apunkt/SurfSense
1
0
Fork 0
mirror of https://github.com/MODSetter/SurfSense.git synced 2026-04-26 01:06:23 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

Pass vision_llm in legacy process_file_in_background path

Browse source
This commit is contained in:
CREDO23 2026-04-09 15:28:21 +02:00
parent 55661bcde6
commit ff2a9c77f9
1 changed files with 12 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

14
surfsense_backend/app/tasks/document_processors/file_processors.py
View file

@ -118,9 +118,13 @@ async def _log_page_divergence(
async def _process_non_document_upload(ctx: _ProcessingContext) -> Document | None:
"""Extract content from a non-document file (plaintext/direct_convert/audio) via the unified ETL pipeline."""
"""Extract content from a non-document file (plaintext/direct_convert/audio/image) via the unified ETL pipeline."""
from app.etl_pipeline.etl_document import EtlRequest
from app.etl_pipeline.etl_pipeline_service import EtlPipelineService
from app.etl_pipeline.file_classifier import (
FileCategory,
classify_file as etl_classify,
)
await _notify(ctx, "parsing", "Processing file")
await ctx.task_logger.log_task_progress(
@ -129,7 +133,13 @@ async def _process_non_document_upload(ctx: _ProcessingContext) -> Document | No
{"processing_stage": "extracting"},
)
etl_result = await EtlPipelineService().extract(
vision_llm = None
if etl_classify(ctx.filename) == FileCategory.IMAGE:
from app.services.llm_service import get_vision_llm
vision_llm = await get_vision_llm(ctx.session, ctx.search_space_id)
etl_result = await EtlPipelineService(vision_llm=vision_llm).extract(
EtlRequest(file_path=ctx.file_path, filename=ctx.filename)
)