pengjunjie/ragflow
1
0
Fork 1
mirror of https://github.com/infiniflow/ragflow.git synced 2025-12-08 20:42:30 +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

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"]