Puppeteer script to check that Google Analytics Client ID is consistent across different platforms

verify-client-ids.js

'use strict';

const puppeteer = require('puppeteer');

// Test scenarios
// 4 different entrypoints
// ~8 different cookie setting scenarios (ie every page has at least 1 video)
// 2 different pre-set options - cookies and cookie-less

const reqUrls = [
    'https://www.mirror.co.uk/comm-part-test/brightcove-lead-11944737',
    'https://www.mirror.co.uk/comm-part-test/brightcove-lead-11944737.amp',
    "https://www-mirror-co-uk.cdn.ampproject.org/c/s/www.mirror.co.uk/comm-part-test/brightcove-lead-11944737.amp",
    'https://www.liverpoolecho.co.uk/comm-part-test/brightcove-lead-14226081',
    'https://www.liverpoolecho.co.uk/comm-part-test/brightcove-lead-14226081.amp',
    'https://www-liverpoolecho-co-uk.cdn.ampproject.org/c/s/www.liverpoolecho.co.uk/comm-part-test/brightcove-lead-14226081.amp',
    'https://www.mirror.co.uk/3am/video/superstar-singer-hrvy-talks-halina-11047508?widgetId=6933150',  // FBIA video widget
];

(async() => {
    const browser = await puppeteer.launch({ headless: true });
    
    const page = await browser.newPage();
    await page.setRequestInterception(true);
    // set user agent to mobile phone
    await page.setUserAgent('Mozilla/5.0 (Linux; Android 6.0; Nexus 5 Build/MRA58N) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/63.0.3239.132 Mobile Safari/537.36');

    // test setting pre-defined cookie
    // page.setCookie({
    //     name : '_ga',
    //     value : 'GA1.3.57927592.1517247373',
    //     domain : '.mirror.co.uk',
    //     path : '/',
    //     expires : 1577836801
    // });

    page.on('request', request => {

        const url = request.url();
        
        // look for tracking script
        if (url.match(/^https?:\/\/www\.google-analytics\.com\/(r\/)?collect/i)) {
            // ignore requests to the wrong GA profiles
            if (url.match(/ua-110513849/i)) {
                let query = url.match(/\?v=1(.*?)$/)[1];
                let params = decodeURI(query).split('&');
                for (let i=0;i<params.length;i++) {
                    let key = params[i];
                    // log if client ID, type, or event action
                    // if (key.match(/^(tid|cid|cd2|t|ea)=/)) {
                    //     console.log(key);
                    // }
                    if (key.match(/^(cd2|t)=/)) {
                        console.log(key);
                    }
                }
                console.log('---------------------');
            }
        }

        // look for call to CID server
        if (url.match(/^https?:\/\/stmg-prod\.mirror\.co\.uk\/analytics\.config\.json/i)) {
            console.log("STMG endpoint called\n---------------------");
        }
        request.continue();
    });

    try {
        // randomise entrypoints
        var randReqUrls = shuffle(reqUrls);
        for (var i = 0;i<randReqUrls.length;i++) {
            console.log("-------------------------\nFetching " + randReqUrls[i] + "\n-------------------------");
            await page.goto(randReqUrls[i], { waitUntil: 'networkidle2' });
            await page.waitFor(1000);
        }

    } catch (err) {
        console.log("Couldn't fetch page " + err);
    }
    
    await browser.close();
})();

function shuffle(array) {
  var currentIndex = array.length, temporaryValue, randomIndex;

  // While there remain elements to shuffle...
  while (0 !== currentIndex) {

    // Pick a remaining element...
    randomIndex = Math.floor(Math.random() * currentIndex);
    currentIndex -= 1;

    // And swap it with the current element.
    temporaryValue = array[currentIndex];
    array[currentIndex] = array[randomIndex];
    array[randomIndex] = temporaryValue;
  }

  return array;
}