Fix: Unified terminology to "Pipeline" and optimized related component logic. #9869 (#10394)

### What problem does this PR solve?

Fix: Unified terminology to "Pipeline" and optimized related component
logic. #9869

- Added logic to clear pipeline_id when parseType changes in the chunk
method dialog.
- Fixed an issue in the Tooltip form component that prevented clicks
from triggering saves.
### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

admin/admin_client.py

@@ -1,9 +1,13 @@
 import argparse
 import base64
+
+from Cryptodome.PublicKey import RSA
+from Cryptodome.Cipher import PKCS1_v1_5 as Cipher_pkcs1_v1_5
 from typing import Dict, List, Any
 from lark import Lark, Transformer, Tree
 import requests
 from requests.auth import HTTPBasicAuth
+from api.common.base64 import encode_to_base64
 
 GRAMMAR = r"""
 start: command
@@ -19,6 +23,8 @@ sql_command: list_services
            | show_user
            | drop_user
            | alter_user
+           | create_user
+           | activate_user
            | list_datasets
            | list_agents
 
@@ -35,6 +41,7 @@ meta_arg: /[^\\s"']+/ | quoted_string
 LIST: "LIST"i
 SERVICES: "SERVICES"i
 SHOW: "SHOW"i
+CREATE: "CREATE"i
 SERVICE: "SERVICE"i
 SHUTDOWN: "SHUTDOWN"i
 STARTUP: "STARTUP"i
@@ -43,6 +50,7 @@ USERS: "USERS"i
 DROP: "DROP"i
 USER: "USER"i
 ALTER: "ALTER"i
+ACTIVE: "ACTIVE"i
 PASSWORD: "PASSWORD"i
 DATASETS: "DATASETS"i
 OF: "OF"i
@@ -58,12 +66,15 @@ list_users: LIST USERS ";"
 drop_user: DROP USER quoted_string ";"
 alter_user: ALTER USER PASSWORD quoted_string quoted_string ";"
 show_user: SHOW USER quoted_string ";"
+create_user: CREATE USER quoted_string quoted_string ";"
+activate_user: ALTER USER ACTIVE quoted_string status ";"
 
 list_datasets: LIST DATASETS OF quoted_string ";"
 list_agents: LIST AGENTS OF quoted_string ";"
 
 identifier: WORD
 quoted_string: QUOTED_STRING
+status: WORD
 
 QUOTED_STRING: /'[^']+'/ | /"[^"]+"/
 WORD: /[a-zA-Z0-9_\-\.]+/
@@ -118,6 +129,16 @@ class AdminTransformer(Transformer):
         new_password = items[4]
         return {"type": "alter_user", "username": user_name, "password": new_password}
 
+    def create_user(self, items):
+        user_name = items[2]
+        password = items[3]
+        return {"type": "create_user", "username": user_name, "password": password, "role": "user"}
+
+    def activate_user(self, items):
+        user_name = items[3]
+        activate_status = items[4]
+        return {"type": "activate_user", "activate_status": activate_status, "username": user_name}
+
     def list_datasets(self, items):
         user_name = items[3]
         return {"type": "list_datasets", "username": user_name}
@@ -147,9 +168,12 @@ class AdminTransformer(Transformer):
         return items
 
 
-def encode_to_base64(input_string):
-    base64_encoded = base64.b64encode(input_string.encode('utf-8'))
-    return base64_encoded.decode('utf-8')
+def encrypt(input_string):
+    pub = '-----BEGIN PUBLIC KEY-----\nMIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEArq9XTUSeYr2+N1h3Afl/z8Dse/2yD0ZGrKwx+EEEcdsBLca9Ynmx3nIB5obmLlSfmskLpBo0UACBmB5rEjBp2Q2f3AG3Hjd4B+gNCG6BDaawuDlgANIhGnaTLrIqWrrcm4EMzJOnAOI1fgzJRsOOUEfaS318Eq9OVO3apEyCCt0lOQK6PuksduOjVxtltDav+guVAA068NrPYmRNabVKRNLJpL8w4D44sfth5RvZ3q9t+6RTArpEtc5sh5ChzvqPOzKGMXW83C95TxmXqpbK6olN4RevSfVjEAgCydH6HN6OhtOQEcnrU97r9H0iZOWwbw3pVrZiUkuRD1R56Wzs2wIDAQAB\n-----END PUBLIC KEY-----'
+    pub_key = RSA.importKey(pub)
+    cipher = Cipher_pkcs1_v1_5.new(pub_key)
+    cipher_text = cipher.encrypt(base64.b64encode(input_string.encode('utf-8')))
+    return base64.b64encode(cipher_text).decode("utf-8")
 
