Update app.py

app.py

@@ -1,23 +1,15 @@
-from smolagents import CodeAgent,DuckDuckGoSearchTool, HfApiModel,load_tool,tool
+from smolagents import CodeAgent, HfApiModel, tool
 import datetime
 import requests
 import pytz
 import yaml
+import os
+import pickle
+import time
+import gradio as gr
 from tools.final_answer import FinalAnswerTool
 
-from Gradio_UI import GradioUI
-
-# Below is an example of a tool that does nothing. Amaze us with your creativity !
-@tool
-def my_custom_tool(arg1:str, arg2:int)-> str: #it's import to specify the return type
-    #Keep this format for the description / args / args description but feel free to modify the tool
-    """A tool that does nothing yet 
-    Args:
-        arg1: the first argument
-        arg2: the second argument
-    """
-    return "What magic will you build ?"
-
+# Herramienta para obtener la hora actual en una zona horaria
 @tool
 def get_current_time_in_timezone(timezone: str) -> str:
     """A tool that fetches the current local time in a specified timezone.
@@ -33,30 +25,214 @@ def get_current_time_in_timezone(timezone: str) -> str:
     except Exception as e:
         return f"Error fetching time for timezone '{timezone}': {str(e)}"
 
+# Herramienta para reconocer canciones usando AudD
+@tool
+def recognize_song(audio_path: str) -> dict:
+    """Reconoce una canción a partir de un archivo de audio
+    Args:
+        audio_path: ruta al archivo de audio a reconocer
+    """
+    AUDD_API_TOKEN = os.getenv("AUDD_API_TOKEN")
+    
+    if not os.path.exists(audio_path):
+        return {"error": "El archivo de audio no existe"}
+    
+    try:
+        with open(audio_path, 'rb') as file:
+            data = {
+                'api_token': AUDD_API_TOKEN,
+                'return': 'spotify,apple_music'
+            }
+            files = {
+                'file': file
+            }
+            response = requests.post('https://api.audd.io/', data=data, files=files)
+            
+        if response.status_code != 200:
+            return {"error": f"Error en la API: {response.status_code}"}
+            
+        result = response.json()
+        
+        if result['status'] == 'error':
+            return {"error": result['error']['error_message']}
+            
+        if not result.get('result'):
+            return {"error": "No se pudo reconocer la canción"}
+            
+        song_info = result['result']
+        
+        return {
+            "🎶 Canción": song_info.get('title', 'Desconocido'),
+            "🎤 Artista": song_info.get('artist', 'Desconocido'),
+            "📀 Álbum": song_info.get('album', 'Desconocido'),
+            "🎧 Spotify": song_info.get('spotify', {}).get('external_urls', {}).get('spotify', 'No disponible'),
+            "🍏 Apple Music": song_info.get('apple_music', {}).get('url', 'No disponible')
+        }
+        
+    except Exception as e:
+        return {"error": f"Error al procesar el audio: {str(e)}"}
 
