TOC Project 2020

Template Code for TOC Project 2020

A Line bot based on a finite state machine

Setup

Prerequisite

Python 3.6
Pipenv
Facebook Page and App
HTTPS Server

Install Dependency

pip3 install pipenv

pipenv --three

pipenv install

pipenv shell

pygraphviz (For visualizing Finite State Machine)
- Setup pygraphviz on Ubuntu
- Note: macOS Install error

Secret Data

You should generate a .env file to set Environment Variables refer to our .env.sample. LINE_CHANNEL_SECRET and LINE_CHANNEL_ACCESS_TOKEN MUST be set to proper values. Otherwise, you might not be able to run your code.

Run Locally

You can either setup https server or using ngrok as a proxy.

a. Ngrok installation

macOS, Windows, Linux

or you can use Homebrew (MAC)

brew cask install ngrok

ngrok would be used in the following instruction

ngrok http 8000

After that, ngrok would generate a https URL.

Run the sever

python3 app.py

b. Servo

Or You can use servo to expose local servers to the internet.

Finite State Machine

Usage

The initial state is set to user.

Every time user state is triggered to advance to another state, it will go_back to user state after the bot replies corresponding message.

user
- Input: "go to state1"
  - Reply: "I'm entering state1"
- Input: "go to state2"
  - Reply: "I'm entering state2"

Deploy

Setting to deploy webhooks on Heroku.

Heroku CLI installation

macOS, Windows

or you can use Homebrew (MAC)

brew tap heroku/brew && brew install heroku

or you can use Snap (Ubuntu 16+)

sudo snap install --classic heroku

Connect to Heroku

Register Heroku: https://signup.heroku.com
Create Heroku project from website
CLI Login

heroku login

Upload project to Heroku

Add local project to Heroku project

heroku git:remote -a {HEROKU_APP_NAME}

Upload project

git add .
git commit -m "Add code"
git push -f heroku master

Set Environment - Line Messaging API Secret Keys

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

Your Project is now running on Heroku!

url: {HEROKU_APP_NAME}.herokuapp.com/callback

debug command: heroku logs --tail --app {HEROKU_APP_NAME}
If fail with pygraphviz install errors

run commands below can solve the problems
```
heroku buildpacks:set heroku/python
heroku buildpacks:add --index 1 heroku-community/apt
```
refference: https://hackmd.io/@ccw/B1Xw7E8kN?type=view#Q2-如何在-Heroku-使用-pygraphviz

Reference

Pipenv ❤️ @chihsuan

TOC-Project-2019 ❤️ @winonecheng

Flask Architecture ❤️ @Sirius207

Line line-bot-sdk-python

ncku-yee/TOC-Project-2020