Merge pull request #883 from YanSte/fix-return-none

Optimised returns

lightrag/kg/chroma_impl.py

@@ -108,9 +108,9 @@ class ChromaVectorDBStorage(BaseVectorStorage):
             raise
 
     async def upsert(self, data: dict[str, dict[str, Any]]) -> None:
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
         if not data:
-            logger.warning("Empty data provided to vector DB")
+            return
-            return []
 
         try:
             ids = list(data.keys())

lightrag/kg/faiss_impl.py

@@ -79,10 +79,9 @@ class FaissVectorDBStorage(BaseVectorStorage):
            ...
         }
         """
-        logger.info(f"Inserting {len(data)} vectors to {self.namespace}")
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
         if not data:
-            logger.warning("You are inserting empty data to the vector DB")
+            return
-            return []
 
         current_time = time.time()
 

lightrag/kg/json_doc_status_impl.py

@@ -58,6 +58,10 @@ class JsonDocStatusStorage(DocStatusStorage):
         write_json(self._data, self._file_name)
 
     async def upsert(self, data: dict[str, dict[str, Any]]) -> None:
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
+        if not data:
+            return
+
         self._data.update(data)
         await self.index_done_callback()
 

lightrag/kg/json_kv_impl.py

@@ -43,6 +43,9 @@ class JsonKVStorage(BaseKVStorage):
         return set(keys) - set(self._data.keys())
 
     async def upsert(self, data: dict[str, dict[str, Any]]) -> None:
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
+        if not data:
+            return
         left_data = {k: v for k, v in data.items() if k not in self._data}
         self._data.update(left_data)
 

lightrag/kg/milvus_impl.py

@@ -75,11 +75,11 @@ class MilvusVectorDBStorage(BaseVectorStorage):
         )
 
     async def upsert(self, data: dict[str, dict[str, Any]]) -> None:
-        logger.info(f"Inserting {len(data)} vectors to {self.namespace}")
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
-        if not len(data):
+        if not data:
-            logger.warning("You insert an empty data to vector DB")
+            return
-            return []
+
-        list_data = [
+        list_data: list[dict[str, Any]] = [
             {
                 "id": k,
                 **{k1: v1 for k1, v1 in v.items() if k1 in self.meta_fields},

lightrag/kg/mongo_impl.py

@@ -108,8 +108,12 @@ class MongoKVStorage(BaseKVStorage):
         return keys - existing_ids
 
     async def upsert(self, data: dict[str, dict[str, Any]]) -> None:
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
+        if not data:
+            return
+
         if is_namespace(self.namespace, NameSpace.KV_STORE_LLM_RESPONSE_CACHE):
-            update_tasks = []
+            update_tasks: list[Any] = []
             for mode, items in data.items():
                 for k, v in items.items():
                     key = f"{mode}_{k}"
@@ -181,7 +185,10 @@ class MongoDocStatusStorage(DocStatusStorage):
         return data - existing_ids
 
     async def upsert(self, data: dict[str, dict[str, Any]]) -> None:
-        update_tasks = []
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
+        if not data:
+            return
+        update_tasks: list[Any] = []
         for k, v in data.items():
             data[k]["_id"] = k
             update_tasks.append(
@@ -855,10 +862,9 @@ class MongoVectorDBStorage(BaseVectorStorage):
             logger.debug("vector index already exist")
 
     async def upsert(self, data: dict[str, dict[str, Any]]) -> None:
-        logger.debug(f"Inserting {len(data)} vectors to {self.namespace}")
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
         if not data:
-            logger.warning("You are inserting an empty data set to vector DB")
+            return
-            return []
 
         list_data = [
             {

lightrag/kg/nano_vector_db_impl.py

@@ -45,10 +45,9 @@ class NanoVectorDBStorage(BaseVectorStorage):
         )
 
     async def upsert(self, data: dict[str, dict[str, Any]]) -> None:
-        logger.info(f"Inserting {len(data)} vectors to {self.namespace}")
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
-        if not len(data):
+        if not data:
-            logger.warning("You insert an empty data to vector DB")
+            return
-            return []
 
         current_time = time.time()
         list_data = [

lightrag/kg/oracle_impl.py

@@ -326,6 +326,10 @@ class OracleKVStorage(BaseKVStorage):
 
