botanalytics 0.2.9

Conversational analytics & engagement tool for chatbots

Botanalytics - Conversational analytics & engagement tool for chatbots

Python SDK currently supports

• Google Assistant
• Amazon Alexa
• Facebook
• Slack
• Generic
• Rasa

If you want to use nodejs instead, checkout Botanalytics Node.js SDK

If you want to use ruby instead, checkout Botanalytics Ruby SDK

Setup

Create a free account at https://www.botanalytics.co and get a Token.

Botanalytics is available via pip.

pip install botanalytics

Google Assistant

from botanalytics.google import GoogleAssistant
from botanalytics.google import GoogleAssistantCloudFunctions
import os

# Optional callback function, if you specify it, you can handle failed
#  attemps the way you want
def err_callback(err, reason, payload):
    pass

# Debug(optional)->bool, token(required)->str,
# callback(optional)->function, is_async(optional)-> bool
botanalytics = GoogleAssistant(
                debug=True,
                token=os.environ['BOTANALYTICS_API_TOKEN'],
                callback=err_callback,
                is_async=True
                )
# request_payload -> dict, response_payload -> dict
botanalytics.log(request_payload, response_payload)

# For Google Cloud Functions
# Debug(optional)->bool, token(required)->str, is_async(optional)->bool
botanalytics_gc = GoogleAssistantCloudFunctions(
                    debug=True,
                    token=os.environ['BOTANALYTICS_API_TOKEN'],
                    is_async=True)
# request_payload->dict, response_payload->dict
botanalytics_gc.log(request_payload, response_payload)

Amazon Alexa

from botanalytics.amazon import AmazonAlexa, AmazonAlexaLambda
import os

# Optional callback function, if you specify it, you can handle failed
#  attemps the way you want
def err_callback(err, reason, payload):
    pass

# Debug(optional)->bool, token(required)->str,
# callback(optional)->function, is_async(optional)-> bool
botanalytics = AmazonAlexa(
                    debug=True,
                    token=os.environ['BOTANALYTICS_API_TOKEN'],
                    callback=err_callback,
                    is_async=True)
# request_payload -> dict, response_payload -> dict
botanalytics.log(request_payload, response_payload)

# For lambda
# Debug(optional)->bool, token(required)->str, is_async(optional)->bool
botanalytics_aal = AmazonAlexaLambda(
                        debug=True,
                        token=os.environ['BOTANALYTICS_API_TOKEN'],
                        is_async=True)
# request_payload->dict, response_payload->dict
botanalytics_aal.log(request_payload, response_payload)

Facebook

from botanalytics.facebook import FacebookMessenger
from http.server import BaseHTTPRequestHandler, HTTPServer
from urllib.parse import parse_qs
import requests
import json
import os

# Optional callback function, if you specify it, you can handle failed
#  attemps the way you want
def err_callback(err, reason, payload):
    pass

botanalytics_token = os.environ['BOTANALYTICS_API_TOKEN']
fb_page_token = os.environ["FACEBOOK_PAGE_TOKEN"]
# Debug(optional)->bool, token(required)->str, fb_token(optional)-> str,
# callback(optional)->function, is_async(optional)-> bool
botanalytics = FacebookMessenger(
                        debug=True,
                        token=botanalytics_token,
                        fb_token=fb_page_token,
                        callback=err_callback,
                        is_async=True
                        )

class BasicRequestHandler(BaseHTTPRequestHandler):
    def __init__(self, *args, **kwargs):
        super(ServerHandler, self).__init__(*args, **kwargs)

    def do_POST(self):
        if self.path == '/webhook/facebook':
            print('Incoming facebook message')
            self.handle_facebook_message()
        else:
            # Handle your paths

    def handle_facebook_message(self):
        request_body = str(
                self.rfile.read(
                        int(self.headers['content-length'])
                        ),
                'utf-8')
        message_payload = json.loads(request_body)
        #log incoming message
        botanalytics.log_incoming(message_payload)

        for entry in message_payload['entry']:
            page_id = entry['id']
            message_time = entry['time']
            for event in entry['messaging']:
                sender_id = event['sender']['id']
                message_to_build= {}

                #
                # Handle your event and build your message
                #

                # log outgoing message
                botanalytics.log_outgoing(message_to_build, sender_id)

                # send your message
                response_message = {
                    "recipient": {"id":sender_id},
                    "message": message_to_build,
                }
                requests.post(
                    'https://graph.facebook.com/v2.6/me/messages?access_token='+fb_page_token,
                     json = response_message)

        self.send_response(200)

def run(server_class=HTTPServer, handler_class=BasicRequestHandler,
        port=8000):
    server_address = ('', port)
    httpd = server_class(server_address, handler_class)
    print('Starting httpd...')
    try:
        httpd.serve_forever()
    except BaseException as ki:
        print(str(ki))


if __name__ == "__main__":
    run()

