Skip to content

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

[FIX] Prompt Studio V2 stabilization fixes (#1237)

harini-venkataramanb740bfab7713
diff --git a/backend/prompt_studio/prompt_studio_core_v2/prompt_studio_helper.py b/backend/prompt_studio/prompt_studio_core_v2/prompt_studio_helper.py
index 4238c0936..a94eca4f7 100644
--- a/backend/prompt_studio/prompt_studio_core_v2/prompt_studio_helper.py
+++ b/backend/prompt_studio/prompt_studio_core_v2/prompt_studio_helper.py
@@ -377,6 +377,13 @@ class PromptStudioHelper:
fs=fs_instance,
tool=util,
)
+ if DocumentIndexingService.is_document_indexing(
+ org_id=org_id, user_id=user_id, doc_id_key=doc_id
+ ):
+ return {
+ "status": IndexingStatus.PENDING_STATUS.value,
+ "output": IndexingStatus.DOCUMENT_BEING_INDEXED.value,
+ }
extracted_text = PromptStudioHelper.dynamic_extractor(
profile_manager=default_profile,
file_path=file_path,
@@ -827,6 +834,13 @@ class PromptStudioHelper:
fs=fs_instance,
tool=util,
)
+ if DocumentIndexingService.is_document_indexing(
+ org_id=org_id, user_id=user_id, doc_id_key=doc_id
+ ):
+ return {
+ "status": IndexingStatus.PENDING_STATUS.value,
+ "output": IndexingStatus.DOCUMENT_BEING_INDEXED.value,
+ }
logger.info(f"Extracting text from {file_path} for {doc_id}")
extracted_text = PromptStudioHelper.dynamic_extractor(
profile_manager=profile_manager,
diff --git a/tools/structure/src/main.py b/tools/structure/src/main.py
index d87851dc6..ba1da29eb 100644
--- a/tools/structure/src/main.py
+++ b/tools/structure/src/main.py
@@ -179,7 +179,11 @@ class StructureTool(BaseTool):
payload[SettingsKeys.OUTPUTS] = outputs
payload[SettingsKeys.FILE_HASH] = summarize_file_hash
break
- if reindex or not summarize_as_source:
+ if (
+ reindex
+ or not summarize_as_source
+ and not output[SettingsKeys.CHUNK_SIZE] == 0
+ ):
self.stream_log("Sucessfully extracted text, indexing..")
STHelper.dynamic_indexing(
tool_settings=tool_settings,