LineTalk - TOC Project 2021

A Line bot based on a finite state machine for TOC Project 2021
Github repository: https://github.com/Eddielin0926/FSM-Line-Bot

Usage

• Enter @bot for calling the menu
  

• Select 設定個人資訊 to set your personal profile

• Select 配對設定 to set your personal preference

• Select 開始配對 to start talking to others

• While chatting, enter @bot to call the menu

• Select 我要猜拳🖐 to start a rock paper scissors

• Select 幫我想話題💬 to list some random topic

• Select 我要離開🏃 to leave the chat

Setup

Prerequisite

• Conda (Anaconda/Miniconda)

Install Dependency

• Create Conda environment

$ conda create --name toc python=3.6
$ conda activate toc

• Install pygraphviz

(toc) $ conda install -c conda-forge pygraphviz

• Install other dependencies

(toc) pip install python-dotenv pygraphviz transitions line-bot-sdk flask colorama

Secret Data

• Creat .env file and fill in your channel info

LINE_CHANNEL_SECRET=<your_line_channel_secret>
LINE_CHANNEL_ACCESS_TOKEN=<your_line_channel_access_token>
PORT=8000

Ngrok

• Download Ngrok and execute

ngrok http 8000

Line Webhook

• Update webhook url Concate the url generate by Ngrok with /webhook (/callback) and put it to line bot channel

Run Locally

• Start Conda environment

$ conda activate toc

• Execute program

(toc) $ python app.py

Finite State Machine

Deploy

Deploy the Line-bot server on Heroku.

Heroku

• macOS, Windows

Connect to Heroku

• Register Heroku: https://signup.heroku.com

• Create Heroku project from website

• CLI Login

$ heroku login

Setup Environment on Heroku

• Setup environment variables

heroku config:set LINE_CHANNEL_SECRET=<your_line_channel_secret>
heroku config:set LINE_CHANNEL_ACCESS_TOKEN=<your_line_channel_access_token>

• Fix pygraphviz intall error

heroku buildpacks:set heroku/python
heroku buildpacks:add --index 1 heroku-community/apt

reference: https://hackmd.io/@ccw/B1Xw7E8kN?type=view#Q2-如何在-Heroku-使用-pygraphviz

Connect Github Repository to Heroku

• Connect to your Github repository
  

• If you have pushed to Github, you can use manual deploy
  

• You can also use automatic deploy. Once you push your repository, Heroku will start deploy.
  

Complete

• url: {HEROKU_APP_NAME}.herokuapp.com/callback
• logs: heroku logs --tail --app {HEROKU_APP_NAME}

Reference

• TOC-Project-2020
• Pipenv ❤️ @chihsuan
• TOC-Project-2019 ❤️ @winonecheng
• Flask Architecture ❤️ @Sirius207
• Line line-bot-sdk-python
• More details in the Slides and FAQ