Skip to content

LLM-Driven Extraction of Unstructured Data — Built for API Deployments & ETL Pipeline Workflows

Modified the Index Document Method to Always Perform Hard Reindex (#303)

Tahier Hussain62dedc6219de
diff --git a/backend/prompt_studio/prompt_studio_core/prompt_studio_helper.py b/backend/prompt_studio/prompt_studio_core/prompt_studio_helper.py
index 8f8849ce4..1b1106a71 100644
--- a/backend/prompt_studio/prompt_studio_core/prompt_studio_helper.py
+++ b/backend/prompt_studio/prompt_studio_core/prompt_studio_helper.py
@@ -278,6 +278,7 @@ class PromptStudioHelper:
org_id=org_id,
document_id=document_id,
is_summary=is_summary,
+ reindex=True,
)
logger.info(f"[{tool_id}] Indexing successful for doc: {file_name}")
@@ -627,6 +628,7 @@ class PromptStudioHelper:
org_id: str,
document_id: str,
is_summary: bool = False,
+ reindex: bool = False,
) -> str:
"""Used to index a file based on the passed arguments.
@@ -681,7 +683,7 @@ class PromptStudioHelper:
file_hash=file_hash,
chunk_size=profile_manager.chunk_size,
chunk_overlap=profile_manager.chunk_overlap,
- reindex=profile_manager.reindex,
+ reindex=reindex,
output_file_path=extract_file_path,
)