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';
9// For more information, see https://crawlee.dev
10import { PlaywrightCrawler } from 'crawlee';
11// this is ESM project, and as such, it requires you to specify extensions in your relative imports
12// read more about this here: https://nodejs.org/docs/latest-v18.x/api/esm.html#mandatory-file-extensions
13// note that we need to use `.js` even when inside TS files
14import { router } from './routes.js';
15import { firefox } from 'playwright';
16import { launchOptions as camoufoxLaunchOptions } from 'camoufox-js';
17
18interface Input {
19    startUrls: string[];
20    maxRequestsPerCrawl: number;
21}
22
23// Initialize the Apify SDK
24await Actor.init();
25
26// Structure of input is defined in input_schema.json
27const {
28    startUrls = ['https://crawlee.dev'],
29    maxRequestsPerCrawl = 100,
30} = await Actor.getInput<Input>() ?? {} as Input;
31
32const proxyConfiguration = await Actor.createProxyConfiguration();
33
34const crawler = new PlaywrightCrawler({
35    proxyConfiguration,
36    maxRequestsPerCrawl,
37    requestHandler: router,
38    launchContext: {
39        launcher: firefox,
40        launchOptions: await camoufoxLaunchOptions({
41            headless: true,
42            // fonts: ['Times New Roman'] // <- custom Camoufox options
43        }),
44    }
45});
46
47await crawler.run(startUrls);
48
49// Exit successfully
50await Actor.exit();