Update app.py

app.py

@@ -177,7 +177,10 @@ async def chat(
 
 def extract_text_from_webpage(html_content):
     """Extracts visible text from HTML content using BeautifulSoup."""
-    return BeautifulSoup(html_content).get_text(strip=True)
+    soup = BeautifulSoup(html_content)
+    for tag in soup(["script", "style", "header", "footer"]):
+        tag.extract()
+    return soup.get_text(strip=True)
 
 async def fetch_and_extract(url, max_chars, proxy: Optional[str] = None):
     """Fetches a URL and extracts text asynchronously."""
@@ -242,7 +245,10 @@ async def web_search_and_extract(
 
 def extract_text_from_webpage2(html_content):
     """Extracts visible text from HTML content using BeautifulSoup."""
-    return BeautifulSoup(html_content).get_text(strip=True)
+    soup = BeautifulSoup(html_content)
+    for tag in soup(["script", "style", "header", "footer"]):
+        tag.extract()
+    return soup.get_text(strip=True)
 
 def fetch_and_extract2(url, max_chars):
     """Fetches a URL and extracts text using threading."""