fix: chroma examples error due to None parsing in flattens (#215)

examples/document-search/chroma_otel.py

@@ -51,7 +51,7 @@
 
 import asyncio
 
-from chromadb import PersistentClient
+from chromadb import EphemeralClient
 from opentelemetry import trace
 from opentelemetry.exporter.otlp.proto.grpc.trace_exporter import OTLPSpanExporter
 from opentelemetry.sdk.resources import SERVICE_NAME, Resource
@@ -102,7 +102,7 @@ async def main() -> None:
         model="text-embedding-3-small",
     )
     vector_store = ChromaVectorStore(
-        client=PersistentClient("./chroma"),
+        client=EphemeralClient(),
         index_name="jokes",
     )
     document_search = DocumentSearch(

packages/ragbits-core/src/ragbits/core/utils/dict_transformations.py

@@ -1,9 +1,9 @@
 from typing import Any
 
+SimpleTypes = str | int | float | bool | None
 
-def flatten_dict(
-    input_dict: dict[str, Any], parent_key: str = "", sep: str = "."
-) -> dict[str, str | int | float | bool]:
+
+def flatten_dict(input_dict: dict[str, Any], parent_key: str = "", sep: str = ".") -> dict[str, SimpleTypes]:
     """
     Recursively flatten a nested dictionary and lists, converting non-primitive types to strings.
 
@@ -18,7 +18,7 @@ def flatten_dict(
     Raises:
         ValueError: If the dictionary cannot be safely flattened due to the presence of the separator in the dict key.
     """
-    items: dict[str, str | int | float | bool] = {}
+    items: dict[str, SimpleTypes] = {}
     for k, v in input_dict.items():
         if sep in k:
             raise ValueError(f"Separator '{sep}' found in key '{parent_key}' Cannot flatten dictionary safely.")
@@ -36,9 +36,9 @@ def flatten_dict(
                 if isinstance(item, dict):
                     items = {**items, **flatten_dict(item, list_key, sep=sep)}
                 else:
-                    items[list_key] = item if isinstance(item, str | int | float | bool) else str(item)
+                    items[list_key] = item if isinstance(item, SimpleTypes) else str(item)
         else:
-            items[new_key] = v if isinstance(v, str | int | float | bool) else str(v)
+            items[new_key] = v if isinstance(v, SimpleTypes) else str(v)
 
     return items
 

packages/ragbits-core/src/ragbits/core/vector_stores/chroma.py

@@ -80,7 +80,7 @@ async def store(self, entries: list[VectorStoreEntry]) -> None:
         metadatas = [entry.metadata for entry in entries]
 
         # Flatten metadata
-        flattened_metadatas = [flatten_dict(metadata) for metadata in metadatas]
+        flattened_metadatas = [self._flatten_metadata(metadata) for metadata in metadatas]
 
         metadatas = (
             flattened_metadatas
@@ -180,3 +180,8 @@ async def list(
             )
             for id, metadata, embedding, document in zip(ids, metadatas, embeddings, documents, strict=True)
         ]
+
+    @staticmethod
+    def _flatten_metadata(metadata: dict) -> dict:
+        """Flattens the metadata dictionary. Removes any None values as they are not supported by ChromaDB."""
+        return {k: v for k, v in flatten_dict(metadata).items() if v is not None}