########## PilosusBot ##########

PilosusBot is a Telegram bot_ written in Python Flask. It reads Telegram chat messages, calculates sentiment scores (polarity score) and replies with the randomly selected message of the matching polarity.

.. contents:: Table of Contents

================ About PilosusBot

Stack

Application is based on Flask, a light-weight Python web framework that gets use of Werkzeug toolkit and Jinja2 template engine. Although Flask works fine with both Python 2 and 3, PilosusBot's written with Python 3 in focus.

Database-agnostic, application uses SQLAlchemy ORM, which enables user to choose between DBMS ranging from a simple SQLite to an enterprise solution of user's choice.

Asynchronous tasks tackled with Celery distributed task queue, which is to be used with a message broker RabbitMQ.

Redis is used for bounded in-memory queue for tracking messages processed so far.

The app relies on indicoio_ API for sentiment analysis as well as Telegram Bot API. It also uses Sentry_ for real-time error tracking. Make sure you get API tokens before you started app's deployment.

As a fallback for indicoio API (paid service) app uses polyglot_ package for natural language processing. Installing it may be tricky requiring icu and libicu-devel packages being installed to your OS.

================== Deployment & Usage

Based on Pili App_, PilosusBot follows its same steps of deployment. manage.py options are basically the same as in Pili too. Please, refer to Pili's documentation for help.

.env file

.env file is kept under project's root directory. It's an entry point of app's configuration and deployment. Standard .env file may look like this::

Testing/Development/Production switch

FLASK_CONFIG=production

Flask Variables

SECRET_KEY=... DATABASE_URL=postgresql+psycopg2://user:passwd@your_db_server:5432/your_prod_db TEST_DATABASE_URL=postgresql+psycopg2://user:[email protected]:5432/your_testing_db

Server

SERVER_MAX_CONNECTIONS=30

Mail server

MAIL_USERNAME=... MAIL_PASSWORD=... MAIL_SERVER=... MAIL_PORT=587 MAIL_USE_TLS=True

Celery

CELERY_BROKER_URL=amqp://guest@localhost// CELERY_RESULT_BACKEND=rpc:// CELERY_QUEUE_ASSESS=assess CELERY_QUEUE_SELECT=select CELERY_QUEUE_SEND=send

App

APP_ADMIN_EMAIL=... APP_ADMIN_NAME=...

Third-Party API

TELEGRAM_TOKEN=... INDICO_TOKEN=... SENTRY_DSN_SECRET=... SENTRY_DSN_PUBLIC=...

See also config.py for more configuration options.

======= Credits

The app partly relies on Miguel Grinberg's Flasky_ code (e.g. authentication module and DB models), which is a Flask application developed as an example for the excellent Flask Web Development_ book. Go grab your copy of the book, it's absolutely worth it!

.. _Flasky: https://github.com/miguelgrinberg/flasky .. _Flask Web Development: http://shop.oreilly.com/product/0636920031116.do .. _Telegram bot: https://core.telegram.org/bots .. _Pili App: https://github.com/pilosus/pili .. _indicoio: https://indico.io/ .. _polyglot: http://polyglot.readthedocs.io/en/latest/ .. _Sentry: https://sentry.io/

======= License

See LICENSE file.