This chapter is from the book
This chapter is from the book
This chapter is from the book
3.24 item
|
Element type |
item |
|
Attributes |
repeat | weight | xml:lang |
|
Parents |
item | one-of | rule |
|
Children |
count | item | one-of | ruleref | tag | token |
|
Description |
Wraps a rule expansion and associates with it a weight to indicate the probability of this element's rule expansion being matched and a tag used to fill one or more fields if this item is matched in a grammar. The item element can also delimit different rule expansions in a one-of element. This is one of the VoiceXML components likely to change: The W3C is currently working on the Semantic Interpretation for Speech Recognition Specification which will subsume this element. |
DTD
<!ELEMENT item (%rule-expansion; | tag)* > <!ATTLIST item weight NMTOKEN #IMPLIED repeat CDATA #IMPLIED xml:lang NMTOKEN #IMPLIED >
Language model
Attributes
-
repeat : CDATA
Specifies that the contents of this item may be heard repeatedly. The permissible patterns of repeat may be expressed as in Table 3-2.
-
tag : CDATA (deprecated)
In the early version of the VoiceXML 2.0 Specification the tag attribute represented an expression to be interpreted if this item element's rule-expansion is matched. This is typically an ECMAScript expression.
Table 3-2. The permissible patterns of repeat
|
Pattern |
Description |
|---|---|
|
0-1 |
This item is optional. |
|
|
This item may occur zero or more times. |
|
|
This item may occur one or more times. |
|
|
This item may occur n or more times (where n is an integer). |
|
n-m |
This item may occur between n and m times inclusive (where n and m are both integers and m > n). Older implementations may still use the count element to specify repeats. An example of this is shown below. |
-
weight : NMTOKEN
Specifies the expected probability of this item's rule-expansion being spoken. This is used to provide a hint to the speech recognition engine.
-
xml:lang : NMTOKEN
Specifies the spoken language used with this element. For example, en-US for US English.
Children
-
A rule-expansion
To be enumerated within a one-of element.
Examples
Example 3-27 A weighted grammar definition
<?xml version="1.0" encoding="iso-8859-1"?>
<vxml version="2.0">
<form id="test">
<field name="favcolor">
<prompt>What is your favorite color?</prompt>
<grammar xml:lang="en-US" version="1.0" root="colors">
<rule id="colors" scope="public">
<one-of>
<item><tag>red</tag>red</item>
<item weight="0.5"><tag>red</tag>burgundy</item>
<item><tag>blue</tag>blue</item>
<item weight="0.5"><tag>blue</tag>indigo</item>
<item><tag>green</tag>green</item>
</one-of>
</rule>
</grammar>
<filled>
<prompt>
You said <value expr="favcolor"/>.
</prompt>
</filled>
</field>
</form>
</vxml>
An example of the deprecated form of item where the tag attribute is used is shown for the grammar element in 3.20, “grammar,”.
Example 3-28 The deprecated count element
<?xml version="1.0" encoding="iso-8859-1"?>
<vxml version="2.0">
<form id="test">
<field name="howmuch">
<prompt>
Do you think this dialog is very good <break/>
or very very good?
</prompt>
<grammar xml:lang="en-US" version="1.0" root="ROOT">
<rule id="ROOT" scope="public">
<count number="optional">I think its</count>
<count number="0+">very</count>
good
</rule>
</grammar>
</field>
<filled>
<prompt>
I'm glad you liked it!
</prompt>
</filled>
</form>
</vxml>
Though the count element is now deprecated, on many interpreters, the count element (Example 3-28) is still the only way to indicate a repeat or an optional rule expansion.