kksweet8845/popnewthing

PopNewThing

This is a paper crawling chatbot, which crawls the paper from google AI and open AI. It is implemented by NodeJs and MongoDB. However, this chatbot is in free trail right now, so the push message to client will be forbiddened. It can't be deployed on the heroku, because it use the Selenium WebDriver to crawl the dynamical website, which is needed to download the corresponding webdriver to activate the browser. The workflow is implmented by xstate.

FSM

This app can be viewd with interactive website in here. The overview of the digram is shown as following.

How to deploy on cloud machine

In this app, I deploy the service on the google-cloud-platform, It is easy to use and maintain.

• On the VM, clone the this repo

$ git clone [email protected]:kksweet8845/PopNewThing.git

• Type npm install to install the required package

$ npm install

• Put the webdriver binary source file in directory, driver to the ENVIRONMENT PATH

$ echo $PATH
$ mv webdriver /one/of/env-path/webdriver

• Set up the config file.

  • execute npm run p to create a new config.js file.

    • This file contain some property which is needed to be specified.

  • execute npm run pdb to create a new database config.js file

    • If there is authentication restriction on your database server, you need to specified the user name and password when connecting to the databse.

• Activate the mongoDB, you need to install the mongodb in your server

$ mkdir history
$ mkdir mongodb-data
$ mongod -dbpath=./mongodb-data

• Set up the chrome

$ chmod u+x ./scripts/chrome-setup.sh
$ ./scripts/chrome-setup.sh

• Activate the service

$ npm run d

QR code