refresh function corrected, deleting document working with opensearch (…

…run-llama#7664)
Arize-ai · Sep 25, 2023 · 33b49de · 33b49de
1 parent 0b016a6
commit 33b49de
Show file tree

Hide file tree

Showing 3 changed files with 16 additions and 6 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,5 +1,10 @@
 # ChangeLog
 
+## Unreleased
+
+### Bug Fixes / Nits
+- Fixed `refresh_ref_docs()` bug with order of operations (#7664)
+
 ## [0.8.33] - 2023-09-25
 
 ### New Features

diff --git a/llama_index/indices/base.py b/llama_index/indices/base.py
@@ -275,7 +275,9 @@ def update_ref_doc(self, document: Document, **update_kwargs: Any) -> None:
         """
         with self._service_context.callback_manager.as_trace("update"):
             self.delete_ref_doc(
-                document.get_doc_id(), **update_kwargs.pop("delete_kwargs", {})
+                document.get_doc_id(),
+                delete_from_docstore=True,
+                **update_kwargs.pop("delete_kwargs", {}),
             )
             self.insert(document, **update_kwargs.pop("insert_kwargs", {}))
 
@@ -309,14 +311,14 @@ def refresh_ref_docs(
                 existing_doc_hash = self._docstore.get_document_hash(
                     document.get_doc_id()
                 )
-                if existing_doc_hash != document.hash:
+                if existing_doc_hash is None:
+                    self.insert(document, **update_kwargs.pop("insert_kwargs", {}))
+                    refreshed_documents[i] = True
+                elif existing_doc_hash != document.hash:
                     self.update_ref_doc(
                         document, **update_kwargs.pop("update_kwargs", {})
                     )
                     refreshed_documents[i] = True
-                elif existing_doc_hash is None:
-                    self.insert(document, **update_kwargs.pop("insert_kwargs", {}))
-                    refreshed_documents[i] = True
 
             return refreshed_documents
 

diff --git a/llama_index/indices/vector_store/base.py b/llama_index/indices/vector_store/base.py
@@ -3,7 +3,7 @@
 An index that that is built on top of an existing vector store.
 
 """
-
+import logging
 from typing import Any, Dict, List, Optional, Sequence
 
 from llama_index.async_utils import run_async_tasks
@@ -17,6 +17,8 @@
 from llama_index.storage.storage_context import StorageContext
 from llama_index.vector_stores.types import VectorStore
 
+logger = logging.getLogger(__name__)
+
 
 class VectorStoreIndex(BaseIndex[IndexDict]):
     """Vector Store Index.
@@ -281,6 +283,7 @@ def delete_ref_doc(
             if ref_doc_info is not None:
                 for node_id in ref_doc_info.node_ids:
                     self._index_struct.delete(node_id)
+                    self._vector_store.delete(node_id)
 
         # delete from docstore only if needed
         if (