Crawlee + Playwright + Camoufox

1/**
2 * This template is a production ready boilerplate for developing with `PlaywrightCrawler`.
3 * Use this to bootstrap your projects using the most up-to-date code.
4 * If you're looking for examples or want to learn more, see README.
5 */
6
7// For more information, see https://docs.apify.com/sdk/js
8import { Actor } from 'apify';
9import { launchOptions as camoufoxLaunchOptions } from 'camoufox-js';
10// For more information, see https://crawlee.dev
11import { PlaywrightCrawler } from 'crawlee';
12import { firefox } from 'playwright';
13
14// this is ESM project, and as such, it requires you to specify extensions in your relative imports
15// read more about this here: https://nodejs.org/docs/latest-v18.x/api/esm.html#mandatory-file-extensions
16// note that we need to use `.js` even when inside TS files
17import { router } from './routes.js';
18
19interface Input {
20    startUrls: string[];
21    maxRequestsPerCrawl: number;
22}
23
24// Initialize the Apify SDK
25await Actor.init();
26
27// Structure of input is defined in input_schema.json
28const { startUrls = ['https://apify.com'], maxRequestsPerCrawl = 100 } =
29    (await Actor.getInput<Input>()) ?? ({} as Input);
30
31const proxyConfiguration = await Actor.createProxyConfiguration();
32
33const crawler = new PlaywrightCrawler({
34    proxyConfiguration,
35    maxRequestsPerCrawl,
36    requestHandler: router,
37    launchContext: {
38        launcher: firefox,
39        launchOptions: await camoufoxLaunchOptions({
40            headless: true,
41            proxy: await proxyConfiguration?.newUrl(),
42            geoip: true,
43            // fonts: ['Times New Roman'] // <- custom Camoufox options
44        }),
45    },
46});
47
48await crawler.run(startUrls);
49
50// Exit successfully
51await Actor.exit();