the interpret-as and format attributes have been specified, and even then it may not apply.

format

Defines the specific format to be used for the text where this could be ambiguous: for example, if the interpret-as attribute is set to “spell”, the matching format could specify “alphanumeric” to indicate that only actual digits and letters are to be spoken, while spaces and punctuation are to be ignored. If the format is instead “strict”, this indicates that all characters in the string are to be spoken, including spaces and punctuation.

This attribute may only be used when the interpret-as attribute has been specified, and even then it may not apply.

interpret-as

Indicates the content type of the <say-as> element, and thus how it should be spoken. For example, the text 3/4 may be a fraction (three quarters), or it may be a date (March 4th); the interpret-as attribute indicates which interpretation applies.

Children

Parents

Sample code

<prompt>The access code is
  <say-as interpret-as="spell" format="alphanumeric">123456</say-as>
  Use this code to open the safe, which will contain
  <say-as interpret-as="number">123456</say-as>
  dollars in local currency.
</prompt>