aog

A simple and easy to use python client library for Actions on google webhook. A complete demo actions on google project built with aog python library is availble here.

Installation

You can install using pip install aog or pip3 install aog.

Code sample

import json
import os
from flask import Flask
from flask import request
from flask import make_response
from aog import conv

# Flask app should start in global layout
app = Flask(__name__)

@app.route('/', methods=['POST'])
def fullfillment():
    req = request.get_json(silent=True, force=True)

    print("Request:")
    print(json.dumps(req, indent=4))

    res = makeFullfillment(req)
    res = json.dumps(res, indent=4)
    #print(res)
    final = make_response(res)
    final.headers['Content-Type'] = 'application/json'
    return final

def makeFullfillment(req):
  # Wrte your python code here
  #isfrom function is used to check from which dialogflow intent request came from.
  if conv.isfrom(req,'ask_intent'): 
    # ask method expect a reply from user. after the message mic will be open for user to reply.
    res=conv.ask("You made me ask this","You made me print this")
    return res

Simple Responses

Simple responses can appear on audio-only, screen-only, or both surfaces. They take the form of a chat bubble visually, and TTS/SSML sound.

Sample Code

# ask method expect a reply from user. after the message mic will be open for user to reply.
res=conv.ask(speech="You made me ask this",displayText="You made me display this")
return res

# close method terminates conversation with the message. 
res=conv.close("Good bye")
return res

Extracting parameters (Entities) from conversation

You can access the parameter obtained by from user's.

color=get_parameters(req,'color')

Rich responses

Rich responses can appear on screen-only or audio and screen experiences. They can contain the following components:

• One or two simple responses (chat bubbles)
• An optional basic card
• Optional suggestion chips
• An optional link-out chip
• An option interface (list or carousel)

Requirements : actions.capability.SCREEN_OUTPUT you can check the requirement using conv.has_capability(req,actions.capability.SCREEN_OUTPUT)

Basic card

A basic card displays information that can include the following:

• Image
• Title
• Sub-title
• Text body
• Link button
• Border

Supported on actions.capability.SCREEN_OUTPUT. Required components : formatted text or image. Optional components : Titile, Sub-title, link button, Border (Image scale)

Sample code

if conv.has_capability(req,'actions.capability.SCREEN_OUTPUT'):  
      res=conv.basic_card(formatted_text="Content in the basic card",speech="You made me say this",\
                        title="Title of the Card",subtitle="subtitle",card_image=url,link_title="Link Title",link_url=url,image_scale="CROPPED")
      return res

List

The single-select list presents the user with a vertical list of multiple items and allows the user to select a single one. Selecting an item from the list generates a user query (chat bubble) containing the title of the list item.

Minimum 2 items are required. Supported on actions.capability.SCREEN_OUTPUT Required components : speech,items. Optional components : titile, description, image_url, image_scale, image_text,key

Sample code

 url="https://developers.google.com/actions/assistant.png"
    res=conv.list(speech="this is a list",title="List title",items=
                            [{'title':"item title",'description':'item description','image_url':url,'image_text':'image here','key':'your_key','synonyms':['one1','two1']},\
                            {'title':"item title1",'description':'item description1','image_url':url,'image_text':'image here1','key':'your_key1','synonyms':['one2','two2']},\
                            {'title':"item title2",'description':'item description1','image_url':url,'image_text':'image here1','key':'your_key2','synonyms':['one3','two3']}])
    return res

Handling a selected item

Once user select an item, intent with event actions_intent_OPTION will be invoked. from this event you can get selected item's key using key=conv.list_item_selectd(req) method.

Carousel

The carousel scrolls horizontally and allows for selecting one item.

Minimum 2 items are required. Supported on actions.capability.SCREEN_OUTPUT Required components : speech,items. Optional components : titile, description, image_url, image_scale, image_text,key

Sample code

 url="https://developers.google.com/actions/assistant.png"
    res=conv.carousels(speech="this is a list",title="List title",items=
                            [{'title':"item title",'description':'item description','image_url':url,'image_text':'image here','key':'your_key','synonyms':['one1','two1']},\
                            {'title':"item title1",'description':'item description1','image_url':url,'image_text':'image here1','key':'your_key1','synonyms':['one2','two2']},\
                            {'title':"item title2",'description':'item description1','image_url':url,'image_text':'image here1','key':'your_key2','synonyms':['one3','two3']}])
    return res

Handling a selected item

Once user select an item, intent with event actions_intent_OPTION will be invoked. from this event you can get selected item's key using key=conv.list_item_selectd(req) method.

Suggestion Chip

Use suggestion chips to hint at responses to continue or pivot the conversation.

Sample code

res=conv.suggestion_chips(speech="Here are some suggestions for you.",suggestions=['suggestion1','suggestion2','suggestion3'])
    return res

Required components : speech,suggestions.

Helper Intents for taking information from user

You can obtain these values from user.

• Name
  • Display name
  • Given name
  • Family name
• Device location (Coordinates)
• Confirmation ( Generic confirmation from the user (yes/no question) )
• Date Time
• Last seen ( lastSeen property to determine whether the user has interacted with your Action before.)
• User id ( An unique string identifier )

Sample code

res=conv.ask_permission(['NAME','DEVICE_PRECISE_LOCATION','DEVICE_COARSE_LOCATION'],context="To get you know better")
    return res

Minimum one permission is required. context could contain explaination of obtaining permission.

Handling permission

User can either accept or reject permission requests. bothways, an intent with event actions_intent_PERMISSION will be invoked. from this event you can handle the permissions.

Sample code

if conv.permission_granted(req): #Checking weather permission is granted or not.
      name=conv.get_username(req) # return type is a dictionary contains {displayName,givenName,familyName}
      location=conv.get_userlocation(req) # return type is a dictionary contains {latitude,longitude}

Date and Time

You can obtain a date and time from users by using conv.ask_datetime(). You can specify custom prompts when asking the user for a date and time.

Sample code

res=conv.ask_datetime(\
      initial="When do you want to come in?",\
      date="What day was that?",\
      time="What time"
      )
    return res

Handling Date and Time

After user responds to date and time intent, an intent with event actions_intent_DATETIME will be invoked. from this intent you can handle the result using conv.get_datetime(req).

date,time=conv.get_datetime(req) #date dictionary= {'year','date','month'} ,time dictionary ={hours and time} 

Confirmation

You can ask a generic confirmation from the user (yes/no question) and get the resulting answer. The grammar for "yes" and "no" naturally expands to things like "Yea" or "Nope", making it usable in many situations.

res=conv.ask_confirmation("Do you like python ?")
return res

Handling Confirmation

Aftter user responds to confirmation message, and intent with event actions_intent_CONFIRMATION will be invoked. from this event we can obtain user reponse using boolValue=conv.get_confirmation(req)

Last seen

You can use conv.get_lastseen(req) to determine whether the user has interacted with your Action before. method returns None if it's a first time user. else last seen time will be returned

User Id

You can obtain a unique string id to identify user from conv object using conv.get_userid(req).

More Info

For more information visit actions on google documentation.