 
 class AdminCommandParser:
@@ -220,6 +244,9 @@ class AdminCLI:
         if not data:
             print("No data to print")
             return
+        if isinstance(data, dict):
+            # handle single row data
+            data = [data]
 
         columns = list(data[0].keys())
         col_widths = {}
@@ -335,6 +362,10 @@ class AdminCLI:
                 self._handle_drop_user(command_dict)
             case 'alter_user':
                 self._handle_alter_user(command_dict)
+            case 'create_user':
+                self._handle_create_user(command_dict)
+            case 'activate_user':
+                self._handle_activate_user(command_dict)
             case 'list_datasets':
                 self._handle_list_datasets(command_dict)
             case 'list_agents':
@@ -349,9 +380,8 @@ class AdminCLI:
 
         url = f'http://{self.host}:{self.port}/api/v1/admin/services'
         response = requests.get(url, auth=HTTPBasicAuth(self.admin_account, self.admin_password))
-        res_json = dict
+        res_json = response.json()
         if response.status_code == 200:
-            res_json = response.json()
             self._print_table_simple(res_json['data'])
         else:
             print(f"Fail to get all users, code: {res_json['code']}, message: {res_json['message']}")
@@ -377,9 +407,8 @@ class AdminCLI:
 
         url = f'http://{self.host}:{self.port}/api/v1/admin/users'
         response = requests.get(url, auth=HTTPBasicAuth(self.admin_account, self.admin_password))
-        res_json = dict
+        res_json = response.json()
         if response.status_code == 200:
-            res_json = response.json()
             self._print_table_simple(res_json['data'])
         else:
             print(f"Fail to get all users, code: {res_json['code']}, message: {res_json['message']}")
@@ -388,11 +417,25 @@ class AdminCLI:
         username_tree: Tree = command['username']
         username: str = username_tree.children[0].strip("'\"")
         print(f"Showing user: {username}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{username}'
+        response = requests.get(url, auth=HTTPBasicAuth(self.admin_account, self.admin_password))
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(f"Fail to get user {username}, code: {res_json['code']}, message: {res_json['message']}")
 
     def _handle_drop_user(self, command):
         username_tree: Tree = command['username']
         username: str = username_tree.children[0].strip("'\"")
         print(f"Drop user: {username}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{username}'
+        response = requests.delete(url, auth=HTTPBasicAuth(self.admin_account, self.admin_password))
+        res_json = response.json()
+        if response.status_code == 200:
+            print(res_json["message"])
+        else:
+            print(f"Fail to drop user, code: {res_json['code']}, message: {res_json['message']}")
 
     def _handle_alter_user(self, command):
         username_tree: Tree = command['username']
@@ -400,16 +443,75 @@ class AdminCLI:
         password_tree: Tree = command['password']
         password: str = password_tree.children[0].strip("'\"")
         print(f"Alter user: {username}, password: {password}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{username}/password'
+        response = requests.put(url, auth=HTTPBasicAuth(self.admin_account, self.admin_password),
+                                json={'new_password': encrypt(password)})
+        res_json = response.json()
+        if response.status_code == 200:
+            print(res_json["message"])
+        else:
+            print(f"Fail to alter password, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _handle_create_user(self, command):
+        username_tree: Tree = command['username']
+        username: str = username_tree.children[0].strip("'\"")
+        password_tree: Tree = command['password']
+        password: str = password_tree.children[0].strip("'\"")
+        role: str = command['role']
+        print(f"Create user: {username}, password: {password}, role: {role}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/users'
+        response = requests.post(
+            url,
+            auth=HTTPBasicAuth(self.admin_account, self.admin_password),
+            json={'username': username, 'password': encrypt(password), 'role': role}
+        )
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(f"Fail to create user {username}, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _handle_activate_user(self, command):
+        username_tree: Tree = command['username']
+        username: str = username_tree.children[0].strip("'\"")
+        activate_tree: Tree = command['activate_status']
+        activate_status: str = activate_tree.children[0].strip("'\"")
+        if activate_status.lower() in ['on', 'off']:
+            print(f"Alter user {username} activate status, turn {activate_status.lower()}.")
+            url = f'http://{self.host}:{self.port}/api/v1/admin/users/{username}/activate'
+            response = requests.put(url, auth=HTTPBasicAuth(self.admin_account, self.admin_password),
+                                    json={'activate_status': activate_status})
+            res_json = response.json()
+            if response.status_code == 200:
+                print(res_json["message"])
+            else:
+                print(f"Fail to alter activate status, code: {res_json['code']}, message: {res_json['message']}")
+        else:
+            print(f"Unknown activate status: {activate_status}.")
 
