highperfocused/danswer
1
0
Fork 0
mirror of https://github.com/danswer-ai/danswer.git synced 2025-10-11 05:36:03 +02:00
Code Issues Packages Projects Releases Wiki Activity

Fix Title for docs without (#1827)

Browse Source
This commit is contained in:
Yuhong Sun
2024-07-14 13:51:11 -07:00
committed by GitHub
parent 56b175f597
commit da31da33e7
2 changed files with 5 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
backend/danswer/document_index/vespa/index.py
View File

@@ -605,7 +605,7 @@ def _vespa_hit_to_inference_chunk(
section_continuation=fields[SECTION_CONTINUATION], section_continuation=fields[SECTION_CONTINUATION],
document_id=fields[DOCUMENT_ID], document_id=fields[DOCUMENT_ID],
source_type=fields[SOURCE_TYPE], source_type=fields[SOURCE_TYPE],
title=fields[TITLE], title=fields.get(TITLE),
semantic_identifier=fields[SEMANTIC_IDENTIFIER], semantic_identifier=fields[SEMANTIC_IDENTIFIER],
boost=fields.get(BOOST, 1), boost=fields.get(BOOST, 1),
recency_bias=fields.get("matchfeatures", {}).get(RECENCY_BIAS, 1.0), recency_bias=fields.get("matchfeatures", {}).get(RECENCY_BIAS, 1.0),
@@ -614,7 +614,7 @@ def _vespa_hit_to_inference_chunk(
primary_owners=fields.get(PRIMARY_OWNERS), primary_owners=fields.get(PRIMARY_OWNERS),
secondary_owners=fields.get(SECONDARY_OWNERS), secondary_owners=fields.get(SECONDARY_OWNERS),
metadata=metadata, metadata=metadata,
metadata_suffix=fields.get(METADATA_SUFFIX) or "", metadata_suffix=fields.get(METADATA_SUFFIX),
match_highlights=match_highlights, match_highlights=match_highlights,
updated_at=updated_at, updated_at=updated_at,
) )

5
backend/danswer/search/models.py
View File

@@ -190,11 +190,12 @@ class InferenceChunk(BaseChunk):
class InferenceChunkUncleaned(InferenceChunk): class InferenceChunkUncleaned(InferenceChunk):
title: str # Separate from Semantic Identifier though often same title: str | None # Separate from Semantic Identifier though often same
metadata_suffix: str metadata_suffix: str | None
def to_inference_chunk(self) -> InferenceChunk: def to_inference_chunk(self) -> InferenceChunk:
# Create a dict of all fields except 'title' and 'metadata_suffix' # Create a dict of all fields except 'title' and 'metadata_suffix'
# Assumes the cleaning has already been applied and just needs to translate to the right type
inference_chunk_data = { inference_chunk_data = {
k: v k: v
for k, v in self.dict().items() for k, v in self.dict().items()