pengjunjie/ragflow
1
0
Fork 1
mirror of https://github.com/infiniflow/ragflow.git synced 2025-12-23 06:46:40 +08:00
Code Issues Packages Projects Releases Wiki Activity

Refa: make RAGFlow more asynchronous (#11601)

Browse Source 
### What problem does this PR solve?

Try to make this more asynchronous. Verified in chat and agent
scenarios, reducing blocking behavior. #11551, #11579.

However, the impact of these changes still requires further
investigation to ensure everything works as expected.

### Type of change

- [x] Refactoring
This commit is contained in:
Yongteng Lei
2025-12-01 14:24:06 +08:00
committed by GitHub
parent 6ea4248bdc
commit b6c4722687
36 changed files with 1162 additions and 359 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
api/apps/sdk/agents.py
View File

@ -25,7 +25,7 @@ from api.db.services.canvas_service import UserCanvasService
from api.db.services.user_canvas_version import UserCanvasVersionService
from common.constants import RetCode
from common.misc_utils import get_uuid
from api.utils.api_utils import get_data_error_result, get_error_data_result, get_json_result, token_required
from api.utils.api_utils import get_data_error_result, get_error_data_result, get_json_result, get_request_json, token_required
from api.utils.api_utils import get_result
from quart import request, Response
@ -53,7 +53,7 @@ def list_agents(tenant_id):
@manager.route("/agents", methods=["POST"]) # noqa: F821
@token_required
async def create_agent(tenant_id: str):
req: dict[str, Any] = cast(dict[str, Any], await request.json)
req: dict[str, Any] = cast(dict[str, Any], await get_request_json())
req["user_id"] = tenant_id
if req.get("dsl") is not None:
@ -90,7 +90,7 @@ async def create_agent(tenant_id: str):
@manager.route("/agents/<agent_id>", methods=["PUT"]) # noqa: F821
@token_required
async def update_agent(tenant_id: str, agent_id: str):
req: dict[str, Any] = {k: v for k, v in cast(dict[str, Any], (await request.json)).items() if v is not None}
req: dict[str, Any] = {k: v for k, v in cast(dict[str, Any], (await get_request_json())).items() if v is not None}
req["user_id"] = tenant_id
if req.get("dsl") is not None:
@ -136,7 +136,7 @@ def delete_agent(tenant_id: str, agent_id: str):
@manager.route('/webhook/<agent_id>', methods=['POST']) # noqa: F821
@token_required
async def webhook(tenant_id: str, agent_id: str):
req = await request.json
req = await get_request_json()
if not UserCanvasService.accessible(req["id"], tenant_id):
return get_json_result(
data=False, message='Only owner of canvas authorized for this operation.',

8
api/apps/sdk/chat.py
View File

@ -21,13 +21,13 @@ from api.db.services.tenant_llm_service import TenantLLMService
from api.db.services.user_service import TenantService
from common.misc_utils import get_uuid
from common.constants import RetCode, StatusEnum
from api.utils.api_utils import check_duplicate_ids, get_error_data_result, get_result, token_required, request_json
from api.utils.api_utils import check_duplicate_ids, get_error_data_result, get_result, token_required, get_request_json
@manager.route("/chats", methods=["POST"]) # noqa: F821
@token_required
async def create(tenant_id):
req = await request_json()
req = await get_request_json()
ids = [i for i in req.get("dataset_ids", []) if i]
for kb_id in ids:
kbs = KnowledgebaseService.accessible(kb_id=kb_id, user_id=tenant_id)
@ -146,7 +146,7 @@ async def create(tenant_id):
async def update(tenant_id, chat_id):
if not DialogService.query(tenant_id=tenant_id, id=chat_id, status=StatusEnum.VALID.value):
return get_error_data_result(message="You do not own the chat")
req = await request_json()
req = await get_request_json()
ids = req.get("dataset_ids", [])
if "show_quotation" in req:
req["do_refer"] = req.pop("show_quotation")
@ -229,7 +229,7 @@ async def update(tenant_id, chat_id):
async def delete_chats(tenant_id):
errors = []
success_count = 0
req = await request_json()
req = await get_request_json()
if not req:
ids = None
else:

