Update app.py

app.py

@@ -25,78 +25,78 @@ def my_custom_tool(arg1:str, arg2:int)-> str: #it's import to specify the return
 
 #from datasets import load_dataset
 #ds = load_dataset("nfliu/decontextualization")
-def get_page_content(url):
+def get_page_content(page_url):
     """
     Obtiene el contenido de una página web
     """
+    if not page_url or page_url == 'No URL available':
+        return "No content available - invalid URL"
+        
     try:
-        # Añadir headers para evitar bloqueos
         headers = {
             'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36'
         }
-        response = requests.get(url, headers=headers, timeout=10)
+        response = requests.get(page_url, headers=headers, timeout=10)
         response.raise_for_status()
         
-        # Parsear el contenido HTML
         soup = BeautifulSoup(response.text, 'html.parser')
         
         # Eliminar scripts y estilos
         for script in soup(["script", "style"]):
             script.decompose()
             
-        # Obtener el texto
         text = soup.get_text()
-        
-        # Limpiar el texto
         lines = (line.strip() for line in text.splitlines())
         text = ' '.join(chunk for chunk in lines if chunk)
         
-        return text[:1000]  # Devolver los primeros 1000 caracteres
+        return text[:1000]
         
     except Exception as e:
-        return f"Error al acceder a la página: {str(e)}"
-@tool
+        return f"Error accessing the page: {str(e)}"
+
 def answer_nfl_question_tool(question: str) -> str:
-    """
-    A tool that answers the user's questions about the NFL rules, tactics and players.
-    
-    Args:
-        question: The user's question that agent have to lookup the answer using the DuckDuckGo browser
-    """
-    # Prepare the duckduckgo query
-    search_query = f"NFL {question}"
-    print(f"🔍 Searching DuckDuckGo with query: {search_query}")
-    
-    # Search using DuckDuckGo
-    with DDGS() as ddgs:
-        search_results = list(ddgs.text(search_query, max_results=3))
-    
-    print(f"📊 Found {len(search_results)} results")
-    
-    if not search_results:
-        return "❌ We haven't found answers to this question."
-    
-    # Format results
-    detailed_results = []
-    detailed_results.append(f"Search Results for: {search_query}\n")
-    detailed_results.append("=" * 50 + "\n")
-    
-    # Process each search result
-    for i, result in enumerate(search_results, 1):
-        detailed_results.append(f"\nSource {i}: {result.get('link') or result.get('url', 'No URL available')}\n")
-        detailed_results.append("-" * 30 + "\n")
-        detailed_results.append(f"Title: {result.get('title', 'No title available')}\n")
-        detailed_results.append(f"Description: {result.get('snippet', 'No description available')}\n")
-
-    # Obtener y añadir el contenido de la página
-    print(f"📄 Accessing page {i}...")
-    page_content = get_page_content(url)
-    detailed_results.append(f"Content: {page_content}\n")
-    
-    # Añadir un pequeño delay entre peticiones
-    time.sleep(1)
-    
-    return "\n".join(detailed_results)
+    try:
+        search_query = f"NFL {question}"
+        print(f"🔍 Searching DuckDuckGo with query: {search_query}")
+        
+        with DDGS() as ddgs:
+            search_results = list(ddgs.text(search_query, max_results=3))
+        
+        print(f"📊 Found {len(search_results)} results")
+        
+        if not search_results:
+            return "❌ No results found for this question."
+        
+        detailed_results = []
+        detailed_results.append(f"Search Results for: {search_query}\n")
+        detailed_results.append("=" * 50 + "\n")
+        
+        for i, result in enumerate(search_results, 1):
+            # Verificar que result es un diccionario y tiene las claves necesarias
+            if not isinstance(result, dict):
+                continue
+                
+            page_url = result.get('link') or result.get('url')
+            if not page_url:
+                continue
+                
+            title = result.get('title', 'No title available')
+            
+            detailed_results.append(f"\nSource {i}: {page_url}\n")
+            detailed_results.append("-" * 30 + "\n")
+            detailed_results.append(f"Title: {title}\n")
+            
+            print(f"📄 Accessing page {i}...")
+            page_content = get_page_content(page_url)
+            detailed_results.append(f"Content: {page_content}\n")
+            
+            time.sleep(1)
+        
+        return "\n".join(detailed_results)
+        
+    except Exception as e:
+        return f"Error during search: {str(e)}"
+
     
 
 @tool