Scraping with Puppeteer

Not going into the details of explaining what Puppeteer is, go read that on their website. We just want to code

What can Puppeteer do

Puppeteer can do lots of things, but the most used ones are:

automate form filling and submissions
crawl a website and extract data from it
take screenshots of a website, page and even sections

What we'll be building

A MVP of a simple scraper that can scrape a website and return the data in a JSON format.

Requirements

Node.js; You can install this the traditional way by downloading and installing it, or you can use NVM to manage multiple versions of Node.js.
VSCode or any other text editor

Project setup

Please follow this guide to set up a Node.js project.

Install puppeteer

npm install puppeteer

Open the project with VSCode. If you don't have the code available on your PATH, please follow this guide

code .

Let's code

Alright, we're done setting up the project. Let's start coding! We'll use a local news website called Starnieuws as our example.

Basics

Basic usage of the puppeteer package

const puppeteer = require('puppeteer');
const url = 'https://www.starnieuws.com/';

// this is a async function that calls itself, no need to call it manually
(async () => {
    // launch a Chromium browser, set the headless property to true if you're deploying to production
    browser = await puppeteer.launch({
        headless: false,
    });

    // create new page object
    const page = await browser.newPage();

    // set viewport width and height
    await page.setViewport({
        width: 1920,
        height: 1080,
    });

    // nagivate to the url
    await page.goto(url);
})();

Run it with node <filename>.js via your VSCode Terminal. Voila, a browser should open and you should be able to see the website.

Catch errors (try catch)

We'll add a try catch block to catch any errors that might occur within the async function

let browser = null;

try {
    // launch headless Chromium browser
    browser = await puppeteer.launch({
        headless: false,
    });

    // create new page object
    const page = await browser.newPage();

    // set viewport width and height
    await page.setViewport({
        width: 1920,
        height: 1080,
    });

    await page.goto(url);

    // do something with the page ...

} catch (err) {
    console.log(`Error: ${err.message}`);
} finally {
    if (browser) {
        await browser.close();
    }
    console.log(`\nScraping ${url} done!`);
}

Scraping

Let's scrape something now. We want the title of the news article and the URL. Looking at the source code, we can see that it's in this order .headlines_content > ul > li. The list sits under the .headlines_content class.

Scraping contents looks like this in puppeteer, read more about evaluate() here

The element looks like this <a href="https://www.starnieuws.com/index.php/welcome/index/nieuwsitem/71066">LVV: Stijgend waterpeil Nannizwamp wordt gemonitord</a> so we want the href attribute, and the text content. Add this function under the await page.goto(url); code block

let data = await page.evaluate(() => {
    let results = [];
    let items = document.querySelectorAll('.headlines_content > ul > li');

    // iterate through the items and push the data to the results array
    items.forEach((item) => {
        results.push({
            title: item.querySelector('a').innerText,
            url: item.querySelector('a').href,
        })
    })

    return results
})

// do something with the scraped data
console.log(data)

Complete source code

Your complete code should look like this

app.js

const puppeteer = require('puppeteer');
const url = 'https://www.starnieuws.com/';

(async () => {
    let browser = null;

    try {
        // launch headless Chromium browser
        browser = await puppeteer.launch({
            headless: false,
        });

        // create new page object
        const page = await browser.newPage();

        // set viewport width and height
        await page.setViewport({
            width: 1920,
            height: 1080,
        });

        await page.goto(url);

        let data = await page.evaluate(() => {
            let results = []
            let items = document.querySelectorAll('.headlines_content > ul > li')
            items.forEach((item) => {
                results.push({
                    title: item.querySelector('a').innerText,
                    url: item.querySelector('a').href,
                })
            })
            return results
        })

        // do something with the scraped data
        console.log(data)

    } catch (err) {
        console.log(`Error: ${err.message}`);
    } finally {
        if (browser) {
            await browser.close();
        }
        console.log(`\nScraping ${url} done!`);
    }
})();

Run it! node app.js and you should see something like this

Storing your results

As a file. We can achieve this by storing the output to a file with the fs module. Read more about it here
In a Database (DB), you can use MongoDB, MySQL, PostgreSQL, etc.

Wrapping up

So, what have we learned?

Setting up a git repo and initializing npm
Getting started with Puppeteer
Catching errors with try catch blocks
Scraping data

Thank you so much for reading and following along, see you soon

Last update: February 5, 2023 11:36:20
Created: February 5, 2023 11:36:20