6
api/apps/sdk/dify_retrieval.py
View File

@ -15,12 +15,12 @@
#
import logging
from quart import request, jsonify
from quart import jsonify
from api.db.services.document_service import DocumentService
from api.db.services.knowledgebase_service import KnowledgebaseService
from api.db.services.llm_service import LLMBundle
from api.utils.api_utils import validate_request, build_error_result, apikey_required
from api.utils.api_utils import apikey_required, build_error_result, get_request_json, validate_request
from rag.app.tag import label_question
from api.db.services.dialog_service import meta_filter, convert_conditions
from common.constants import RetCode, LLMType
@ -113,7 +113,7 @@ async def retrieval(tenant_id):
404:
description: Knowledge base or document not found
"""
req = await request.json
req = await get_request_json()
question = req["query"]
kb_id = req["knowledge_id"]
use_kg = req.get("use_kg", False)

32
api/apps/sdk/doc.py
View File

@ -36,7 +36,7 @@ from api.db.services.tenant_llm_service import TenantLLMService
from api.db.services.task_service import TaskService, queue_tasks
from api.db.services.dialog_service import meta_filter, convert_conditions
from api.utils.api_utils import check_duplicate_ids, construct_json_result, get_error_data_result, get_parser_config, get_result, server_error_response, token_required, \
request_json
get_request_json
from rag.app.qa import beAdoc, rmPrefix
from rag.app.tag import label_question
from rag.nlp import rag_tokenizer, search
@ -231,7 +231,7 @@ async def update_doc(tenant_id, dataset_id, document_id):
schema:
type: object
"""
req = await request_json()
req = await get_request_json()
if not KnowledgebaseService.query(id=dataset_id, tenant_id=tenant_id):
return get_error_data_result(message="You don't own the dataset.")
e, kb = KnowledgebaseService.get_by_id(dataset_id)
@ -536,7 +536,7 @@ def list_docs(dataset_id, tenant_id):
return get_error_data_result(message=f"You don't own the dataset {dataset_id}. ")
q = request.args
document_id = q.get("id")
document_id = q.get("id")
name = q.get("name")
if document_id and not DocumentService.query(id=document_id, kb_id=dataset_id):
@ -545,16 +545,16 @@ def list_docs(dataset_id, tenant_id):
return get_error_data_result(message=f"You don't own the document {name}.")
page = int(q.get("page", 1))
page_size = int(q.get("page_size", 30))
page_size = int(q.get("page_size", 30))
orderby = q.get("orderby", "create_time")
desc = str(q.get("desc", "true")).strip().lower() != "false"
keywords = q.get("keywords", "")
# filters - align with OpenAPI parameter names
suffix = q.getlist("suffix")
run_status = q.getlist("run")
create_time_from = int(q.get("create_time_from", 0))
create_time_to = int(q.get("create_time_to", 0))
suffix = q.getlist("suffix")
run_status = q.getlist("run")
create_time_from = int(q.get("create_time_from", 0))
create_time_to = int(q.get("create_time_to", 0))
# map run status (accept text or numeric) - align with API parameter
run_status_text_to_numeric = {"UNSTART": "0", "RUNNING": "1", "CANCEL": "2", "DONE": "3", "FAIL": "4"}
@ -575,7 +575,7 @@ def list_docs(dataset_id, tenant_id):
# rename keys + map run status back to text for output
key_mapping = {
"chunk_num": "chunk_count",
"kb_id": "dataset_id",
"kb_id": "dataset_id",
"token_num": "token_count",
"parser_id": "chunk_method",
}
@ -631,7 +631,7 @@ async def delete(tenant_id, dataset_id):
"""
if not KnowledgebaseService.accessible(kb_id=dataset_id, user_id=tenant_id):
return get_error_data_result(message=f"You don't own the dataset {dataset_id}. ")
req = await request_json()
req = await get_request_json()
if not req:
doc_ids = None
else:
@ -741,7 +741,7 @@ async def parse(tenant_id, dataset_id):
"""
if not KnowledgebaseService.accessible(kb_id=dataset_id, user_id=tenant_id):
return get_error_data_result(message=f"You don't own the dataset {dataset_id}.")
req = await request_json()
req = await get_request_json()
if not req.get("document_ids"):
return get_error_data_result("`document_ids` is required")
doc_list = req.get("document_ids")
@ -824,7 +824,7 @@ async def stop_parsing(tenant_id, dataset_id):
"""
if not KnowledgebaseService.accessible(kb_id=dataset_id, user_id=tenant_id):
return get_error_data_result(message=f"You don't own the dataset {dataset_id}.")
req = await request_json()
req = await get_request_json()
if not req.get("document_ids"):
return get_error_data_result("`document_ids` is required")
@ -1096,7 +1096,7 @@ async def add_chunk(tenant_id, dataset_id, document_id):
if not doc:
return get_error_data_result(message=f"You don't own the document {document_id}.")
doc = doc[0]
req = await request_json()
req = await get_request_json()
if not str(req.get("content", "")).strip():
return get_error_data_result(message="`content` is required")
if "important_keywords" in req:
@ -1202,7 +1202,7 @@ async def rm_chunk(tenant_id, dataset_id, document_id):
docs = DocumentService.get_by_ids([document_id])
if not docs:
raise LookupError(f"Can't find the document with ID {document_id}!")
req = await request_json()
req = await get_request_json()
condition = {"doc_id": document_id}
if "chunk_ids" in req:
unique_chunk_ids, duplicate_messages = check_duplicate_ids(req["chunk_ids"], "chunk")
@ -1288,7 +1288,7 @@ async def update_chunk(tenant_id, dataset_id, document_id, chunk_id):
if not doc:
return get_error_data_result(message=f"You don't own the document {document_id}.")
doc = doc[0]
req = await request_json()
req = await get_request_json()
if "content" in req and req["content"] is not None:
content = req["content"]
else:
@ -1411,7 +1411,7 @@ async def retrieval_test(tenant_id):
format: float
description: Similarity score.
"""
req = await request_json()
req = await get_request_json()
if not req.get("dataset_ids"):
return get_error_data_result("`dataset_ids` is required.")
kb_ids = req["dataset_ids"]

