From 591358c5a54919b7b8fc0819679c11ffe1d3472e Mon Sep 17 00:00:00 2001 From: MohamedBassem Date: Tue, 5 Mar 2024 00:33:59 +0000 Subject: fix: Use puppeteer adblocker to block cookies notices --- packages/workers/crawler.ts | 6 ++++++ packages/workers/package.json | 1 + 2 files changed, 7 insertions(+) (limited to 'packages/workers') diff --git a/packages/workers/crawler.ts b/packages/workers/crawler.ts index f1ee07f3..fbbee730 100644 --- a/packages/workers/crawler.ts +++ b/packages/workers/crawler.ts @@ -18,6 +18,7 @@ import { db } from "@hoarder/db"; import { Browser } from "puppeteer"; import puppeteer from "puppeteer-extra"; import StealthPlugin from "puppeteer-extra-plugin-stealth"; +import AdblockerPlugin from "puppeteer-extra-plugin-adblocker"; import metascraper from "metascraper"; @@ -70,6 +71,11 @@ async function launchBrowser() { export class CrawlerWorker { static async build() { puppeteer.use(StealthPlugin()); + puppeteer.use( + AdblockerPlugin({ + blockTrackersAndAnnoyances: true, + }), + ); await launchBrowser(); logger.info("Starting crawler worker ..."); diff --git a/packages/workers/package.json b/packages/workers/package.json index a7b62462..f2fc164c 100644 --- a/packages/workers/package.json +++ b/packages/workers/package.json @@ -26,6 +26,7 @@ "openai": "^4.26.1", "puppeteer": "^22.0.0", "puppeteer-extra": "^3.3.6", + "puppeteer-extra-plugin-adblocker": "^2.13.6", "puppeteer-extra-plugin-stealth": "^2.11.2", "tsx": "^4.7.1", "typescript": "^5", -- cgit v1.2.3-70-g09d2