     def _handle_list_datasets(self, command):
         username_tree: Tree = command['username']
         username: str = username_tree.children[0].strip("'\"")
         print(f"Listing all datasets of user: {username}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{username}/datasets'
+        response = requests.get(url, auth=HTTPBasicAuth(self.admin_account, self.admin_password))
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(f"Fail to get all datasets of {username}, code: {res_json['code']}, message: {res_json['message']}")
 
     def _handle_list_agents(self, command):
         username_tree: Tree = command['username']
         username: str = username_tree.children[0].strip("'\"")
         print(f"Listing all agents of user: {username}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{username}/agents'
+        response = requests.get(url, auth=HTTPBasicAuth(self.admin_account, self.admin_password))
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(f"Fail to get all agents of {username}, code: {res_json['code']}, message: {res_json['message']}")
 
     def _handle_meta_command(self, command):
         meta_command = command['command']
@@ -436,6 +538,7 @@ Commands:
   DROP USER <user>
   CREATE USER <user> <password>
   ALTER USER PASSWORD <user> <new_password>
+  ALTER USER ACTIVE <user> <on/off>
   LIST DATASETS OF <user>
   LIST AGENTS OF <user>
 

admin/admin_server.py

@@ -10,6 +10,7 @@ from flask import Flask
 from routes import admin_bp
 from api.utils.log_utils import init_root_logger
 from api.constants import SERVICE_CONF
+from api import settings
 from config import load_configurations, SERVICE_CONFIGS
 
 stop_event = threading.Event()
@@ -26,7 +27,7 @@ if __name__ == '__main__':
 
     app = Flask(__name__)
     app.register_blueprint(admin_bp)
-
+    settings.init_settings()
     SERVICE_CONFIGS.configs = load_configurations(SERVICE_CONF)
 
     try:

admin/config.py

@@ -4,7 +4,7 @@ from enum import Enum
 
 from pydantic import BaseModel
 from typing import Any
-from api.utils import read_config
+from api.utils.configs import read_config
 from urllib.parse import urlparse
 
 

admin/routes.py

@@ -1,7 +1,8 @@
 from flask import Blueprint, request
+
 from auth import login_verify
 from responses import success_response, error_response
-from services import UserMgr, ServiceMgr
+from services import UserMgr, ServiceMgr, UserServiceMgr
 from exceptions import AdminException
 
 admin_bp = Blueprint('admin', __name__, url_prefix='/api/v1/admin')
@@ -38,21 +39,29 @@ def create_user():
         password = data['password']
         role = data.get('role', 'user')
 
-        user = UserMgr.create_user(username, password, role)
-        return success_response(user, "User created successfully", 201)
+        res = UserMgr.create_user(username, password, role)
+        if res["success"]:
+            user_info = res["user_info"]
+            user_info.pop("password") # do not return password
+            return success_response(user_info, "User created successfully")
+        else:
+            return error_response("create user failed")
 
     except AdminException as e:
         return error_response(e.message, e.code)
     except Exception as e:
-        return error_response(str(e), 500)
+        return error_response(str(e))
 
 
 @admin_bp.route('/users/<username>', methods=['DELETE'])
 @login_verify
 def delete_user(username):
     try:
-        UserMgr.delete_user(username)
-        return success_response(None, "User and all data deleted successfully")
+        res = UserMgr.delete_user(username)
+        if res["success"]:
+            return success_response(None, res["message"])
+        else:
+            return error_response(res["message"])
 
     except AdminException as e:
         return error_response(e.message, e.code)
@@ -69,8 +78,8 @@ def change_password(username):
             return error_response("New password is required", 400)
 
         new_password = data['new_password']
-        UserMgr.update_user_password(username, new_password)
-        return success_response(None, "Password updated successfully")
+        msg = UserMgr.update_user_password(username, new_password)
+        return success_response(None, msg)
 
     except AdminException as e:
         return error_response(e.message, e.code)
@@ -78,6 +87,21 @@ def change_password(username):
         return error_response(str(e), 500)
 