25
api/apps/sdk/files.py
View File

@ -23,12 +23,11 @@ from pathlib import Path
from api.db.services.document_service import DocumentService
from api.db.services.file2document_service import File2DocumentService
from api.db.services.knowledgebase_service import KnowledgebaseService
from api.utils.api_utils import server_error_response, token_required
from api.utils.api_utils import get_json_result, get_request_json, server_error_response, token_required
from common.misc_utils import get_uuid
from api.db import FileType
from api.db.services import duplicate_name
from api.db.services.file_service import FileService
from api.utils.api_utils import get_json_result
from api.utils.file_utils import filename_type
from common import settings
from common.constants import RetCode
@ -193,9 +192,9 @@ async def create(tenant_id):
type:
type: string
"""
req = await request.json
pf_id = await request.json.get("parent_id")
input_file_type = await request.json.get("type")
req = await get_request_json()
pf_id = req.get("parent_id")
input_file_type = req.get("type")
if not pf_id:
root_folder = FileService.get_root_folder(tenant_id)
pf_id = root_folder["id"]
@ -229,7 +228,7 @@ async def create(tenant_id):
@manager.route('/file/list', methods=['GET']) # noqa: F821
@token_required
def list_files(tenant_id):
async def list_files(tenant_id):
"""
List files under a specific folder.
---
@ -321,7 +320,7 @@ def list_files(tenant_id):
@manager.route('/file/root_folder', methods=['GET']) # noqa: F821
@token_required
def get_root_folder(tenant_id):
async def get_root_folder(tenant_id):
"""
Get user's root folder.
---
@ -357,7 +356,7 @@ def get_root_folder(tenant_id):
@manager.route('/file/parent_folder', methods=['GET']) # noqa: F821
@token_required
def get_parent_folder():
async def get_parent_folder():
"""
Get parent folder info of a file.
---
@ -402,7 +401,7 @@ def get_parent_folder():
@manager.route('/file/all_parent_folder', methods=['GET']) # noqa: F821
@token_required
def get_all_parent_folders(tenant_id):
async def get_all_parent_folders(tenant_id):
"""
Get all parent folders of a file.
---
@ -481,7 +480,7 @@ async def rm(tenant_id):
type: boolean
example: true
"""
req = await request.json
req = await get_request_json()
file_ids = req["file_ids"]
try:
for file_id in file_ids:
@ -556,7 +555,7 @@ async def rename(tenant_id):
type: boolean
example: true
"""
req = await request.json
req = await get_request_json()
try:
e, file = FileService.get_by_id(req["file_id"])
if not e:
@ -667,7 +666,7 @@ async def move(tenant_id):
type: boolean
example: true
"""
req = await request.json
req = await get_request_json()
try:
file_ids = req["src_file_ids"]
parent_id = req["dest_file_id"]
@ -694,7 +693,7 @@ async def move(tenant_id):
@manager.route('/file/convert', methods=['POST']) # noqa: F821
@token_required
async def convert(tenant_id):
req = await request.json
req = await get_request_json()
kb_ids = req["kb_ids"]
file_ids = req["file_ids"]
file2documents = []

