vikarti.anatra/main
1
0
Fork 0
mirror of https://github.com/manualdousuario/marreta.git synced 2026-04-29 12:00:16 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

adicionada regra removeElementsByTag

Browse source
This commit is contained in:
Renan Bernordi 2025-01-23 16:48:47 -03:00
parent 2444c30a3e
commit d4ab2c4cd8
3 changed files with 25 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

9
app/data/domain_rules.php
View file

@ -32,6 +32,8 @@
* - fetchFromSelenium: Use Selenium for extraction / Usa Selenium para extração
* - socialReferrers: Add random social media headers / Adiciona headers randomicos de redes sociais
* - fromGoogleBot: Adds simulation of request coming from Google Bot / Adiciona simulação de requisição vinda do Google Bot
* - removeElementsByTag: Remove specific elements via DOM / Remove elementos especificos via DOM
* - removeCustomAttr: Remove custom attributes from elements / Remove custom attributes from elements
*/
return [
'nsctotal.com.br' => [
@ -39,8 +41,11 @@ return [
],
'elcorreo.com' => [
'idElementRemove' => ['didomi-popup','engagement-top'],
'classAttrRemove' => ['didomi-popup-open'],
'fromGoogleBot' => true
'classElementRemove' => ['content-exclusive-bg'],
'classAttrRemove' => ['didomi-popup-open','paywall'],
'fromGoogleBot' => true,
'removeElementsByTag' => ['style'],
'removeCustomAttr' => ['hidden','data-*']
],
'globo.com' => [
'idElementRemove' => ['cookie-banner-lgpd', 'paywall-cpt', 'mc-read-more-wrapper', 'paywall-cookie-content', 'paywall-cpt'],