Generic

from botanalytics.generic import Generic
import os

# Optional callback function, if you specify it, you can handle failed
#  attemps the way you want
def err_callback(err, reason, payload):
    pass

# Debug(optional)->bool, token(required)->str,
# callback(optional)->function, is_async(optional)-> bool
botanalytics = Generic(
                    debug=True,
                    token=os.environ['BOTANALYTICS_API_TOKEN'],
                    callback=err_callback,
                    is_async=True)
# message -> dict
botanalytics.log(message)

Rasa

There are two options to integrate Rasa with Botanalytics

OPTION 1: Endpoints.yml

Add Botanalytics to your endpoints.yml

Add a line to your endpoints.yml so that rasa-core is configured to send events to Botanalytics:

event_broker:
    type: botanalytics.rasa.Rasa
    api_token: BOTANALYTICS_TOKEN
    debug: true

OPTION 2: Event Broker Include Botanalytics

from botanalytics.rasa import Rasa

Consume events with Botanalytics

def _callback(ch, method, properties, body):
    event = json.loads(body)
    db = Rasa(token=os.environ['BOTANALYTICS_TOKEN'])
    db.publish(event)

SlackRTMApi

from botanalytics.slack import SlackRTMApi
from slackclient import SlackClient
import os

sc = SlackClient(os.environ["SLACK_API_TOKEN"])

# Optional callback function, if you specify it, you can handle failed
#  attemps the way you want
def err_callback(err, reason, payload):
    pass

# Debug(optional)->bool, token(required)->str,
# slack_client_instance(required)-> SlackClient,
# callback(optional)->function, is_async(optional)-> bool
botanalytics = SlackRTMApi(
                     debug=True,
                     token=os.environ['BOTANALYTICS_API_TOKEN'],
                     slack_client_instance=sc, callback=err_callback,
                     is_async=True
                     )
if sc.rtm_connect():
  while sc.server.connected is True:
        for message in sc.rtm_read():
            # message -> dict
            botanalytics.log_incoming(message)
            # Handle incoming message
            #       .
            #       .
            sc.rtm_send_message("welcome-test", "test")
            botanalytics.log_outgoing("welcome-test", "test")
            time.sleep(1)
else:
    print "Connection Failed"

SlackEventApi

from botanalytics.slack import SlackEventApi
from http.server import BaseHTTPRequestHandler, HTTPServer
from urllib.parse import parse_qs
import json

# Optional callback function, if you specify it, you can handle failed
#  attemps the way you want
def err_callback(err, reason, payload):
    pass

# Debug(optional)->bool, token(required)->str,
# slack_token(required)-> str, callback(optional)->function
# is_async(optional)-> bool
botanalytics = SlackEventApi(
                 debug=True,
                 token=os.environ['BOTANALYTICS_API_TOKEN'],
                 slack_token=os.environ["SLACK_API_TOKEN"],
                 callback=err_callback,
                 is_async=True
                 )

class BasicRequestHandler(BaseHTTPRequestHandler):
    def __init__(self, *args, **kwargs):
        super(ServerHandler, self).__init__(*args, **kwargs)

    def do_POST(self):
        if self.path == '/slack/events':
            print('Incoming event')
            self.handle_slack_event()

        elif self.path == '/slack/interactive':
            print('Incoming interactive event')
            self.handle_slack_interactive()
        else:
            # Handle your paths

    def handle_slack_interactive(self):
        byte_dictionary = parse_qs(
                            self.rfile.read(
                                int(self.headers['content-length'])),
                                keep_blank_values=1
                            )
        interactive_payload = json.loads(
                                byte_dictionary[b'payload'][0].decode()
                              )
        botanalytics.log(interactive_payload)
        #
        # Handle interactive payload
        #
        self.send_response(200)

    def handle_slack_event(self):
        request_body = str(
                        self.rfile.read(
                            int(self.headers['content-length'])
                         ),
                        'utf-8'
                       )
        event_payload = json.loads(request_body)
        botanalytics.log(event_payload)
        #
        #Handle event message
        #
        self.send_response(200)

def run(server_class=HTTPServer, handler_class=BasicRequestHandler,
                port=8000):
    server_address = ('', port)
    httpd = server_class(server_address, handler_class)
    print('Starting httpd...')
    try:
        httpd.serve_forever()
    except BaseException as ki:
        print(str(ki))


if __name__ == "__main__":
    run()

Follow the instructions at https://botanalytics.co/docs