Remove content fallback and standardize doc status handling

- Remove content_summary fallback logic
- Standardize doc status processing
- Handle missing file_path consistently

lightrag/kg/json_doc_status_impl.py

@@ -95,9 +95,6 @@ class JsonDocStatusStorage(DocStatusStorage):
                     try:
                         # Make a copy of the data to avoid modifying the original
                         data = v.copy()
-                        # If content is missing, use content_summary as content
-                        if "content" not in data and "content_summary" in data:
-                            data["content"] = data["content_summary"]
                         # If file_path is not in data, use document id as file path
                         if "file_path" not in data:
                             data["file_path"] = "no-file-path"

lightrag/kg/mongo_impl.py

@@ -372,19 +372,19 @@ class MongoDocStatusStorage(DocStatusStorage):
         """Get all documents with a specific status"""
         cursor = self._data.find({"status": status.value})
         result = await cursor.to_list()
-        return {
+        processed_result = {}
-            doc["_id"]: DocProcessingStatus(
+        for doc in result:
-                content_summary=doc.get("content_summary"),
+            try:
-                content_length=doc["content_length"],
+                # Make a copy of the data to avoid modifying the original
-                file_path=doc.get("file_path", doc["_id"]),
+                data = doc.copy()
-                status=doc["status"],
+                # If file_path is not in data, use document id as file path
-                created_at=doc.get("created_at"),
+                if "file_path" not in data:
-                updated_at=doc.get("updated_at"),
+                    data["file_path"] = "no-file-path"
-                chunks_count=doc.get("chunks_count", -1),
+                processed_result[doc["_id"]] = DocProcessingStatus(**data)
-                chunks_list=doc.get("chunks_list", []),
+            except KeyError as e:
-            )
+                logger.error(f"Missing required field for document {doc['_id']}: {e}")
-            for doc in result
+                continue
-        }
+        return processed_result
 
     async def index_done_callback(self) -> None:
         # Mongo handles persistence automatically