@opentermsarchive/engine 10.1.3 → 10.3.0

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@opentermsarchive/engine",
-  "version": "10.1.3",
+  "version": "10.3.0",
   "description": "Tracks and makes visible changes to the terms of online services",
   "homepage": "https://opentermsarchive.org",
   "bugs": {
@@ -113,7 +113,7 @@
     "supertest": "^7.1.4"
   },
   "peerDependencies": {
-    "@opentermsarchive/terms-types": "^2.1.0"
+    "@opentermsarchive/terms-types": "^2.2.0"
   },
   "engines": {
     "node": ">=20 <=24"

package/src/archivist/fetcher/fullDomFetcher.js

@@ -8,6 +8,7 @@ puppeteer.use(stealthPlugin());
 let browser;
 
 export default async function fetch(url, cssSelectors, config) {
+  let context;
   let page;
   let response;
   const selectors = [].concat(cssSelectors);
@@ -17,16 +18,13 @@ export default async function fetch(url, cssSelectors, config) {
   }
 
   try {
-    page = await browser.newPage();
+    context = await browser.createBrowserContext(); // Create an isolated browser context to ensure complete isolation between fetches (cookies, localStorage, sessionStorage, IndexedDB, cache)
+    page = await context.newPage();
 
+    await page.setViewport({ width: 1920, height: 1080 }); // Set a realistic viewport size to avoid detection based on default Puppeteer dimensions (800x600)
     await page.setDefaultNavigationTimeout(config.navigationTimeout);
     await page.setExtraHTTPHeaders({ 'Accept-Language': config.language });
 
-    await page.setCacheEnabled(false); // Disable cache to ensure fresh content on each fetch and prevent stale data from previous requests
-    const client = await page.target().createCDPSession();
-
-    await client.send('Network.clearBrowserCookies'); // Clear cookies to ensure clean state between fetches and prevent session persistence across different URLs
-
     if (browser.proxyCredentials?.username && browser.proxyCredentials?.password) {
       await page.authenticate(browser.proxyCredentials);
     }
@@ -78,6 +76,9 @@ export default async function fetch(url, cssSelectors, config) {
     if (page) {
       await page.close();
     }
+    if (context) {
+      await context.close(); // Close the isolated context to free resources and ensure complete cleanup
+    }
   }
 }