estel/dify
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

Feat/support parent child chunk (#12092)

Browse Source
This commit is contained in:
Jyong
2024-12-25 19:49:07 +08:00
committed by GitHub
parent 017d7538ae
commit 9231fdbf4c
54 changed files with 2578 additions and 808 deletions
Download Patch File Download Diff File Expand all files Collapse all files

32
api/core/rag/retrieval/dataset_retrieval.py
View File

@@ -166,43 +166,29 @@ class DatasetRetrieval:
"content": item.page_content,
}
retrieval_resource_list.append(source)
document_score_list = {}
# deal with dify documents
if dify_documents:
for item in dify_documents:
if item.metadata.get("score"):
document_score_list[item.metadata["doc_id"]] = item.metadata["score"]
index_node_ids = [document.metadata["doc_id"] for document in dify_documents]
segments = DocumentSegment.query.filter(
DocumentSegment.dataset_id.in_(dataset_ids),
DocumentSegment.status == "completed",
DocumentSegment.enabled == True,
DocumentSegment.index_node_id.in_(index_node_ids),
).all()
if segments:
index_node_id_to_position = {id: position for position, id in enumerate(index_node_ids)}
sorted_segments = sorted(
segments, key=lambda segment: index_node_id_to_position.get(segment.index_node_id, float("inf"))
)
for segment in sorted_segments:
records = RetrievalService.format_retrieval_documents(dify_documents)
if records:
for record in records:
segment = record.segment
if segment.answer:
document_context_list.append(
DocumentContext(
content=f"question:{segment.get_sign_content()} answer:{segment.answer}",
score=document_score_list.get(segment.index_node_id, None),
score=record.score,
)
)
else:
document_context_list.append(
DocumentContext(
content=segment.get_sign_content(),
score=document_score_list.get(segment.index_node_id, None),
score=record.score,
)
)
if show_retrieve_source:
for segment in sorted_segments:
for record in records:
segment = record.segment
dataset = Dataset.query.filter_by(id=segment.dataset_id).first()
document = DatasetDocument.query.filter(
DatasetDocument.id == segment.document_id,
@@ -218,7 +204,7 @@ class DatasetRetrieval:
"data_source_type": document.data_source_type,
"segment_id": segment.id,
"retriever_from": invoke_from.to_source(),
"score": document_score_list.get(segment.index_node_id, 0.0),
"score": record.score or 0.0,
}
if invoke_from.to_source() == "dev":