From dd96819709b4e0eef46eefbd5004ff25f5cdd8cd Mon Sep 17 00:00:00 2001 From: Jet Li Date: Sat, 26 Apr 2025 19:53:44 -0700 Subject: [PATCH] fix(cleaner): update document cleaner job --- apps/content/backend/content/document_cleaner.py | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/apps/content/backend/content/document_cleaner.py b/apps/content/backend/content/document_cleaner.py index 9bc4d30..54feb20 100644 --- a/apps/content/backend/content/document_cleaner.py +++ b/apps/content/backend/content/document_cleaner.py @@ -1,3 +1,4 @@ +from datetime import datetime, timedelta from typing import Set from backend.content.models import DocumentDoc from backend.content.models import ContentFolderDoc @@ -32,11 +33,16 @@ class DocumentCleaner: print( f"Valid document IDs (from updated content directories): {valid_document_ids}" ) + # Calculate the time 1 day ago + time_1_day_ago = datetime.now(datetime.UTC) - timedelta(days=1) # Retrieve all document IDs from DocumentDoc all_document_ids = set() async for doc in DocumentDoc.find( - {"created_by": {"$regex": "^content-service-"}} + { + "created_by": {"$regex": "^content-service-"}, + "create_time": {"$lt": time_1_day_ago}, + } ): # Fetch all documents that are created by content-service if doc.document_id: # Ensure document_id is not None all_document_ids.add(doc.document_id)