46
api/apps/sdk/session.py
View File

@ -35,7 +35,7 @@ from api.db.services.search_service import SearchService
from api.db.services.user_service import UserTenantService
from common.misc_utils import get_uuid
from api.utils.api_utils import check_duplicate_ids, get_data_openai, get_error_data_result, get_json_result, \
get_result, server_error_response, token_required, validate_request
get_result, get_request_json, server_error_response, token_required, validate_request
from rag.app.tag import label_question
from rag.prompts.template import load_prompt
from rag.prompts.generator import cross_languages, gen_meta_filter, keyword_extraction, chunks_format
@ -45,7 +45,7 @@ from common import settings
@manager.route("/chats/<chat_id>/sessions", methods=["POST"]) # noqa: F821
@token_required
async def create(tenant_id, chat_id):
req = await request.json
req = await get_request_json()
req["dialog_id"] = chat_id
dia = DialogService.query(tenant_id=tenant_id, id=req["dialog_id"], status=StatusEnum.VALID.value)
if not dia:
@ -73,7 +73,7 @@ async def create(tenant_id, chat_id):
@manager.route("/agents/<agent_id>/sessions", methods=["POST"]) # noqa: F821
@token_required
def create_agent_session(tenant_id, agent_id):
async def create_agent_session(tenant_id, agent_id):
user_id = request.args.get("user_id", tenant_id)
e, cvs = UserCanvasService.get_by_id(agent_id)
if not e:
@ -98,7 +98,7 @@ def create_agent_session(tenant_id, agent_id):
@manager.route("/chats/<chat_id>/sessions/<session_id>", methods=["PUT"]) # noqa: F821
@token_required
async def update(tenant_id, chat_id, session_id):
req = await request.json
req = await get_request_json()
req["dialog_id"] = chat_id
conv_id = session_id
conv = ConversationService.query(id=conv_id, dialog_id=chat_id)
@ -120,7 +120,7 @@ async def update(tenant_id, chat_id, session_id):
@manager.route("/chats/<chat_id>/completions", methods=["POST"]) # noqa: F821
@token_required
async def chat_completion(tenant_id, chat_id):
req = await request.json
req = await get_request_json()
if not req:
req = {"question": ""}
if not req.get("session_id"):
@ -206,7 +206,7 @@ async def chat_completion_openai_like(tenant_id, chat_id):
if reference:
print(completion.choices[0].message.reference)
"""
req = await request.get_json()
req = await get_request_json()
need_reference = bool(req.get("reference", False))
@ -384,7 +384,7 @@ async def chat_completion_openai_like(tenant_id, chat_id):
@validate_request("model", "messages") # noqa: F821
@token_required
async def agents_completion_openai_compatibility(tenant_id, agent_id):
req = await request.json
req = await get_request_json()
tiktokenenc = tiktoken.get_encoding("cl100k_base")
messages = req.get("messages", [])
if not messages:
@ -442,7 +442,7 @@ async def agents_completion_openai_compatibility(tenant_id, agent_id):
@manager.route("/agents/<agent_id>/completions", methods=["POST"]) # noqa: F821
@token_required
async def agent_completions(tenant_id, agent_id):
req = await request.json
req = await get_request_json()
if req.get("stream", True):
@ -491,7 +491,7 @@ async def agent_completions(tenant_id, agent_id):
@manager.route("/chats/<chat_id>/sessions", methods=["GET"]) # noqa: F821
@token_required
def list_session(tenant_id, chat_id):
async def list_session(tenant_id, chat_id):
if not DialogService.query(tenant_id=tenant_id, id=chat_id, status=StatusEnum.VALID.value):
return get_error_data_result(message=f"You don't own the assistant {chat_id}.")
id = request.args.get("id")
@ -545,7 +545,7 @@ def list_session(tenant_id, chat_id):
@manager.route("/agents/<agent_id>/sessions", methods=["GET"]) # noqa: F821
@token_required
def list_agent_session(tenant_id, agent_id):
async def list_agent_session(tenant_id, agent_id):
if not UserCanvasService.query(user_id=tenant_id, id=agent_id):
return get_error_data_result(message=f"You don't own the agent {agent_id}.")
id = request.args.get("id")
@ -614,7 +614,7 @@ async def delete(tenant_id, chat_id):
errors = []
success_count = 0
