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

Fix: debug hierachical merging... (#10337)

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


### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)
This commit is contained in:
Kevin Hu
2025-09-29 09:29:33 +08:00
committed by GitHub
parent 664bc0b961
commit 71f69cdb75
13 changed files with 113 additions and 46 deletions
Download Patch File Download Diff File Expand all files Collapse all files

31
api/apps/canvas_app.py
View File

@ -19,17 +19,19 @@ import re
import sys
from functools import partial
import flask
import trio
from flask import request, Response
from flask_login import login_required, current_user
from agent.component import LLM
from api import settings
from api.db import CanvasCategory, FileType
from api.db.services.canvas_service import CanvasTemplateService, UserCanvasService, API4ConversationService
from api.db.services.document_service import DocumentService
from api.db.services.file_service import FileService
from api.db.services.pipeline_operation_log_service import PipelineOperationLogService
from api.db.services.task_service import queue_dataflow, CANVAS_DEBUG_DOC_ID
from api.db.services.task_service import queue_dataflow, CANVAS_DEBUG_DOC_ID, TaskService
from api.db.services.user_service import TenantService
from api.db.services.user_canvas_version import UserCanvasVersionService
from api.settings import RetCode
@ -37,11 +39,12 @@ from api.utils import get_uuid
from api.utils.api_utils import get_json_result, server_error_response, validate_request, get_data_error_result
from agent.canvas import Canvas
from peewee import MySQLDatabase, PostgresqlDatabase
from api.db.db_models import APIToken
from api.db.db_models import APIToken, Task
import time
from api.utils.file_utils import filename_type, read_potential_broken_pdf
from rag.flow.pipeline import Pipeline
from rag.nlp import search
from rag.utils.redis_conn import REDIS_CONN
@ -189,6 +192,15 @@ def rerun():
if 0 < doc["progress"] < 1:
return get_data_error_result(message=f"`{doc['name']}` is processing...")
if settings.docStoreConn.indexExist(search.index_name(current_user.id), doc["kb_id"]):
settings.docStoreConn.delete({"doc_id": doc["id"]}, search.index_name(current_user.id), doc["kb_id"])
doc["progress_msg"] = ""
doc["chunk_num"] = 0
doc["token_num"] = 0
DocumentService.clear_chunk_num_when_rerun(doc["id"])
DocumentService.update_by_id(id, doc)
TaskService.filter_delete([Task.doc_id == id])
dsl = req["dsl"]
dsl["path"] = [req["component_id"]]
PipelineOperationLogService.update_by_id(req["id"], {"dsl": dsl})
@ -420,8 +432,8 @@ def getversion( version_id):
@login_required
def list_canvas():
keywords = request.args.get("keywords", "")
page_number = int(request.args.get("page", 1))
items_per_page = int(request.args.get("page_size", 150))
page_number = int(request.args.get("page", 0))
items_per_page = int(request.args.get("page_size", 0))
orderby = request.args.get("orderby", "create_time")
canvas_category = request.args.get("canvas_category")
if request.args.get("desc", "true").lower() == "false":
@ -429,9 +441,12 @@ def list_canvas():
else:
desc = True
owner_ids = request.args.get("owner_ids", [])
if owner_ids and isinstance(owner_ids, str):
owner_ids = [owner_ids]
if not owner_ids:
tenants = TenantService.get_joined_tenants_by_user_id(current_user.id)
tenants = [m["tenant_id"] for m in tenants]
tenants.append(current_user.id)
canvas, total = UserCanvasService.get_by_tenant_ids(
tenants, current_user.id, page_number,
items_per_page, orderby, desc, keywords, canvas_category)
@ -525,3 +540,11 @@ def prompts():
#"context_ranking": RANK_MEMORY,
"citation_guidelines": CITATION_PROMPT_TEMPLATE
})
@manager.route('/download', methods=['GET']) # noqa: F821
def download():
id = request.args.get("id")
created_by = request.args.get("created_by")
blob = FileService.get_blob(created_by, id)
return flask.make_response(blob)