     ################ INSERT METHODS ################
     async def upsert(self, data: dict[str, dict[str, Any]]) -> None:
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
+        if not data:
+            return
+
         if is_namespace(self.namespace, NameSpace.KV_STORE_TEXT_CHUNKS):
             list_data = [
                 {

lightrag/kg/postgres_impl.py

@@ -347,6 +347,10 @@ class PGKVStorage(BaseKVStorage):
 
     ################ INSERT METHODS ################
     async def upsert(self, data: dict[str, dict[str, Any]]) -> None:
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
+        if not data:
+            return
+
         if is_namespace(self.namespace, NameSpace.KV_STORE_TEXT_CHUNKS):
             pass
         elif is_namespace(self.namespace, NameSpace.KV_STORE_FULL_DOCS):
@@ -448,10 +452,10 @@ class PGVectorStorage(BaseVectorStorage):
         return upsert_sql, data
 
     async def upsert(self, data: dict[str, dict[str, Any]]) -> None:
-        logger.info(f"Inserting {len(data)} vectors to {self.namespace}")
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
-        if not len(data):
+        if not data:
-            logger.warning("You insert an empty data to vector DB")
+            return
-            return []
+
         current_time = time.time()
         list_data = [
             {
@@ -612,6 +616,10 @@ class PGDocStatusStorage(DocStatusStorage):
         Args:
             data: dictionary of document IDs and their status data
         """
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
+        if not data:
+            return
+
         sql = """insert into LIGHTRAG_DOC_STATUS(workspace,id,content,content_summary,content_length,chunks_count,status)
                  values($1,$2,$3,$4,$5,$6,$7)
                   on conflict(id,workspace) do update set

lightrag/kg/qdrant_impl.py

@@ -87,9 +87,9 @@ class QdrantVectorDBStorage(BaseVectorStorage):
         )
 
     async def upsert(self, data: dict[str, dict[str, Any]]) -> None:
-        if not len(data):
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
-            logger.warning("You insert an empty data to vector DB")
+        if not data:
-            return []
+            return
         list_data = [
             {
                 "id": k,

lightrag/kg/redis_impl.py

@@ -49,6 +49,9 @@ class RedisKVStorage(BaseKVStorage):
         return set(keys) - existing_ids
 
     async def upsert(self, data: dict[str, dict[str, Any]]) -> None:
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
+        if not data:
+            return
         pipe = self._redis.pipeline()
 
         for k, v in data.items():

lightrag/kg/tidb_impl.py

@@ -211,6 +211,9 @@ class TiDBKVStorage(BaseKVStorage):
 
     ################ INSERT full_doc AND chunks ################
     async def upsert(self, data: dict[str, dict[str, Any]]) -> None:
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
+        if not data:
+            return
         left_data = {k: v for k, v in data.items() if k not in self._data}
         self._data.update(left_data)
         if is_namespace(self.namespace, NameSpace.KV_STORE_TEXT_CHUNKS):
@@ -318,12 +321,12 @@ class TiDBVectorDBStorage(BaseVectorStorage):
 
     ###### INSERT entities And relationships ######
     async def upsert(self, data: dict[str, dict[str, Any]]) -> None:
-        # ignore, upsert in TiDBKVStorage already
+        logger.info(f"Inserting {len(data)} to {self.namespace}")
-        if not len(data):
+        if not data:
-            logger.warning("You insert an empty data to vector DB")
+            return
-            return []
         if is_namespace(self.namespace, NameSpace.VECTOR_STORE_CHUNKS):
-            return []
+            return
+
         logger.info(f"Inserting {len(data)} vectors to {self.namespace}")
 
         list_data = [