 
+@admin_bp.route('/users/<username>/activate', methods=['PUT'])
+@login_verify
+def alter_user_activate_status(username):
+    try:
+        data = request.get_json()
+        if not data or 'activate_status' not in data:
+            return error_response("Activation status is required", 400)
+        activate_status = data['activate_status']
+        msg = UserMgr.update_user_activate_status(username, activate_status)
+        return success_response(None, msg)
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e), 500)
+
 @admin_bp.route('/users/<username>', methods=['GET'])
 @login_verify
 def get_user_details(username):
@@ -90,6 +114,31 @@ def get_user_details(username):
     except Exception as e:
         return error_response(str(e), 500)
 
+@admin_bp.route('/users/<username>/datasets', methods=['GET'])
+@login_verify
+def get_user_datasets(username):
+    try:
+        datasets_list = UserServiceMgr.get_user_datasets(username)
+        return success_response(datasets_list)
+
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users/<username>/agents', methods=['GET'])
+@login_verify
+def get_user_agents(username):
+    try:
+        agents_list = UserServiceMgr.get_user_agents(username)
+        return success_response(agents_list)
+
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e), 500)
+
 
 @admin_bp.route('/services', methods=['GET'])
 @login_verify

admin/services.py

@@ -1,5 +1,13 @@
+import re
+from werkzeug.security import check_password_hash
+from api.db import ActiveEnum
 from api.db.services import UserService
-from exceptions import AdminException
+from api.db.joint_services.user_account_service import create_new_user, delete_user_data
+from api.db.services.canvas_service import UserCanvasService
+from api.db.services.user_service import TenantService
+from api.db.services.knowledgebase_service import KnowledgebaseService
+from api.utils.crypt import decrypt
+from exceptions import AdminException, UserAlreadyExistsError, UserNotFoundError
 from config import SERVICE_CONFIGS
 
 class UserMgr:
@@ -13,19 +21,132 @@ class UserMgr:
 
     @staticmethod
     def get_user_details(username):
-        raise AdminException("get_user_details: not implemented")
+        # use email to query
+        users = UserService.query_user_by_email(username)
+        result = []
+        for user in users:
+            result.append({
+                'email': user.email,
+                'language': user.language,
+                'last_login_time': user.last_login_time,
+                'is_authenticated': user.is_authenticated,
+                'is_active': user.is_active,
+                'is_anonymous': user.is_anonymous,
+                'login_channel': user.login_channel,
+                'status': user.status,
+                'is_superuser': user.is_superuser,
+                'create_date': user.create_date,
+                'update_date': user.update_date
+            })
+        return result
 
     @staticmethod
-    def create_user(username, password, role="user"):
-        raise AdminException("create_user: not implemented")
+    def create_user(username, password, role="user") -> dict:
+        # Validate the email address
+        if not re.match(r"^[\w\._-]+@([\w_-]+\.)+[\w-]{2,}$", username):
+            raise AdminException(f"Invalid email address: {username}!")
+        # Check if the email address is already used
+        if UserService.query(email=username):
+            raise UserAlreadyExistsError(username)
+        # Construct user info data
+        user_info_dict = {
+            "email": username,
+            "nickname": "",  # ask user to edit it manually in settings.
+            "password": decrypt(password),
+            "login_channel": "password",
+            "is_superuser": role == "admin",
+        }
+        return create_new_user(user_info_dict)
 
     @staticmethod
     def delete_user(username):
-        raise AdminException("delete_user: not implemented")
+        # use email to delete
+        user_list = UserService.query_user_by_email(username)
+        if not user_list:
+            raise UserNotFoundError(username)
+        if len(user_list) > 1:
+            raise AdminException(f"Exist more than 1 user: {username}!")
+        usr = user_list[0]
+        return delete_user_data(usr.id)
 