35
api/apps/kb_app.py
View File

@ -68,7 +68,34 @@ def create():
e, t = TenantService.get_by_id(current_user.id)
if not e:
return get_data_error_result(message="Tenant not found.")
#req["embd_id"] = t.embd_id
req["parser_config"] = {
"layout_recognize": "DeepDOC",
"chunk_token_num": 512,
"delimiter": "\n",
"auto_keywords": 0,
"auto_questions": 0,
"html4excel": False,
"topn_tags": 3,
"raptor": {
"use_raptor": True,
"prompt": "Please summarize the following paragraphs. Be careful with the numbers, do not make things up. Paragraphs as following:\n {cluster_content}\nThe above is the content you need to summarize.",
"max_token": 256,
"threshold": 0.1,
"max_cluster": 64,
"random_seed": 0
},
"graphrag": {
"use_graphrag": True,
"entity_types": [
"organization",
"person",
"geo",
"event",
"category"
],
"method": "light"
}
}
if not KnowledgebaseService.save(**req):
return get_data_error_result()
return get_json_result(data={"kb_id": req["id"]})
@ -729,19 +756,21 @@ def delete_kb_task():
if not pipeline_task_type or pipeline_task_type not in [PipelineTaskType.GRAPH_RAG, PipelineTaskType.RAPTOR, PipelineTaskType.MINDMAP]:
return get_error_data_result(message="Invalid task type")
kb_task_id = ""
match pipeline_task_type:
case PipelineTaskType.GRAPH_RAG:
settings.docStoreConn.delete({"knowledge_graph_kwd": ["graph", "subgraph", "entity", "relation"]}, search.index_name(kb.tenant_id), kb_id)
kb_task_id = "graphrag_task_id"
kb_task_finish_at = "graphrag_task_finish_at"
case PipelineTaskType.RAPTOR:
kb_task_id = "raptor_task_id"
kb_task_finish_at = "raptor_task_finish_at"
case PipelineTaskType.MINDMAP:
kb_task_id = "mindmap_task_id"
kb_task_finish_at = "mindmap_task_finish_at"
case _:
return get_error_data_result(message="Internal Error: Invalid task type")
ok = KnowledgebaseService.update_by_id(kb_id, {kb_task_id: ""})
ok = KnowledgebaseService.update_by_id(kb_id, {kb_task_id: "", kb_task_finish_at: None})
if not ok:
return server_error_response(f"Internal error: cannot delete task {pipeline_task_type}")

20
api/db/services/canvas_service.py
View File

@ -18,7 +18,7 @@ import logging
import time
from uuid import uuid4
from agent.canvas import Canvas
from api.db import CanvasCategory, TenantPermission
from api.db import CanvasCategory
from api.db.db_models import DB, CanvasTemplate, User, UserCanvas, API4Conversation
from api.db.services.api_service import API4ConversationService
from api.db.services.common_service import CommonService
@ -104,6 +104,7 @@ class UserCanvasService(CommonService):
cls.model.dsl,
cls.model.description,
cls.model.permission,
cls.model.user_id.alias("tenant_id"),
User.nickname,
User.avatar.alias('tenant_avatar'),
cls.model.update_time,
@ -111,16 +112,15 @@ class UserCanvasService(CommonService):
]
if keywords:
agents = cls.model.select(*fields).join(User, on=(cls.model.user_id == User.id)).where(
((cls.model.user_id.in_(joined_tenant_ids) & (cls.model.permission ==
TenantPermission.TEAM.value)) | (
cls.model.user_id == user_id)),
(fn.LOWER(cls.model.title).contains(keywords.lower()))
cls.model.user_id.in_(joined_tenant_ids),
fn.LOWER(cls.model.title).contains(keywords.lower())
#(((cls.model.user_id.in_(joined_tenant_ids)) & (cls.model.permission == TenantPermission.TEAM.value)) | (cls.model.user_id == user_id)),
#(fn.LOWER(cls.model.title).contains(keywords.lower()))
)
else:
agents = cls.model.select(*fields).join(User, on=(cls.model.user_id == User.id)).where(
((cls.model.user_id.in_(joined_tenant_ids) & (cls.model.permission ==
TenantPermission.TEAM.value)) | (
cls.model.user_id == user_id))
cls.model.user_id.in_(joined_tenant_ids)
#(((cls.model.user_id.in_(joined_tenant_ids)) & (cls.model.permission == TenantPermission.TEAM.value)) | (cls.model.user_id == user_id))
)
if canvas_category:
agents = agents.where(cls.model.canvas_category == canvas_category)
@ -128,8 +128,10 @@ class UserCanvasService(CommonService):
agents = agents.order_by(cls.model.getter_by(orderby).desc())
else:
agents = agents.order_by(cls.model.getter_by(orderby).asc())
count = agents.count()
agents = agents.paginate(page_number, items_per_page)
if page_number and items_per_page:
agents = agents.paginate(page_number, items_per_page)
return list(agents.dicts()), count
@classmethod

