GeneralBots/botserver
2
1
Fork 0
Code Issues Pull requests Projects Releases 323 Packages Wiki Activity Actions

Remove images and CSS from crawled pages
Some checks failed
GBCI / build (push) Failing after 16m12s
Details

Browse source
This commit is contained in:
Rodrigo Rodriguez (Pragmatismo) 2025-09-26 14:47:13 -03:00
parent d6b6d4c09f
commit c9612fcf23
1 changed files with 17 additions and 8 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

25
packages/kb.gbapp/services/KBService.ts
View file

@ -1208,14 +1208,23 @@ export class KBService implements IGBKBService {
await min.core['setConfig'](min, 'Color1', colors[0].hex()); await min.core['setConfig'](min, 'Color1', colors[0].hex());
await min.core['setConfig'](min, 'Color2', colors[1].hex()); await min.core['setConfig'](min, 'Color2', colors[1].hex());
} }
// Disables images in crawling.
await page.route('**/*', route => { // Remove images and disable CSS after page loads
const type = route.request().resourceType(); await page.evaluate(() => {
if (type === 'image' || type === 'stylesheet') { // Remove all images
route.abort(); document.querySelectorAll('img').forEach(img => img.remove());
} else {
route.continue(); // Disable all stylesheets
} document.querySelectorAll('link[rel="stylesheet"]').forEach(link => {
link['disabled'] = true;
link.remove();
});
// Remove inline styles
document.querySelectorAll('style').forEach(style => style.remove());
// Remove style attributes
document.querySelectorAll('[style]').forEach(el => el.removeAttribute('style'));
}); });
page.on('dialog', async dialog => { page.on('dialog', async dialog => {