Interactive Canvas API reference

The Interactive Canvas API is a JavaScript library that you use to interface with your Conversational Action using the interactiveCanvas object in your web app.

Note: The interactiveCanvas object is attached to the window object. If you use a front-end framework to develop your web app, you must ensure that you have access to window to set up the API.

For more information about Interactive Canvas, see the Interactive Canvas documentation.

Methods

Warning: Some of the Interactive Canvas APIs are currently in Developer Preview. You can test the APIs in the simulator, but do not deploy an Action that use these features to alpha, beta, or production channels. Actions deployed using these features will not function on end-user devices.

ready

ready(callbacks) returns undefined

Called by the Interactive Canvas web app once it has loaded to register callbacks.

Parameters
callbacks	InteractiveCanvasCallbacks An object that contains method callbacks.

Note: You must invoke the ready API in the web app as soon as the page is loaded. Otherwise, the Action closes.

sendTextQuery

sendTextQuery(textQuery) returns Promise<state>

Sends a text query to the conversational action.

Parameters
textQuery	string The query to send to the conversational action.
state	string One of the following: READY: Requests can be accepted by the platform. BLOCKED: There is a query in flight and the new query will be dropped. UNKNOWN: The current query state couldn't be determined.

Caution: textQuery must be 640 characters or less.

getHeaderHeightPx

getHeaderHeightPx() returns Promise<height>

Determines the height of the header at the top of a display.

Parameters
height	number The height in pixels.

outputTts

outputTts(text, openMic) returns undefined;

Plays SSML text.

Parameters
text	string The SSML supported string to be played.
openMic	boolean Whether to open mic after the TTS plays.

createIntentHandler

Caution: This API is currently in Developer Preview. You can test this API in the simulator, but do not deploy an Action that uses this feature to alpha, beta, or production channels. Actions deployed using this feature will not function on end-user devices.

createIntentHandler(intentId, callback) returns IntentHandler

Creates an intent handler for an NLU custom intent.

Parameters
intentId	string Intent name as defined in Actions Builder.
callback	function(MatchedIntent) Function to be executed when the intent is matched.
IntentHandler	object Intent handler object that can be registered using expect().

IntentHandler

An object representing an intent handler. Can be registered using expect() or prompt().

MatchedIntent

An object containing the data for a matched intent.

Methods
getIntentArg(parameterName)	function(string) Method to get the intent parameter of a matched intent.

createNumberSlot

Caution: This API is currently in Developer Preview. You can test this API in the simulator, but do not deploy an Action that uses this feature to alpha, beta, or production channels. Actions deployed using this feature will not function on end-user devices.

createNumberSlot(callback, hints?) returns Slot<number>

Creates a slot of type number.

Parameters
callback	function(FilledSlot) Function to be executed when the slot is filled.
hints	Hints Short words that the user might say when filling a slot. hints help the NLU system decide which slot to fill.
Slot	object Slot object that can be registered using prompt().

createTextSlot

Caution: This API is currently in Developer Preview. You can test this API in the simulator, but do not deploy an Action that uses this feature to alpha, beta, or production channels. Actions deployed using this feature will not function on end-user devices.

createTextSlot(callback, hints?) returns Slot<string>

Creates a slot of type string.

Parameters
callback	function(FilledSlot) Function to be executed when the slot is filled.
hints	Hints Short words that the user might say when filling a slot. hints help the NLU system decide which slot to fill.
Slot	object Slot object that can be registered using prompt().

createConfirmationSlot

Caution: This API is currently in Developer Preview. You can test this API in the simulator, but do not deploy an Action that uses this feature to alpha, beta, or production channels. Actions deployed using this feature will not function on end-user devices.

createConfirmationSlot(callback, hints?) returns Slot<boolean>

Creates a slot of type boolean.

