Auto-commit: OCR workflow improvements, performance optimizations, and bug fixes

This commit is contained in:
2026-01-11 18:21:16 +08:00
parent 642dd0ea5f
commit 1ddd49f913
97 changed files with 5909 additions and 451 deletions

View File

@@ -1610,8 +1610,12 @@ async def background_delete_documents(
def create_document_routes(
rag: LightRAG, doc_manager: DocumentManager, api_key: Optional[str] = None
workspace_manager, api_key: Optional[str] = None
):
# Get default RAG instance and document manager from workspace manager
rag = workspace_manager.get_rag()
doc_manager = workspace_manager.get_document_manager()
# Create combined auth dependency for document routes
combined_auth = get_combined_auth_dependency(api_key)

View File

@@ -25,7 +25,19 @@ class RelationUpdateRequest(BaseModel):
updated_data: Dict[str, Any]
def create_graph_routes(rag, api_key: Optional[str] = None):
def create_graph_routes(rag_or_manager, api_key: Optional[str] = None):
# Accept either a LightRAG instance or a WorkspaceManager
from lightrag.api.workspace_manager import WorkspaceManager
from lightrag import LightRAG
if isinstance(rag_or_manager, WorkspaceManager):
# Get default RAG instance from workspace manager
rag = rag_or_manager.get_rag()
elif isinstance(rag_or_manager, LightRAG):
rag = rag_or_manager
else:
raise TypeError(f"Expected LightRAG or WorkspaceManager, got {type(rag_or_manager)}")
combined_auth = get_combined_auth_dependency(api_key)
@router.get("/graph/label/list", dependencies=[Depends(combined_auth)])

View File

@@ -219,9 +219,22 @@ def parse_query_mode(query: str) -> tuple[str, SearchMode, bool, Optional[str]]:
class OllamaAPI:
def __init__(self, rag: LightRAG, top_k: int = 60, api_key: Optional[str] = None):
self.rag = rag
self.ollama_server_infos = rag.ollama_server_infos
def __init__(self, rag_or_manager, top_k: int = 60, api_key: Optional[str] = None):
# Accept either a LightRAG instance or a WorkspaceManager
from lightrag.api.workspace_manager import WorkspaceManager
from lightrag import LightRAG
if isinstance(rag_or_manager, WorkspaceManager):
# Get default RAG instance from workspace manager
self.workspace_manager = rag_or_manager
self.rag = rag_or_manager.get_rag() # Get default workspace RAG
elif isinstance(rag_or_manager, LightRAG):
self.workspace_manager = None
self.rag = rag_or_manager
else:
raise TypeError(f"Expected LightRAG or WorkspaceManager, got {type(rag_or_manager)}")
self.ollama_server_infos = self.rag.ollama_server_infos
self.top_k = top_k
self.api_key = api_key
self.router = APIRouter(tags=["ollama"])

View File

@@ -149,7 +149,19 @@ class QueryDataResponse(BaseModel):
)
def create_query_routes(rag, api_key: Optional[str] = None, top_k: int = 60):
def create_query_routes(rag_or_manager, api_key: Optional[str] = None, top_k: int = 60):
# Accept either a LightRAG instance or a WorkspaceManager
from lightrag.api.workspace_manager import WorkspaceManager
from lightrag import LightRAG
if isinstance(rag_or_manager, WorkspaceManager):
# Get default RAG instance from workspace manager
rag = rag_or_manager.get_rag()
elif isinstance(rag_or_manager, LightRAG):
rag = rag_or_manager
else:
raise TypeError(f"Expected LightRAG or WorkspaceManager, got {type(rag_or_manager)}")
combined_auth = get_combined_auth_dependency(api_key)
@router.post(

View File

@@ -68,7 +68,19 @@ class SearchDataResponse(BaseModel):
)
def create_search_routes(rag, api_key: Optional[str] = None, top_k: int = 60):
def create_search_routes(rag_or_manager, api_key: Optional[str] = None, top_k: int = 60):
# Accept either a LightRAG instance or a WorkspaceManager
from lightrag.api.workspace_manager import WorkspaceManager
from lightrag import LightRAG
if isinstance(rag_or_manager, WorkspaceManager):
# Get default RAG instance from workspace manager
rag = rag_or_manager.get_rag()
elif isinstance(rag_or_manager, LightRAG):
rag = rag_or_manager
else:
raise TypeError(f"Expected LightRAG or WorkspaceManager, got {type(rag_or_manager)}")
combined_auth = get_combined_auth_dependency(api_key)
@router.post(

View File

@@ -0,0 +1,83 @@
"""
Workspace management routes for LightRAG server.
"""
import logging
from typing import List, Optional
from fastapi import APIRouter, Depends, HTTPException, Query, Request
from pydantic import BaseModel, Field
from lightrag.api.utils_api import get_combined_auth_dependency
from lightrag.api.workspace_manager import WorkspaceManager
logger = logging.getLogger(__name__)
router = APIRouter(
prefix="/workspaces",
tags=["workspaces"],
)
class WorkspaceCreateRequest(BaseModel):
name: str = Field(..., min_length=1, max_length=100, description="Workspace name (alphanumeric, underscores, hyphens)")
class WorkspaceResponse(BaseModel):
name: str
path: str
def get_workspace_manager(request: Request):
"""Dependency to get workspace manager instance."""
# Access workspace manager from app state via request
return request.app.state.workspace_manager
@router.get("/", response_model=List[WorkspaceResponse])
async def list_workspaces(
search: Optional[str] = Query(None, description="Search workspaces by keyword in name"),
workspace_manager: WorkspaceManager = Depends(get_workspace_manager),
):
"""List all existing workspaces with optional keyword search."""
try:
workspaces = workspace_manager.list_workspaces()
# Filter by search keyword if provided
if search:
search_lower = search.lower()
workspaces = [ws for ws in workspaces if search_lower in ws.lower()]
return [
WorkspaceResponse(name=ws, path=str(workspace_manager.base_working_dir / ws))
for ws in workspaces
]
except Exception as e:
logger.error(f"Error listing workspaces: {e}")
raise HTTPException(status_code=500, detail=str(e))
@router.post("/", response_model=WorkspaceResponse)
async def create_workspace(
create_request: WorkspaceCreateRequest,
workspace_manager: WorkspaceManager = Depends(get_workspace_manager),
):
"""Create a new workspace."""
try:
workspace_manager.create_workspace(create_request.name)
return WorkspaceResponse(
name=create_request.name,
path=str(workspace_manager.base_working_dir / create_request.name)
)
except ValueError as e:
raise HTTPException(status_code=400, detail=str(e))
except Exception as e:
logger.error(f"Error creating workspace: {e}")
raise HTTPException(status_code=500, detail=str(e))
@router.delete("/{workspace_name}")
async def delete_workspace(
workspace_name: str,
workspace_manager: WorkspaceManager = Depends(get_workspace_manager),
):
"""Delete a workspace (not implemented)."""
raise HTTPException(status_code=501, detail="Workspace deletion not yet implemented")