Nightmarish facebook to wordpress scraper

Facebook group scraping using NodeJs and NightmareJs.

Quick Setup

npm install

and

Input your facebook credentials in credentialss.js

1. Basic NightmareJS setup

npm install --save nightmare

2. Login to Facebook with NightmareJS

Setup external file to keep all your credentials

I'm setting it up like this so I don't have to think when I push files to git. Just add credentials.js to .gitignore and you can do the same. As you may guess, you'll have to create credentials.js file.

Generate wordpress authorization code

Check this article - don't worry this won't take more than 5 minutes.

// credentials.js

module.exports = {
    facebook_username : [your_facebook_username],
    facebook_password : [your_facebook_password],
    wordpress_auhtorization_code : [your_wordpress_auhtorization_code]
}

Login to facebook

Totally basic setup, just to see how it works, will do some refactoring later.

// FBscraper.js

const   Credentials = require('./credentials'), // Include our credentials
        Nightmare = require('nightmare'),
        nightmare = Nightmare({ show: true }),
        domain = 'https://facebook.com',      // Initial navigation domain
        groups = ['https://www.facebook.com/groups/394556737385005/?sorting_setting=RECENT_ACTIVITY', // Scrapable group array
        'https://www.facebook.com/groups/391908637633014/?sorting_setting=RECENT_ACTIVITY'];

        // add console logging - makes life a bit easier
        nightmare
            .on('console', (log, msg) => {console.log(msg); console.log( Credentials.facebook_username );})
            .useragent("Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.115 Safari/537.36") // browser info - not essential
            .goto(domain)
            .type('input[id="email"]', Credentials.facebook_username)
            .type('input[id="pass"]', Credentials.facebook_password)
            .click('#loginbutton>input')
            .wait(3000) // will wait 3 seconds before closing Nightmare headless browser
            .end()
            .catch((error) => {
                console.error('Something went wrong:', error)
        });

3. Looping through group pages with vo.js and NightmareJS

For this you will need external library vo.js

npm install --save vo

Once we login to facebook, we will loop through our groups. We achieve this using yield generator, which allows us to run asynchronous javascript synchronously.

// FBscraper.js

 .then(() => {

     var run = function*() {
         for (var i = 0; groups.length > i; i++) {
             var post = yield nightmare.goto(groups[i])
                 .wait(3000)
                 .then(function() {
                     console.log('Done yielding');
                 })
         }
     }

     vo(run)((res) => {
         console.log(res);
     })

 })

4. Scraping data with NightmareJS and jQuery

Make sure to install jQuery.

npm install --save jquery

jQuery must be injected into every page you navigate to. Lastly we select the field we want to scrape. Here only important part is evaluate, the rest is entirely up to you, keep in mind that facebook elements/wrappers may be different now and code inside evaluate function won't work. But have no fear, just inspect any element and get any value from it.

// FBscraper.js

.evaluate(() => {
    var posts = [];
    $('.fbUserStory').each(function(index) {
        var post = {};
        var postTitle = $(this).find('._l53>span:last-child').text();
        post.title = postTitle;
        posts.push(post);
    }) return posts;
})

Coming up..... 11/09/2017 Posting data to Wordpress using NightmareJS Any indication of interest in further tutorials is valued.