PyPI - sunholo - Versions diffs - 0.70.2__tar.gz → 0.70.4__tar.gz - Mend

@@ -62,7 +62,7 @@ def chunk_doc_to_docs(documents: list, extension: str = ".md", min_size: int = 8
         if len(content) < min_size:
             combined_documents_content += content + "\n"
-            log.info(f"Appending document as its smaller than {min_size}: length {len(content)} - appended doc length {len(combined_documents_content)}")
+            log.debug(f"Appending document as its smaller than {min_size}: length {len(content)} - appended doc length {len(combined_documents_content)}")
         else:
             if combined_documents_content:
                 combined_documents.append(Document(page_content=combined_documents_content, metadata=document.metadata))
@@ -81,7 +81,7 @@ def chunk_doc_to_docs(documents: list, extension: str = ".md", min_size: int = 8
             # If a chunk is smaller than the min_size, append it to temporary_chunk with a line break and continue
             if len(chunk) < min_size:
                 temporary_chunk += chunk + "\n"
-                log.info(f"Appending chunk as its smaller than {min_size}: length {len(chunk)}")
+                log.debug(f"Appending chunk as its smaller than {min_size}: length {len(chunk)}")
                 continue
             # If there's content in temporary_chunk, append it to the current chunk
@@ -92,7 +92,7 @@ def chunk_doc_to_docs(documents: list, extension: str = ".md", min_size: int = 8
             # If the combined chunk is still less than the min_size, append to temporary_chunk with a line break and continue
             if len(chunk) < min_size:
                 temporary_chunk += chunk + "\n"
-                log.info(f"Appending chunk as its smaller than {min_size}: length {len(chunk)}")
+                log.debug(f"Appending chunk as its smaller than {min_size}: length {len(chunk)}")
                 continue
             log.info(f"Adding chunk of length {len(chunk)}")

@@ -1,9 +1,9 @@
 Metadata-Version: 2.1
 Name: sunholo
-Version: 0.70.2
+Version: 0.70.4
 Summary: Large Language Model DevOps - a package to help deploy LLMs to the Cloud.
 Home-page: https://github.com/sunholo-data/sunholo-py
-Download-URL: https://github.com/sunholo-data/sunholo-py/archive/refs/tags/v0.70.2.tar.gz
+Download-URL: https://github.com/sunholo-data/sunholo-py/archive/refs/tags/v0.70.4.tar.gz
 Author: Holosun ApS
 Author-email: multivac@sunholo.com
 License: Apache License, Version 2.0

@@ -1,7 +1,7 @@
 from setuptools import setup, find_packages
 # Define your base version
-version = '0.70.2'
+version = '0.70.4'
 setup(
     name='sunholo',

@@ -310,9 +310,13 @@ class DiscoveryEngineClient:
         try:
             operation = import_documents_with_retry(self.doc_client, request)
         except ResourceExhausted as e:
-            log.error(f"Operation failed after retries due to quota exceeded: {e}")
+            log.error(f"DiscoveryEngine Operation failed after retries due to quota exceeded: {e}")
+            raise e
         except Exception as e:
-            log.error(f"An unexpected error occurred: {e}")
+            log.error(f"An unexpected DiscoveryEngine error occurred: {e}")
+            raise e
         return operation.operation.name

@@ -142,8 +142,11 @@ def embed_pubsub_chunk(data: dict):
                             continue
                 vectorstore_obj = pick_vectorstore(vectorstore, vector_name=vector_name, embeddings=embeddings)
-                vs_retriever = vectorstore_obj.as_retriever(search_kwargs=dict(k=3))
-                vectorstore_list.append(vs_retriever)
+                if vectorstore_obj:
+                    vs_retriever = vectorstore_obj.as_retriever(search_kwargs=dict(k=3))
+                    vectorstore_list.append(vs_retriever)
+                else:
+                    log.warning(f"No vectorstore added for {vectorstore}")
     # can have multiple vectorstores per embed
     metadata_list = []

@@ -32,11 +32,8 @@ def process_pubsub_message(data: dict) -> tuple:
     messageId = data['message'].get('messageId')
     publishTime = data['message'].get('publishTime')
     vector_name = attributes.get('namespace', None)
-    if vector_name is None:
-        log.warning(f"Did not find key vector_name within attributes: {attributes}")
-    # to show we found it
-    attributes['vector_name'] = vector_name
+    if vector_name:
+        attributes['vector_name'] = vector_name
     log.info(f"Process Pub/Sub was triggered by messageId {messageId} published at {publishTime}")
     log.debug(f"Process Pub/Sub data: {message_data}")

sunholo 0.70.2__tar.gz → 0.70.4__tar.gz

sunholo 0.70.2tar.gz → 0.70.4tar.gz