SessionizeScarper.js

// To run this script, you can either:
// 1. Go to sessionize.com in your browser, open the browser console, and paste the following script
// 2. Run it locally in Node.JS and comment the following imports in the begining of the file to use JSDom as a dom parser.
//    const jsdom = require("jsdom")
//    const { JSDOM } = jsdom
//    global.DOMParser = new JSDOM().window.DOMParser

const sleep = ms => new Promise(resolve => setTimeout(resolve, ms))

const now = new Date();

(async () => {
    const response = await fetch('https://sessionize.com/sitemap/events.xml');
    const xml = await response.text();
    const parser = new DOMParser();
    const doc = parser.parseFromString(xml, 'application/xml');
    const urls = doc.querySelectorAll('urlset > url > loc');
    const eventUrls = Array.from(urls).map(url => url.textContent);
    const events = await Promise.all(eventUrls.map(async (url, index) => {
        try {
            await sleep(index * 1000)
            const response = await fetch(url, { headers: { 'User-Agent': 'Mozilla/5.0' } });
            const html = await response.text();
            const parser = new DOMParser();
            const doc = parser.parseFromString(html, 'text/html');
            const alreadyEnd = doc.querySelector('.alert.alert-danger');
            if (alreadyEnd) return;
            const title = doc.querySelector('title').textContent.split(': Call for Speakers @ Sessionize.com')[0].trim();
            const dateField = doc.querySelector('.m-b-md .fa-calendar').closest('.row').querySelector('h2').textContent;
            const date = new Date(dateField.split(',')[0].split('\n').join('').trim());
            const multipleEvents = !!(dateField.split(',')?.length > 1);
            const description = doc.querySelector('meta[name="description"]').getAttribute('content');
            const endDate = new Date(doc.querySelector('strong[data-date]').getAttribute('data-date').split('|')[0]);
            const location = doc.querySelector('.m-b-md .fa-map-marker')?.closest('.row').querySelector('h2').textContent.replace(/\s{2,}/g, ' ').trim('').split('\n').join('|');
            const flightCovered = !!doc.querySelector('.m-b-md .fa-plane');
            const hotelCovered = !!doc.querySelector('.m-b-md .fa-bed');
            if (endDate < now) return;
            const event = {
                title,
                description,
                date: date.toDateString(),
                endDate: endDate().toDateString(),
                url,
                daysLeft: Math.ceil((endDate - now)/(1000*60*60*24)),
                daysBetween: Math.ceil((date - endDate)/(1000*60*60*24)),
                hotelCovered,
                flightCovered,
                location,
                multipleEvents
            };
            console.log(event);
            return event;
        } catch (error) {
            console.log('--- ERROR ---');
            console.log(url);
            console.log(error);
            console.log('--- END OF ERROR ---');
        }
    }));
    console.log('--- DONE ---');
    console.log(JSON.stringify(events.filter((e) => e !== undefined)));
})();