     @staticmethod
-    def update_user_password(username, new_password):
-        raise AdminException("update_user_password: not implemented")
+    def update_user_password(username, new_password) -> str:
+        # use email to find user. check exist and unique.
+        user_list = UserService.query_user_by_email(username)
+        if not user_list:
+            raise UserNotFoundError(username)
+        elif len(user_list) > 1:
+            raise AdminException(f"Exist more than 1 user: {username}!")
+        # check new_password different from old.
+        usr = user_list[0]
+        psw = decrypt(new_password)
+        if check_password_hash(usr.password, psw):
+            return "Same password, no need to update!"
+        # update password
+        UserService.update_user_password(usr.id, psw)
+        return "Password updated successfully!"
+
+    @staticmethod
+    def update_user_activate_status(username, activate_status: str):
+        # use email to find user. check exist and unique.
+        user_list = UserService.query_user_by_email(username)
+        if not user_list:
+            raise UserNotFoundError(username)
+        elif len(user_list) > 1:
+            raise AdminException(f"Exist more than 1 user: {username}!")
+        # check activate status different from new
+        usr = user_list[0]
+        # format activate_status before handle
+        _activate_status = activate_status.lower()
+        target_status = {
+            'on': ActiveEnum.ACTIVE.value,
+            'off': ActiveEnum.INACTIVE.value,
+        }.get(_activate_status)
+        if not target_status:
+            raise AdminException(f"Invalid activate_status: {activate_status}")
+        if target_status == usr.is_active:
+            return f"User activate status is already {_activate_status}!"
+        # update is_active
+        UserService.update_user(usr.id, {"is_active": target_status})
+        return f"Turn {_activate_status} user activate status successfully!"
+
+class UserServiceMgr:
+
+    @staticmethod
+    def get_user_datasets(username):
+        # use email to find user.
+        user_list = UserService.query_user_by_email(username)
+        if not user_list:
+            raise UserNotFoundError(username)
+        elif len(user_list) > 1:
+            raise AdminException(f"Exist more than 1 user: {username}!")
+        # find tenants
+        usr = user_list[0]
+        tenants = TenantService.get_joined_tenants_by_user_id(usr.id)
+        tenant_ids = [m["tenant_id"] for m in tenants]
+        # filter permitted kb and owned kb
+        return KnowledgebaseService.get_all_kb_by_tenant_ids(tenant_ids, usr.id)
+
+    @staticmethod
+    def get_user_agents(username):
+        # use email to find user.
+        user_list = UserService.query_user_by_email(username)
+        if not user_list:
+            raise UserNotFoundError(username)
+        elif len(user_list) > 1:
+            raise AdminException(f"Exist more than 1 user: {username}!")
+        # find tenants
+        usr = user_list[0]
+        tenants = TenantService.get_joined_tenants_by_user_id(usr.id)
+        tenant_ids = [m["tenant_id"] for m in tenants]
+        # filter permitted agents and owned agents
+        res = UserCanvasService.get_all_agents_by_tenant_ids(tenant_ids, usr.id)
+        return [{
+            'title': r['title'],
+            'permission': r['permission'],
+            'canvas_type': r['canvas_type'],
+            'canvas_category': r['canvas_category']
+        } for r in res]
 
 class ServiceMgr:
 

agent/canvas.py

@@ -27,7 +27,7 @@ from agent.component import component_class
 from agent.component.base import ComponentBase
 from api.db.services.file_service import FileService
 from api.utils import get_uuid, hash_str2int
-from rag.prompts.prompts import chunks_format
+from rag.prompts.generator import chunks_format
 from rag.utils.redis_conn import REDIS_CONN
 
 class Graph:
@@ -153,6 +153,16 @@ class Graph:
     def get_tenant_id(self):
         return self._tenant_id
 
+    def get_variable_value(self, exp: str) -> Any:
+        exp = exp.strip("{").strip("}").strip(" ").strip("{").strip("}")
+        if exp.find("@") < 0:
+            return self.globals[exp]
+        cpn_id, var_nm = exp.split("@")
+        cpn = self.get_component(cpn_id)
+        if not cpn:
+            raise Exception(f"Can't find variable: '{cpn_id}@{var_nm}'")
+        return cpn["obj"].output(var_nm)
+
 
 class Canvas(Graph):
 
@@ -406,16 +416,6 @@ class Canvas(Graph):
             return False
         return True
 
