apunkt/SurfSense
1
0
Fork 0
mirror of https://github.com/MODSetter/SurfSense.git synced 2026-06-24 21:38:09 +02:00
Code Issues Projects Releases Packages Wiki Activity

feat(tasks): route extraction through etl cache

Browse source
This commit is contained in:
CREDO23 2026-06-12 11:23:50 +02:00
parent 1c05980ffb
commit 0dc2ccc003
1 changed files with 5 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

8
surfsense_backend/app/tasks/document_processors/file_processors.py
View file

@ -381,7 +381,6 @@ async def _extract_file_content(
Tuple of (markdown_content, etl_service_name, billable_pages). Tuple of (markdown_content, etl_service_name, billable_pages).
""" """
from app.etl_pipeline.etl_document import EtlRequest, ProcessingMode from app.etl_pipeline.etl_document import EtlRequest, ProcessingMode
from app.etl_pipeline.etl_pipeline_service import EtlPipelineService
from app.etl_pipeline.file_classifier import ( from app.etl_pipeline.file_classifier import (
FileCategory, FileCategory,
classify_file as etl_classify, classify_file as etl_classify,
@ -432,13 +431,16 @@ async def _extract_file_content(
vision_llm = await get_vision_llm(session, search_space_id) vision_llm = await get_vision_llm(session, search_space_id)
result = await EtlPipelineService(vision_llm=vision_llm).extract( from app.etl_pipeline.cache import extract_with_cache
result = await extract_with_cache(
EtlRequest( EtlRequest(
file_path=file_path, file_path=file_path,
filename=filename, filename=filename,
estimated_pages=estimated_pages, estimated_pages=estimated_pages,
processing_mode=mode, processing_mode=mode,
) ),
vision_llm=vision_llm,
) )
with contextlib.suppress(Exception): with contextlib.suppress(Exception):