req = await request.json
req = await get_request_json()
convs = ConversationService.query(dialog_id=chat_id)
if not req:
ids = None
@ -662,7 +662,7 @@ async def delete(tenant_id, chat_id):
async def delete_agent_session(tenant_id, agent_id):
errors = []
success_count = 0
req = await request.json
req = await get_request_json()
cvs = UserCanvasService.query(user_id=tenant_id, id=agent_id)
if not cvs:
return get_error_data_result(f"You don't own the agent {agent_id}")
@ -715,7 +715,7 @@ async def delete_agent_session(tenant_id, agent_id):
@manager.route("/sessions/ask", methods=["POST"]) # noqa: F821
@token_required
async def ask_about(tenant_id):
req = await request.json
req = await get_request_json()
if not req.get("question"):
return get_error_data_result("`question` is required.")
if not req.get("dataset_ids"):
@ -754,7 +754,7 @@ async def ask_about(tenant_id):
@manager.route("/sessions/related_questions", methods=["POST"]) # noqa: F821
@token_required
async def related_questions(tenant_id):
req = await request.json
req = await get_request_json()
if not req.get("question"):
return get_error_data_result("`question` is required.")
question = req["question"]
@ -805,7 +805,7 @@ Related search terms:
@manager.route("/chatbots/<dialog_id>/completions", methods=["POST"]) # noqa: F821
async def chatbot_completions(dialog_id):
req = await request.json
req = await get_request_json()
token = request.headers.get("Authorization").split()
if len(token) != 2:
@ -831,7 +831,7 @@ async def chatbot_completions(dialog_id):
@manager.route("/chatbots/<dialog_id>/info", methods=["GET"]) # noqa: F821
def chatbots_inputs(dialog_id):
async def chatbots_inputs(dialog_id):
token = request.headers.get("Authorization").split()
if len(token) != 2:
return get_error_data_result(message='Authorization is not valid!"')
@ -855,7 +855,7 @@ def chatbots_inputs(dialog_id):
@manager.route("/agentbots/<agent_id>/completions", methods=["POST"]) # noqa: F821
async def agent_bot_completions(agent_id):
req = await request.json
req = await get_request_json()
token = request.headers.get("Authorization").split()
if len(token) != 2:
@ -878,7 +878,7 @@ async def agent_bot_completions(agent_id):
@manager.route("/agentbots/<agent_id>/inputs", methods=["GET"]) # noqa: F821
def begin_inputs(agent_id):
async def begin_inputs(agent_id):
token = request.headers.get("Authorization").split()
if len(token) != 2:
return get_error_data_result(message='Authorization is not valid!"')
@ -908,7 +908,7 @@ async def ask_about_embedded():
if not objs:
return get_error_data_result(message='Authentication error: API key is invalid!"')
req = await request.json
req = await get_request_json()
uid = objs[0].tenant_id
search_id = req.get("search_id", "")
@ -947,7 +947,7 @@ async def retrieval_test_embedded():
if not objs:
return get_error_data_result(message='Authentication error: API key is invalid!"')
req = await request.json
req = await get_request_json()
page = int(req.get("page", 1))
size = int(req.get("size", 30))
question = req["question"]
@ -1046,7 +1046,7 @@ async def related_questions_embedded():
if not objs:
return get_error_data_result(message='Authentication error: API key is invalid!"')
req = await request.json
req = await get_request_json()
tenant_id = objs[0].tenant_id
if not tenant_id:
return get_error_data_result(message="permission denined.")
@ -1081,7 +1081,7 @@ Related search terms:
@manager.route("/searchbots/detail", methods=["GET"]) # noqa: F821
def detail_share_embedded():
async def detail_share_embedded():
token = request.headers.get("Authorization").split()
if len(token) != 2:
return get_error_data_result(message='Authorization is not valid!"')
@ -1123,7 +1123,7 @@ async def mindmap():
return get_error_data_result(message='Authentication error: API key is invalid!"')
tenant_id = objs[0].tenant_id
req = await request.json
req = await get_request_json()
search_id = req.get("search_id", "")
search_app = SearchService.get_detail(search_id) if search_id else {}