-final_answer = FinalAnswerTool()
+# Herramienta para iniciar sesión en Soundeo
+@tool
+def login_soundeo() -> str:
+    """Inicia sesión en Soundeo usando las credenciales almacenadas en secrets"""
+    from selenium import webdriver
+    from selenium.webdriver.chrome.options import Options
+    from selenium.webdriver.common.by import By
+    from selenium.webdriver.support.ui import WebDriverWait
+    from selenium.webdriver.support import expected_conditions as EC
+    
+    # Obtener credenciales de los secrets
+    username = os.getenv("SOUNDEO_USERNAME")
+    password = os.getenv("SOUNDEO_PASSWORD")
+    
+    if not username or not password:
+        return "❌ No se encontraron las credenciales en los secrets"
+    
+    # Configurar opciones de Chrome
+    chrome_options = Options()
+    chrome_options.add_argument("--headless")
+    chrome_options.add_argument("--no-sandbox")
+    chrome_options.add_argument("--disable-dev-shm-usage")
+    chrome_options.add_argument("--disable-gpu")
+    
+    try:
+        # Inicializar el driver
+        driver = webdriver.Chrome(options=chrome_options)
+        
+        # Navegar a Soundeo
+        driver.get("https://soundeo.com/login")
+        time.sleep(3)  # Esperar a que cargue la página
+        
+        # Capturar screenshot para debug
+        driver.save_screenshot("pre_login.png")
+        
+        # Completar formulario de login
+        username_field = WebDriverWait(driver, 10).until(
+            EC.presence_of_element_located((By.ID, "username"))
+        )
+        username_field.send_keys(username)
+        
+        password_field = driver.find_element(By.ID, "password")
+        password_field.send_keys(password)
+        
+        # Hacer clic en el botón de login
+        login_button = driver.find_element(By.XPATH, "//button[@type='submit']")
+        login_button.click()
+        
+        # Esperar a que se complete el login
+        time.sleep(5)
+        
+        # Capturar screenshot para verificar
+        driver.save_screenshot("post_login.png")
+        
+        # Guardar cookies para uso futuro
+        pickle.dump(driver.get_cookies(), open("soundeo_cookies.pkl", "wb"))
+        
+        # Verificar si el login fue exitoso
+        if "dashboard" in driver.current_url or "account" in driver.current_url:
+            result = "✅ Login exitoso en Soundeo"
+        else:
+            result = "❌ Error al iniciar sesión. Verifica tus credenciales."
+        
+        driver.quit()
+        return result
+        
+    except Exception as e:
+        return f"❌ Error con Selenium: {str(e)}"
 
-# If the agent does not answer, the model is overloaded, please use another model or the following Hugging Face Endpoint that also contains qwen2.5 coder:
-# model_id='https://pflgm2locj2t89co.us-east-1.aws.endpoints.huggingface.cloud' 
+# Herramienta para buscar y añadir una canción a la lista de descarga
+@tool
+def add_song_to_download_list(song_title: str, artist: str) -> str:
+    """Busca una canción en Soundeo y la añade a la lista de descarga
+    Args:
+        song_title: título de la canción
+        artist: nombre del artista
+    """
+    from selenium import webdriver
+    from selenium.webdriver.chrome.options import Options
+    from selenium.webdriver.common.by import By
+    from selenium.webdriver.support.ui import WebDriverWait
+    from selenium.webdriver.support import expected_conditions as EC
+    
+    chrome_options = Options()
+    chrome_options.add_argument("--headless")
+    chrome_options.add_argument("--no-sandbox")
+    chrome_options.add_argument("--disable-dev-shm-usage")
+    chrome_options.add_argument("--disable-gpu")
+    
+    try:
+        driver = webdriver.Chrome(options=chrome_options)
+        
+        # Cargar cookies si existen
+        cookies_file = "soundeo_cookies.pkl"
+        if os.path.exists(cookies_file):
+            driver.get("https://soundeo.com")
+            cookies = pickle.load(open(cookies_file, "rb"))
+            for cookie in cookies:
+                driver.add_cookie(cookie)
+        else:
+            driver.quit()
+            return "❌ No hay cookies de sesión. Primero debes iniciar sesión."
+        
+        # Navegar a la búsqueda
+        search_query = f"{song_title} {artist}"
+        search_query = search_query.replace(" ", "+")
+        driver.get(f"https://soundeo.com/search?q={search_query}")
+        time.sleep(5)
+        
+        # Capturar screenshot para debug
+        driver.save_screenshot("search_results.png")
+        
+        # Buscar el primer resultado
+        try:
+            first_result = WebDriverWait(driver, 10).until(
+                EC.presence_of_element_located((By.CSS_SELECTOR, ".track-item"))
+            )
+            
+            # Hacer clic en "Add to Download List"
+            download_button = first_result.find_element(By.CSS_SELECTOR, ".download-button")
+            download_button.click()
+            
+            time.sleep(3)
+            driver.save_screenshot("added_to_list.png")
+            
+            # Obtener información del track añadido
+            track_info = first_result.find_element(By.CSS_SELECTOR, ".track-title").text
+            artist_info = first_result.find_element(By.CSS_SELECTOR, ".track-artist").text
+            
+            driver.quit()
+            return f"✅ Añadida a la lista de descarga: {track_info} - {artist_info}"
+            
+        except Exception as e:
+            driver.quit()
+            return f"❌ No se encontraron resultados o no se pudo añadir: {str(e)}"
+            
+    except Exception as e:
+        return f"❌ Error con Selenium: {str(e)}"
 