9
api/db/services/document_service.py
View File

@ -29,7 +29,8 @@ from peewee import fn, Case, JOIN
from api import settings
from api.constants import IMG_BASE64_PREFIX, FILE_NAME_LEN_LIMIT
from api.db import FileType, LLMType, ParserType, StatusEnum, TaskStatus, UserTenantRole, CanvasCategory
from api.db.db_models import DB, Document, Knowledgebase, Task, Tenant, UserTenant, File2Document, File, UserCanvas
from api.db.db_models import DB, Document, Knowledgebase, Task, Tenant, UserTenant, File2Document, File, UserCanvas, \
User
from api.db.db_utils import bulk_insert_into_db
from api.db.services.common_service import CommonService
from api.db.services.knowledgebase_service import KnowledgebaseService
@ -121,19 +122,21 @@ class DocumentService(CommonService):
orderby, desc, keywords, run_status, types, suffix):
fields = cls.get_cls_model_fields()
if keywords:
docs = cls.model.select(*[*fields, UserCanvas.title.alias("pipeline_name")])\
docs = cls.model.select(*[*fields, UserCanvas.title.alias("pipeline_name"), User.nickname])\
.join(File2Document, on=(File2Document.document_id == cls.model.id))\
.join(File, on=(File.id == File2Document.file_id))\
.join(UserCanvas, on=(cls.model.pipeline_id == UserCanvas.id), join_type=JOIN.LEFT_OUTER)\
.join(User, on=(cls.model.created_by == User.id), join_type=JOIN.LEFT_OUTER)\
.where(
(cls.model.kb_id == kb_id),
(fn.LOWER(cls.model.name).contains(keywords.lower()))
)
else:
docs = cls.model.select(*[*fields, UserCanvas.title.alias("pipeline_name")])\
docs = cls.model.select(*[*fields, UserCanvas.title.alias("pipeline_name"), User.nickname])\
.join(File2Document, on=(File2Document.document_id == cls.model.id))\
.join(UserCanvas, on=(cls.model.pipeline_id == UserCanvas.id), join_type=JOIN.LEFT_OUTER)\
.join(File, on=(File.id == File2Document.file_id))\
.join(User, on=(cls.model.created_by == User.id), join_type=JOIN.LEFT_OUTER)\
.where(cls.model.kb_id == kb_id)
if run_status:

8
api/db/services/pipeline_operation_log_service.py
View File

@ -123,7 +123,7 @@ class PipelineOperationLogService(CommonService):
raise RuntimeError(f"Cannot find knowledge base {document.kb_id} for referred_document {referred_document_id}")
tenant_id = kb_info.tenant_id
title = document.name
title = document.parser_id
avatar = document.thumbnail
if task_type not in VALID_PIPELINE_TASK_TYPES:
@ -228,14 +228,12 @@ class PipelineOperationLogService(CommonService):
@classmethod
@DB.connection_context()
def get_documents_info(cls, id):
fields = [Document.id, Document.name, Document.progress]
fields = [Document.id, Document.name, Document.progress, Document.kb_id]
return (
cls.model.select(*fields)
.join(Document, on=(cls.model.document_id == Document.id))
.where(
cls.model.id == id,
Document.progress > 0,
Document.progress < 1,
cls.model.id == id
)
.dicts()
)

3
api/db/services/task_service.py
View File

@ -358,7 +358,7 @@ def queue_tasks(doc: dict, bucket: str, name: str, priority: int):
page_size = doc["parser_config"].get("task_page_size") or 12
if doc["parser_id"] == "paper":
page_size = doc["parser_config"].get("task_page_size") or 22
if doc["parser_id"] in ["one", "knowledge_graph"] or do_layout != "DeepDOC":
if doc["parser_id"] in ["one", "knowledge_graph"] or do_layout != "DeepDOC" or doc["parser_config"].get("toc", True):
page_size = 10 ** 9
page_ranges = doc["parser_config"].get("pages") or [(1, 10 ** 5)]
for s, e in page_ranges:
@ -505,7 +505,6 @@ def queue_dataflow(tenant_id:str, flow_id:str, task_id:str, doc_id:str=CANVAS_DE
task["kb_id"] = DocumentService.get_knowledgebase_id(doc_id)
task["tenant_id"] = tenant_id
task["task_type"] = "dataflow"
task["dataflow_id"] = flow_id
task["file"] = file