Update functions.py

functions.py

@@ -1,7 +1,14 @@
-# functions.py
 import json
 import requests
+from datasets import load_dataset
 
+# Cargar modalidades y tareas (ahora devuelve una lista de tareas) <button class="citation-flag" data-index="1">
+def cargar_modalidades_tareas():
+    with open("modalidades_tareas.json", "r") as file:
+        MODALIDAD_TAREAS = json.load(file)
+    return MODALIDAD_TAREAS
+
+# Función para actualizar el JSON desde HuggingFace y devolver tareas
 def actualizar_modalidades_tareas_desde_huggingface():
     MODALIDADES = ["text-classification", "token-classification", "translation", "summarization",
                    "speech-recognition", "audio-classification", "music-classification",
@@ -10,11 +17,10 @@ def actualizar_modalidades_tareas_desde_huggingface():
     
     modalidades_tareas = {}
     for task in MODALIDADES:
-        url = f"https://huggingface.co/api/datasets?task={task}&limit=5"  # Buscar datasets por tarea <button class="citation-flag" data-index="1">
+        url = f"https://huggingface.co/api/datasets?task={task}&limit=5"
         response = requests.get(url)
         datasets = response.json()
         
-        # Extraer columnas comunes de los datasets
         columnas = set()
         for dataset in datasets:
             try:
@@ -23,26 +29,20 @@ def actualizar_modalidades_tareas_desde_huggingface():
             except:
                 continue
         
-        # Asignar a modalidades
-        if task in ["text-classification", "token-classification", "translation", "summarization"]:
-            modalidad = "NLP"
-        elif task in ["speech-recognition", "audio-classification", "music-classification"]:
-            modalidad = "Audio"
-        elif task in ["image-classification", "object-detection", "image-segmentation"]:
-            modalidad = "Vision"
-        elif task in ["question-answering", "information-retrieval"]:
-            modalidad = "RAG"
+        modalidad = "NLP" if task in ["text-classification", "token-classification", "translation", "summarization"] else \
+                   "Audio" if task in ["speech-recognition", "audio-classification", "music-classification"] else \
+                   "Vision" if task in ["image-classification", "object-detection", "image-segmentation"] else \
+                   "RAG"
         
-        # Añadir al diccionario
         if modalidad not in modalidades_tareas:
             modalidades_tareas[modalidad] = {"tareas": {}, "columnas_generales": []}
         modalidades_tareas[modalidad]["tareas"][task] = list(columnas)
     
-    # Guardar JSON
     with open("modalidades_tareas.json", "w") as file:
         json.dump(modalidades_tareas, file, indent=4)
     
-    return "Archivo actualizado con tareas y columnas de HuggingFace."
+    # Devolver lista de tareas actualizada
+    return [t for datos in modalidades_tareas.values() for t in datos["tareas"].keys()]
 
     
 # Cargar modalidades y tareas