Parameters
callback	function(FilledSlot) Function to be executed when the slot is filled.
hints	Hints Short words that the user might say when filling a slot. hints help the NLU system decide which slot to fill.
Slot	object Slot object that can be registered using prompt().

createOptionsSlot

Caution: This API is currently in Developer Preview. You can test this API in the simulator, but do not deploy an Action that uses this feature to alpha, beta, or production channels. Actions deployed using this feature will not function on end-user devices.

createOptionsSlot(options, callback, hints?) returns Slot<string>

Creates a slot from a pre-defined list of options.

Parameters
options	array<Option> Function to be executed when the slot is filled.
callback	function(FilledSlot) Function to be executed when the slot is filled.
hints	Hints Short words that the user might say when filling a slot. hints help the NLU system decide which slot to fill.
Slot	object Slot object that can be registered using prompt().

Slot

An object representing a slot to be filled. Can be registered using expect() or prompt().

FilledSlot<T>

An object representing a filled slot.

Parameter
T	boolean|number|string The slot type.

Option

An object representing a slot type entry.

Fields
key	string The unique identifier for the slot type entry.
synonyms	array<string> A list of words used to select the slot option.

Hints

An object representing hints for a slot.

Fields
associatedWords	array<string> A list of words used to help the NLU system decide which slot to fill.

expect

Caution: This API is currently in Developer Preview. You can test this API in the simulator, but do not deploy an Action that uses this feature to alpha, beta, or production channels. Actions deployed using this feature will not function on end-user devices.

expect(expectation) returns HandlerRegistration

Registers the expectation on the client. Can be called on intent handlers and slots.

Parameters
expectation	IntentHandler|Slot The intent handler or slot to be registered.
HandlerRegistration	object The registered intent handler or slot.

HandlerRegistration

An object representing a registered intent handler.

Methods
deleteHandler()	function Unregisters the expected intent handler.

clearExpectations

Caution: This API is currently in Developer Preview. You can test this API in the simulator, but do not deploy an Action that uses this feature to alpha, beta, or production channels. Actions deployed using this feature will not function on end-user devices.

clearExpectations() returns undefined;

Removes all expectations by deleting all previously registered intent handlers and slots.

prompt

Caution: This API is currently in Developer Preview. You can test this API in the simulator, but do not deploy an Action that uses this feature to alpha, beta, or production channels. Actions deployed using this feature will not function on end-user devices.

prompt(tts, expectation) returns Promise<Answer<T>>

Prompts the user with the provided TTS string, and registers the expectation on the client. Can be called on intent handlers and slots.

Parameters
tts	string The TTS string used to prompt the user for the provided expectation.
expectation	IntentHandler|Slot The intent handler or slot to be registered.
Answer<T>	object The user-supplied answer to the prompt.

Answer<T>

An object representing a user-supplied answer to a prompt.

Fields
value	boolean|number|string The value of the user-supplied answer.
status	string The answer status. One of the following: ANSWERED: The user responded to the prompt with an answer. NO_MATCH: The user was prompted for an answer and responded with an answer that didn't match the expected answer or any other expected intent. YIELD: The user responded to the prompt with a different intent.

triggerScene

Caution: This API is currently in Developer Preview. You can test this API in the simulator, but do not deploy an Action that uses this feature to alpha, beta, or production channels. Actions deployed using this feature will not function on end-user devices.

triggerScene(sceneName) returns Promise<TriggerSceneStatus>

Triggers a scene.

Parameters
sceneName	string The name of the scene to trigger.
TriggerSceneStatus	string The status of the call to triggerScene(). One of the following: SUCCESS: The request was successfully processed by the server. BLOCKED: The request was blocked by an existing query in flight. UNKNOWN: The status of the request is unknown.

setUserParam

Caution: This API is currently in Developer Preview. You can test this API in the simulator, but do not deploy an Action that uses this feature to alpha, beta, or production channels. Actions deployed using this feature will not function on end-user devices.

setUserParam(key, value) returns Promise<undefined>;

Sets the key-value pair that's passed and persisted between conversations.

