apunkt/SurfSense
1
0
Fork 0
mirror of https://github.com/MODSetter/SurfSense.git synced 2026-05-12 09:12:40 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 2

Merge pull request #622 from CREDO23/documents-mentions

Browse source 
[Fix] Documents mentions | Use the same structure of document returned from retriever
This commit is contained in:
Rohan Verma 2025-12-24 13:46:31 -08:00 • committed by GitHub
commit fb719faa0d
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 33 additions and 13 deletions
Download patch file Download diff file Expand all files Collapse all files

46
surfsense_backend/app/tasks/chat/stream_new_chat.py
View file

@ -52,21 +52,41 @@ def format_mentioned_documents_as_context(documents: list[Document]) -> str:
"""Format mentioned documents as context for the agent."""
if not documents:
return ""
import json
context_parts = ["<mentioned_documents>"]
context_parts.append(
"The user has explicitly mentioned the following documents from their knowledge base. "
"These documents are directly relevant to the query and should be prioritized as primary sources."
)
for i, doc in enumerate(documents, 1):
context_parts.append(
f"<document index='{i}' id='{doc.id}' title='{doc.title}' type='{doc.document_type.value}'>"
parts = []
for doc in documents:
metadata = doc.document_metadata or {}
chunks = (
[
{"chunk_id": c.id, "content": c.content}
for c in getattr(doc, "chunks", [])
]
if hasattr(doc, "chunks") and doc.chunks
else [{"chunk_id": doc.id, "content": doc.content}]
)
context_parts.append(f"<![CDATA[{doc.content}]]>")
context_parts.append("</document>")
context_parts.append("</mentioned_documents>")
return "\n".join(context_parts)
metadata_json = json.dumps(metadata, ensure_ascii=False)
parts.append("<document>")
parts.append("<document_metadata>")
parts.append(f" <document_id>{doc.id}</document_id>")
parts.append(f" <document_type>{doc.document_type.value}</document_type>")
parts.append(f" <title><![CDATA[{doc.title}]]></title>")
parts.append(" <url><![CDATA[]]></url>")
parts.append(f" <metadata_json><![CDATA[{metadata_json}]]></metadata_json>")
parts.append("</document_metadata>")
parts.append("")
parts.append("<document_content>")
for ch in chunks:
ch_content = ch["content"]
ch_id = ch["chunk_id"]
if ch_id is None:
parts.append(f" <chunk><![CDATA[{ch_content}]]></chunk>")
else:
parts.append(f" <chunk id='{ch_id}'><![CDATA[{ch_content}]]></chunk>")
parts.append("</document_content>")
parts.append("</document>")
parts.append("")
return "\n".join(parts).strip()
async def stream_new_chat(