Webhook client#

class WebhookClient(request)[source]#

Bases: object

A client class for handling webhook requests from Dialogflow.

This class allows to dinamically manipulate contexts and create responses to be sent back to Dialogflow (which will validate the response and send it back to the end-user).

Parameters: request (dict) – The webhook request object (WebhookRequest) from Dialogflow.
Raises: TypeError – If the request is not a dictionary.
Return type: None

See also

For more information about the webhook request object, see the WebhookRequest section in Dialogflow’s API reference.

query#

The original query sent by the end-user.

Type: str

intent#

The intent triggered by Dialogflow.

Type: str

action#

The action defined for the intent.

Type: str

context#

An API class for handling input and output contexts.

Type: Context

contexts#

The array of input contexts.

Type: list(dict)

parameters#

The intent parameters extracted by Dialogflow.

Type: dict

console_messages#

The response messages defined for the intent.

Type: list(RichResponse)

original_request#

The original request object from detectIntent/query.

Type: str

request_source#

The source of the request.

Type: str

locale#

The language code or locale of the original request.

Type: str

session#

The session id of the conversation.

Type: str

add(responses)[source]#

Add response messages to be sent back to Dialogflow.

Examples

Adding a simple text response as a string:

>>> agent.add('Hi! How can I help you?')

Adding multiple rich responses one at a time:

>>> agent.add(Text('How are you feeling today?'))
>>> agent.add(QuickReplies(quick_replies=['Happy :)', 'Sad :(']))

Adding multiple rich responses at once:

>>> responses = [
...     Text('How are you feeling today?'),
...     QuickReplies(quick_replies=['Happy :)', 'Sad :('])
... ]
>>> agent.add(responses)

Parameters: responses (str, RichResponse, list(str, RichResponse)) – A single response message or a list of response messages.
Return type: None

property followup_event: Optional[Dict[str, Any]]#

The followup event to be triggered by the response.

Examples

Accessing the followup_event attribute:

>>> agent.followup_event
None

Assigning an event name to the followup_event attribute:

>>> agent.followup_event = 'WELCOME'
>>> agent.followup_event
{'name': 'WELCOME', 'languageCode': 'en-US'}

Assigning an event dictionary to the followup_event attribute:

>>> agent.followup_event = {'name': 'GOODBYE', 'languageCode': 'en-US'}
>>> agent.followup_event
{'name': 'GOODBYE', 'languageCode': 'en-US'}

Raises: TypeError – If the event is not a string or a dictionary.
Type: dict, optional

handle_request(handler)[source]#

Handle the webhook request using a handler or a mapping of handlers.

In order to manipulate the conversation programatically, the handler function must receive an instance of WebhookClient as a parameter. Then, inside the function, WebhookClient’s attributes and methods can be used to access and manipulate the webhook request attributes and generate the webhook response.

Alternatively, this method can receive a mapping of handler functions for each intent.

Note

If a mapping of handler functions is provided, the name of the corresponding intent must be written exactly as it is in Dialogflow.

Finally, once the request has been handled, the generated webhook response can be accessed via the response attribute.

Examples

Creating a simple handler function that sends a text and a collection of quick reply buttons to the end-user (the response is independent of the triggered intent):

>>> def handler(agent: WebhookClient) -> None:
...     agent.add('How are you feeling today?')
...     agent.add(QuickReplies(quick_replies=['Happy :)', 'Sad :(']))

Creating a mapping of handler functions for different intents:

>>> def welcome_handler(agent):
...     agent.add('Hi!')
...     agent.add('How can I help you?')
...
>>> def fallback_handler(agent):
...     agent.add('Sorry, I missed what you said.')
...     agent.add('Can you say that again?')
...
>>> handler = {
...     'Default Welcome Intent': welcome_handler,
...     'Default Fallback Intent': fallback_handler,
... }

Parameters: handler (callable, dict(str, callable)) – The handler function or a mapping of intents to handler functions.
Raises: TypeError – If the handler is not a function or a map of functions.
Returns: The output from the handler function (if any).
Return type: any, optional

property response: Dict[str, Any]#

The generated webhook response object (WebhookResponse).