Parameters
key	string The parameter key.
value	object|array|boolean|number|string The parameter value.

getUserParam

Caution: This API is currently in Developer Preview. You can test this API in the simulator, but do not deploy an Action that uses this feature to alpha, beta, or production channels. Actions deployed using this feature will not function on end-user devices.

getUserParam(key) returns Promise<object|array|boolean|number|string>;

Gets the param value for the given key if it exists and the status of the invocation.

Parameters
key	string The parameter key.

resetUserParam

Caution: This API is currently in Developer Preview. You can test this API in the simulator, but do not deploy an Action that uses this feature to alpha, beta, or production channels. Actions deployed using this feature will not function on end-user devices.

resetUserParam() returns Promise<void>;

Clears all user parameters.

setHomeParam

Caution: This API is currently in Developer Preview. You can test this API in the simulator, but do not deploy an Action that uses this feature to alpha, beta, or production channels. Actions deployed using this feature will not function on end-user devices.

setHomeParam(key, value) returns Promise<undefined>;

Sets the key-value pair that's passed and persisted between conversations.

Parameters
key	string The parameter key.
value	object|array|boolean|number|string The parameter value.

getHomeParam

Caution: This API is currently in Developer Preview. You can test this API in the simulator, but do not deploy an Action that uses this feature to alpha, beta, or production channels. Actions deployed using this feature will not function on end-user devices.

getHomeParam(key) returns Promise<object|array|boolean|number|string>;

Gets the param value for the given key if it exists and the status of the invocation.

Parameters
key	string The parameter key.

resetHomeParam

Caution: This API is currently in Developer Preview. You can test this API in the simulator, but do not deploy an Action that uses this feature to alpha, beta, or production channels. Actions deployed using this feature will not function on end-user devices.

resetHomeParam() returns Promise<void>;

Clears all user parameters.

setCanvasState

setCanvasState(state) returns undefined

Notifies the Assistant that the Canvas web app has updated its state. The setCanvasState method doesn't return a value. The state payload can be accessed through the conv.context.canvas.state property.

Caution: The length of the string representation of the state object must be less than or equal to 50,000 characters. If the string exceeds 50,000 characters in length, an error is thrown.

Parameters
state	Object The updated state of the Canvas app as a JSON object.

InteractiveCanvasCallbacks

These callbacks provide a way for you to respond to information or requests from your Conversational Action while the methods provide a way to send information or requests to your Conversational Action.

onUpdate

onUpdate(data) returns Promise|undefined

Called when a Canvas response is sent from the Conversational Action. TTS will start when the returned Promise is resolved, or timeout after 10 seconds.

Parameters
data	Object Updated data sent by the webhook.

onTtsMark

onTtsMark(markName) returns undefined

Callback that is invoked during TTS playback.

Special marks that always are triggered:

• START indicates the start of the TTS.
• END indicates the end of the TTS.
• ERROR indicates an error playing the TTS.

You can also define custom marks by using SSML and the <mark> tag, like <mark name="custom"/>.

Note: SSML marks are not invoked when the following occurs:

• Before TTS begins, the user taps the microphone, or the Interactive Canvas client invokes sendTextQuery. The TTS is cancelled and any SSML marks in the TTS are ignored.
• During TTS, the user taps the microphone, or the Interactive Canvas client invokes sendTextQuery. The TTS is interrupted and any remaining SSML marks in the TTS are ignored.

Parameters
markName	string Mark name invoked during TTS playback.

onInputStatusChanged

Caution: This API is currently in Developer Preview. You can test this API in the simulator, but do not deploy an Action that uses this feature to alpha, beta, or production channels. Actions deployed using this feature will not function on end-user devices.

onInputStatusChanged(inputStatus) returns undefined

Notifies the Canvas app that the Assistant input status has changed.

Parameters
inputStatus	Enum Status of voice and/or text input processing on the device. Can be the following: LISTENING : Microphone is open. IDLE : Microphone is closed. PROCESSING : Assistant is currently executing a query (microphone is closed).