new read function

tools.py

@@ -71,84 +71,29 @@ def get_youtube_transcript(video_url: str) -> str:
     except Exception as e:
         return f"Unavailable. Error: {str(e)}"
     
-    
 
-def read_document(file_spec: str) -> str:
+
+def read_document(file_path: str) -> str:
     """
     file_spec can be:
       - a local file path (e.g., "data.xlsx")
       - a task_id (e.g., "abc123") to download from GAIA
     """
-    import os, requests, tempfile
-    from urllib.parse import quote
-
-    actual_path = None
-
-    # 1. Si es un task_id (solo letras/números, ~10-12 chars, sin .), asumir que es ID
-    if os.path.basename(file_spec).isalnum() and len(file_spec) in range(5, 20) and '.' not in file_spec:
-        # Es probablemente un task_id → descargar
-        download_url = f"https://agents-course-unit4-scoring.hf.space/files/{quote(file_spec)}"
-        print(f"📥 Downloading from: {download_url}")
+    name = "read_file"
+    description = "Reads a file and returns its content"
+    inputs = {
+        "file_path": {"type": "string", "description": "Path to the file to read"},
+    }
+    output_type = "string"
+
+    def forward(self, file_path: str) -> str:
         try:
-            response = requests.get(download_url, timeout=15)
-            response.raise_for_status()
-
-            # Determinar extensión desde Content-Type o por defecto .xlsx
-            content_type = response.headers.get('content-type', '').lower()
-            if 'pdf' in content_type:
-                suffix = '.pdf'
-            elif 'excel' in content_type or 'sheet' in content_type:
-                suffix = '.xlsx'
-            elif 'json' in content_type:
-                suffix = '.json'
-            elif 'text' in content_type or 'csv' in content_type:
-                suffix = '.csv'
-            else:
-                suffix = '.xlsx'  # fallback
-
-            with tempfile.NamedTemporaryFile(delete=False, suffix=suffix) as tmp:
-                tmp.write(response.content)
-                actual_path = tmp.name
-            print(f"✅ Saved as: {actual_path}")
-
+            with open(file_path, "r") as file:
+                return file.read()
         except Exception as e:
-            return f"Download failed for task_id '{file_spec}': {str(e)}"
+            return f"Error reading file: {str(e)}"
 
-    else:
-        # 2. Buscar localmente
-        possible_paths = [file_spec, f"./{file_spec}", f"/tmp/{file_spec}"]
-        for p in possible_paths:
-            if os.path.exists(p):
-                actual_path = p
-                break
 
-    if not actual_path:
-        return f"File not found locally or downloadable: {file_spec}"
-
-    # 3. Leer (tu lógica existente)
-    try:
-        ext = os.path.splitext(actual_path)[1].lower()
-        if ext in ['.mp3', '.wav', '.m4a', '.flac', '.ogg']:
-            transcription = transcribe_audio_openai(actual_path)
-            result = f"Audio transcription: {transcription}"
-        elif ext in ['.txt', '.pdf', '.docx', '.csv', '.json', '.md', '.xlsx']:
-            reader = SimpleDirectoryReader(input_files=[actual_path])
-            docs = reader.load_data()
-            full_text = "\n\n".join(doc.text for doc in docs) if docs else ""
-            result = f"File: {os.path.basename(actual_path)}\n\n{full_text}"
-        else:
-            result = f"Unsupported file type: {ext}"
-    except Exception as e:
-        result = f"Error reading {actual_path}: {str(e)}"
-    finally:
-        # Limpiar temporal
-        if actual_path and actual_path not in possible_paths:
-            try:
-                os.unlink(actual_path)
-            except:
-                pass
-
-    return result