From 367af9512fa47d9bd41fd9822b81031517b809c0 Mon Sep 17 00:00:00 2001
From: rafaelsideguide <150964962+rafaelsideguide@users.noreply.github.com>
Date: Thu, 31 Oct 2024 10:53:47 -0300
Subject: [PATCH] added iframe links to extractLinksFromHTML

---
 apps/api/src/scraper/WebScraper/crawler.ts | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/apps/api/src/scraper/WebScraper/crawler.ts b/apps/api/src/scraper/WebScraper/crawler.ts
index 72a49fd8..1ae166b4 100644
--- a/apps/api/src/scraper/WebScraper/crawler.ts
+++ b/apps/api/src/scraper/WebScraper/crawler.ts
@@ -333,6 +333,16 @@ export class WebCrawler {
       }
     });
 
+    // Extract links from iframes with inline src
+    $("iframe").each((_, element) => {
+      const src = $(element).attr("src");
+      if (src && src.startsWith("data:text/html")) {
+        const iframeHtml = decodeURIComponent(src.split(",")[1]);
+        const iframeLinks = this.extractLinksFromHTML(iframeHtml, url);
+        links = links.concat(iframeLinks);
+      }
+    });
+
     return links;
   }