union in-memory seed dataset names

Author: Behnam Ousat

pyrit/backend/services/dataset_service.py

@@ -4,7 +4,7 @@
 """
 Dataset service for listing seed datasets.
 
-Wraps ``SeedDatasetProvider`` discovery so the API can list available datasets.
+Wraps ``SeedDatasetProvider`` discovery and memory to list available datasets.
 """
 
 import logging
@@ -15,21 +15,31 @@
     DatasetListResponse,
 )
 from pyrit.datasets import SeedDatasetProvider
+from pyrit.memory import CentralMemory
 
 logger = logging.getLogger(__name__)
 
 
 class DatasetService:
     """Service for listing seed datasets."""
 
+    def __init__(self) -> None:
+        """Initialize the dataset service."""
+        self._memory = CentralMemory.get_memory_instance()
+
     async def list_datasets_async(self) -> DatasetListResponse:
         """
         List all available datasets.
 
+        Combines datasets discoverable via registered providers with those
+        already loaded into memory, since both are available for use.
+
         Returns:
             DatasetListResponse: Available datasets.
         """
-        available = await SeedDatasetProvider.get_all_dataset_names_async()
+        provider_names = await SeedDatasetProvider.get_all_dataset_names_async()
+        memory_names = self._memory.get_seed_dataset_names()
+        available = sorted(set(provider_names) | set(memory_names))
         items = [DatasetInfo(name=name) for name in available]
         return DatasetListResponse(items=items)