apunkt/SurfSense
1
0
Fork 0
mirror of https://github.com/MODSetter/SurfSense.git synced 2026-06-22 21:28:12 +02:00
Code Issues Projects Releases Packages Wiki Activity

feat(roadmap-1.2): removed SearchMode enum and combined hybrid search

Browse source
This commit is contained in:
DESKTOP-RTLN3BA\$punk 2025-12-12 04:17:30 -08:00
parent dc92f313e7
commit 08fb488995
10 changed files with 331 additions and 548 deletions
Download patch file Download diff file Expand all files Collapse all files

9
surfsense_backend/app/agents/researcher/configuration.py
View file

@ -3,18 +3,10 @@
from __future__ import annotations
from dataclasses import dataclass, fields
from enum import Enum
from langchain_core.runnables import RunnableConfig
class SearchMode(Enum):
"""Enum defining the type of search mode."""
CHUNKS = "CHUNKS"
DOCUMENTS = "DOCUMENTS"
@dataclass(kw_only=True)
class Configuration:
"""The configuration for the agent."""
@ -24,7 +16,6 @@ class Configuration:
connectors_to_search: list[str]
user_id: str
search_space_id: int
search_mode: SearchMode
document_ids_to_add_in_context: list[int]
language: str | None = None
top_k: int = 10

26
surfsense_backend/app/agents/researcher/nodes.py
View file

@ -16,7 +16,7 @@ from app.db import Document
from app.services.connector_service import ConnectorService
from app.services.query_service import QueryService
from .configuration import Configuration, SearchMode
from .configuration import Configuration
from .prompts import get_further_questions_system_prompt
from .qna_agent.graph import graph as qna_agent_graph
from .state import State
@ -102,7 +102,7 @@ async def fetch_documents_by_ids(
Fetch documents by their IDs within a search space.
This function ensures that only documents belonging to the search space are fetched.
Similar to SearchMode.DOCUMENTS, it fetches full documents and concatenates their chunks.
It fetches full documents and returns their chunks individually.
Also creates source objects for UI display, grouped by document type.
Args:
@ -526,7 +526,6 @@ async def fetch_relevant_documents(
state: State = None,
top_k: int = 10,
connector_service: ConnectorService = None,
search_mode: SearchMode = SearchMode.CHUNKS,
user_selected_sources: list[dict[str, Any]] | None = None,
start_date: datetime | None = None,
end_date: datetime | None = None,
@ -539,6 +538,8 @@ async def fetch_relevant_documents(
displaying connector names (like "Web Search" instead of "TAVILY_API") and adding
relevant emojis to indicate the type of source being searched.
Uses combined chunk-level and document-level hybrid search with RRF fusion.
Args:
research_questions: List of research questions to find documents for
search_space_id: The search space ID
@ -548,7 +549,6 @@ async def fetch_relevant_documents(
state: The current state containing the streaming service
top_k: Number of top results to retrieve per connector per question
connector_service: An initialized connector service to use for searching
search_mode: Search mode (CHUNKS or DOCUMENTS)
user_selected_sources: Optional list of user-selected source objects
start_date: Optional start date for filtering documents by updated_at
end_date: Optional end date for filtering documents by updated_at
@ -629,7 +629,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -657,7 +656,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -685,7 +683,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -710,7 +707,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -735,7 +731,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -763,7 +758,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -791,7 +785,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -819,7 +812,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -947,7 +939,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -970,7 +961,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -997,7 +987,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -1024,7 +1013,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -1051,7 +1039,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -1078,7 +1065,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -1105,7 +1091,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -1133,7 +1118,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -1161,7 +1145,6 @@ async def fetch_relevant_documents(
user_query=reformulated_query,
search_space_id=search_space_id,
top_k=top_k,
search_mode=search_mode,
start_date=start_date,
end_date=end_date,
)
@ -1444,7 +1427,6 @@ async def handle_qna_workflow(
state=state,
top_k=top_k,
connector_service=connector_service,
search_mode=configuration.search_mode,
user_selected_sources=user_selected_sources,
start_date=start_date,
end_date=end_date,

9
surfsense_backend/app/agents/researcher/qna_agent/nodes.py
View file

@ -89,10 +89,13 @@ async def rerank_documents(state: State, config: RunnableConfig) -> dict[str, An
# Get reranker service from app config
reranker_service = RerankerService.get_reranker_instance()
# If reranking is not enabled, return original documents without processing
# If reranking is not enabled, sort by existing score and return
if not reranker_service:
print("Reranking is disabled. Using original document order.")
return {"reranked_documents": documents}
print("Reranking is disabled. Sorting documents by existing score.")
sorted_documents = sorted(
documents, key=lambda x: x.get("score", 0), reverse=True
)
return {"reranked_documents": sorted_documents}
# Perform reranking
try:

3
surfsense_backend/app/routes/chats_routes.py
View file

@ -29,7 +29,6 @@ from app.utils.validators import (
validate_document_ids,
validate_messages,
validate_research_mode,
validate_search_mode,
validate_search_space_id,
validate_top_k,
)
@ -61,7 +60,6 @@ async def handle_chat_data(
document_ids_to_add_in_context = validate_document_ids(
request_data.get("document_ids_to_add_in_context")
)
search_mode_str = validate_search_mode(request_data.get("search_mode"))
top_k = validate_top_k(request_data.get("top_k"))
# print("RESQUEST DATA:", request_data)
# print("SELECTED CONNECTORS:", selected_connectors)
@ -144,7 +142,6 @@ async def handle_chat_data(
research_mode,
selected_connectors,
langchain_chat_history,
search_mode_str,
document_ids_to_add_in_context,
language,
top_k,

734
surfsense_backend/app/services/connector_service.py
View file

File diff suppressed because it is too large Load diff

8
surfsense_backend/app/tasks/stream_connector_search_results.py
View file

@ -4,7 +4,6 @@ from uuid import UUID
from sqlalchemy.ext.asyncio import AsyncSession
from app.agents.researcher.configuration import SearchMode
from app.agents.researcher.graph import graph as researcher_graph
from app.agents.researcher.state import State
from app.services.streaming_service import StreamingService
@ -18,7 +17,6 @@ async def stream_connector_search_results(
research_mode: str,
selected_connectors: list[str],
langchain_chat_history: list[Any],
search_mode_str: str,
document_ids_to_add_in_context: list[int],
language: str | None = None,
top_k: int = 10,
@ -42,11 +40,6 @@ async def stream_connector_search_results(
# Convert UUID to string if needed
user_id_str = str(user_id) if isinstance(user_id, UUID) else user_id
if search_mode_str == "CHUNKS":
search_mode = SearchMode.CHUNKS
elif search_mode_str == "DOCUMENTS":
search_mode = SearchMode.DOCUMENTS
# Sample configuration
config = {
"configurable": {
@ -54,7 +47,6 @@ async def stream_connector_search_results(
"connectors_to_search": selected_connectors,
"user_id": user_id_str,
"search_space_id": search_space_id,
"search_mode": search_mode,
"document_ids_to_add_in_context": document_ids_to_add_in_context,
"language": language, # Add language to the configuration
"top_k": top_k, # Add top_k to the configuration