Headless WebKit with PhantomJS

April 19, 2013 / Mad Coding, PhantomJS

PhantomJS is just pure awesomeness. I gave a presentation on it at Well.ca and talked about how one can use it to facilitate website testing (with CasperJS). At the time we just encountered a bug with the website, so it was fitting. Having a headless WebKit can do so much for you other than just tesitng too!

Taking Screenshots

For another project I was working on, I wanted to capture webpages and save them as images. Before I discovered PhantomJS, I used CutyCapt and Qt graphics-dojo example. PhantomJS is much much simpler to use though. Check out the official rasterize example for details.

Fetching Actual Website Content

Another thing that’s great about having a headless browser is that you can use it to fetch the actual content of a website. Nowadays so many sites are javascript driven so simply fetching the initial HTML won’t do. Below is how you can use PhantomJS to capture the actual content.

var page = require('webpage').create(),
    system = require('system'),
    fs = require('fs'),
    address, output;

if (system.args.length != 3) {
    console.log('Usage: grab.js URL filename');
} else {
    address = system.args[1];
    output = system.args[2];

    page.open(address, function (status) {
        if (status !== 'success') {
            console.log('Unable to load the address!');
        } else {
            window.setTimeout(function () {
                var results = page.evaluate(function() {
                    return document.documentElement.innerHTML;

                try {
                    var f = fs.open(output, "w");
                } catch (e) {
            }, 200);

Save the code to grab.js and run it via PhantomJS by providing the URL to fetch and the output file to save the content to.