bottoku - Python Chat/Messenger Bot Microframework

bottoku(pronounced as bot-oku) is a chat bot micro framework for Pythonista.

Feature

bottoku is a microframework, however, you can build chat bots without restriction.

For example,

• Integration with web frameworks e.g. flask, django, etc...
• Serverless bots running on AWS Lambda
• Building messages compatible with Line, Facebook, and Slack
• Messages with attachments e.g. images, videos, etc...
• Conversations with contexts: hi! -> who r u? -> john -> hi john, how about...
• Favorite database backend

Compared to Hubot

• No built-in server. If you want to build non-serverless bots, choose favorite web frameworks.
• No built-in binary. bottoku is just a library, so please do pip install to integrate.
• Very flexible routing system. Conditions are not defined by regex, but by Python's lambda functions.
• And of course, bottoku is a Python framework.

Usage

First, pip it.:

pip install bottoku

Then, define actions and views:

# views.py
from bottoku.api.facebook.template import TextMessage
def greeting():  # view is just a function to return a list of messages
    return [TextMesage(text='hello, i am a bot')]

# actions.py
from bottoku import route
@route(lambda message, context: 'hello' in message)  # lambda condition to reply
def say_hello(render, message, context):
    render(greeting())  # call `render` function

Next, define a bot class (application class), it is used to define routes and API clients.:

# mybot.py
from bottoku import Bot
from bottoku.renderer.stdout import StdoutRenderer
from bottoku.api.facebook.renderer import FacebookRenderer
class MyBot(Bot):
    def __init__(self, env):
        super(MyBot, self).__init__(
          env,
          [
              actions.say_hello,
              # other actions...
          ],
          BlackholeRepository(),  # repository to set user's data
          [
              StdoutRenderer(),
              FacebookRenderer(),
          ]  # API clients
        )

Finally, instantiate the bot and call reply in handler.:

# handler.py, lambda handler or web frameworks' webhook handler
def handler():
    message, receiver_id = extract_from_message()
    bot = MyBot(env)
    bot.reply(message, receiver_id)

Currently, I have not prepared docs yet. But you can see examples.

Example

• example: most basic example
• example_api: advanced example to use Facebook and Slack APIs

Bugs / Feature requests

Github's issue or mention to @yamitzky