+# Configuración del agente
+final_answer = FinalAnswerTool()
 model = HfApiModel(
-max_tokens=2096,
-temperature=0.5,
-model_id='Qwen/Qwen2.5-Coder-32B-Instruct',# it is possible that this model may be overloaded
-custom_role_conversions=None,
+    max_tokens=2096,
+    temperature=0.5,
+    model_id='Qwen/Qwen2.5-Coder-32B-Instruct',
+    custom_role_conversions=None,
 )
 
-
-# Import tool from Hub
-image_generation_tool = load_tool("agents-course/text-to-image", trust_remote_code=True)
-
 with open("prompts.yaml", 'r') as stream:
     prompt_templates = yaml.safe_load(stream)
-    
+
 agent = CodeAgent(
     model=model,
-    tools=[final_answer], ## add your tools here (don't remove final answer)
-    max_steps=6,
+    tools=[
+        final_answer,
+        recognize_song,
+        login_soundeo,
+        add_song_to_download_list,
+        get_current_time_in_timezone
+    ],
+    max_steps=8,
     verbosity_level=1,
     grammar=None,
     planning_interval=None,
@@ -65,5 +241,57 @@ agent = CodeAgent(
     prompt_templates=prompt_templates
 )
 
+# Interfaz de usuario con Gradio
+with gr.Blocks() as demo:
+    gr.Markdown("# 🎵 Asistente Musical - Reconocimiento y Descarga")
+    
+    with gr.Tab("Reconocimiento de Música"):
+        with gr.Row():
+            audio_input = gr.Audio(type="filepath", label="Sube o graba un fragmento de audio")
+        
+        query_input = gr.Textbox(
+            label="Consulta (opcional)", 
+            placeholder="Reconoce esta canción y agrégala a mi lista de descarga"
+        )
+        submit_btn = gr.Button("Procesar")
+        output = gr.Markdown()
+        
+        def process_with_agent(audio=None, text_query=""):
+            context = {}
+            
+            if audio:
+                context["audio_path"] = audio
+                if not text_query:
+                    text_query = "Reconoce esta canción y dame información detallada sobre ella."
+            
+            if not text_query:
+                return "❌ Por favor ingresa una consulta o sube un archivo de audio."
+            
+            try:
+                result = agent.run(text_query, context=context)
+                return result
+            except Exception as e:
+                return f"❌ Error al procesar la solicitud: {str(e)}"
+        
+        submit_btn.click(
+            fn=process_with_agent,
+            inputs=[audio_input, query_input],
+            outputs=output
+        )
+    
+    with gr.Tab("Login a Soundeo"):
+        info_text = gr.Markdown("Las credenciales se cargarán automáticamente desde los secrets")
+        login_btn = gr.Button("Iniciar Sesión en Soundeo")
+        login_output = gr.Markdown()
+        
+        def login_process():
+            query = "Inicia sesión en Soundeo usando las credenciales almacenadas"
+            return agent.run(query)
+        
+        login_btn.click(
+            fn=login_process,
+            inputs=[],
+            outputs=login_output
+        )
 
-GradioUI(agent).launch()
+demo.launch()