gmakstutis/LightRAG
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Remove _id field and improve index handling in MongoDB

Browse source 
- Remove MongoDB _id field from documents
- Improve index existence check and creation
This commit is contained in:
yangdx 2025-07-30 04:17:26 +08:00
parent cfb7117dd6
commit 30f71c8acf
1 changed files with 6 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

7
lightrag/kg/mongo_impl.py
View file

@ -383,6 +383,8 @@ class MongoDocStatusStorage(DocStatusStorage):
data = doc.copy() data = doc.copy()
# Remove deprecated content field if it exists # Remove deprecated content field if it exists
data.pop("content", None) data.pop("content", None)
# Remove MongoDB _id field if it exists
data.pop("_id", None)
# If file_path is not in data, use document id as file path # If file_path is not in data, use document id as file path
if "file_path" not in data: if "file_path" not in data:
data["file_path"] = "no-file-path" data["file_path"] = "no-file-path"
@ -410,6 +412,8 @@ class MongoDocStatusStorage(DocStatusStorage):
data = doc.copy() data = doc.copy()
# Remove deprecated content field if it exists # Remove deprecated content field if it exists
data.pop("content", None) data.pop("content", None)
# Remove MongoDB _id field if it exists
data.pop("_id", None)
# If file_path is not in data, use document id as file path # If file_path is not in data, use document id as file path
if "file_path" not in data: if "file_path" not in data:
data["file_path"] = "no-file-path" data["file_path"] = "no-file-path"
@ -456,7 +460,8 @@ class MongoDocStatusStorage(DocStatusStorage):
"""Create track_id index for better query performance""" """Create track_id index for better query performance"""
try: try:
# Check if index already exists # Check if index already exists
existing_indexes = await self._data.list_indexes().to_list(length=None) indexes_cursor = await self._data.list_indexes()
existing_indexes = await indexes_cursor.to_list(length=None)
track_id_index_exists = any( track_id_index_exists = any(
"track_id" in idx.get("key", {}) for idx in existing_indexes "track_id" in idx.get("key", {}) for idx in existing_indexes
) )