-    def get_variable_value(self, exp: str) -> Any:
-        exp = exp.strip("{").strip("}").strip(" ").strip("{").strip("}")
-        if exp.find("@") < 0:
-            return self.globals[exp]
-        cpn_id, var_nm = exp.split("@")
-        cpn = self.get_component(cpn_id)
-        if not cpn:
-            raise Exception(f"Can't find variable: '{cpn_id}@{var_nm}'")
-        return cpn["obj"].output(var_nm)
-
     def get_history(self, window_size):
         convs = []
         if window_size <= 0:
@@ -490,7 +490,8 @@ class Canvas(Graph):
 
         r = self.retrieval[-1]
         for ck in chunks_format({"chunks": chunks}):
-            cid = hash_str2int(ck["id"], 100)
+            cid = hash_str2int(ck["id"], 500)
+            # cid = uuid.uuid5(uuid.NAMESPACE_DNS, ck["id"])
             if cid not in r:
                 r["chunks"][cid] = ck
 

agent/component/agent_with_tools.py

@@ -28,9 +28,8 @@ from api.db.services.llm_service import LLMBundle
 from api.db.services.tenant_llm_service import TenantLLMService
 from api.db.services.mcp_server_service import MCPServerService
 from api.utils.api_utils import timeout
-from rag.prompts import message_fit_in
-from rag.prompts.prompts import next_step, COMPLETE_TASK, analyze_task, \
-    citation_prompt, reflect, rank_memories, kb_prompt, citation_plus, full_question
+from rag.prompts.generator import next_step, COMPLETE_TASK, analyze_task, \
+    citation_prompt, reflect, rank_memories, kb_prompt, citation_plus, full_question, message_fit_in
 from rag.utils.mcp_tool_call_conn import MCPToolCallSession, mcp_tool_metadata_to_openai_tool
 from agent.component.llm import LLMParam, LLM
 
@@ -138,7 +137,7 @@ class Agent(LLM, ToolBase):
             res.update(cpn.get_input_form())
         return res
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 20*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 20*60)))
     def _invoke(self, **kwargs):
         if kwargs.get("user_prompt"):
             usr_pmt = ""

agent/component/base.py

@@ -244,7 +244,7 @@ class ComponentParamBase(ABC):
 
                 if not value_legal:
                     raise ValueError(
-                        "Plase check runtime conf, {} = {} does not match user-parameter restriction".format(
+                        "Please check runtime conf, {} = {} does not match user-parameter restriction".format(
                             variable, value
                         )
                     )
@@ -431,7 +431,7 @@ class ComponentBase(ABC):
         self.set_output("_elapsed_time", time.perf_counter() - self.output("_created_time"))
         return self.output()
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         raise NotImplementedError()
 

agent/component/categorize.py

@@ -28,7 +28,7 @@ from rag.llm.chat_model import ERROR_PREFIX
 class CategorizeParam(LLMParam):
 
     """
-    Define the Categorize component parameters.
+    Define the categorize component parameters.
     """
     def __init__(self):
         super().__init__()
@@ -80,7 +80,7 @@ Here's description of each category:
  - Prioritize the most specific applicable category
  - Return only the category name without explanations
  - Use "Other" only when no other category fits
- 
+
  """.format(
             "\n - ".join(list(self.category_description.keys())),
             "\n".join(descriptions)
@@ -96,7 +96,7 @@ Here's description of each category:
 class Categorize(LLM, ABC):
     component_name = "Categorize"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         msg = self._canvas.get_history(self._param.message_history_window_size)
         if not msg:
@@ -112,7 +112,7 @@ class Categorize(LLM, ABC):
 
         user_prompt = """
 ---- Real Data ----
-{} → 
+{} →
 """.format(" | ".join(["{}: \"{}\"".format(c["role"].upper(), re.sub(r"\n", "", c["content"], flags=re.DOTALL)) for c in msg]))
         ans = chat_mdl.chat(self._param.sys_prompt, [{"role": "user", "content": user_prompt}], self._param.gen_conf())
         logging.info(f"input: {user_prompt}, answer: {str(ans)}")
@@ -134,4 +134,4 @@ class Categorize(LLM, ABC):
         self.set_output("_next", cpn_ids)
 
     def thoughts(self) -> str:
-        return "Which should it falls into {}? ...".format(",".join([f"`{c}`" for c, _ in self._param.category_description.items()]))
+        return "Which should it falls into {}? ...".format(",".join([f"`{c}`" for c, _ in self._param.category_description.items()]))

agent/component/invoke.py

@@ -53,7 +53,7 @@ class InvokeParam(ComponentParamBase):
 class Invoke(ComponentBase, ABC):
     component_name = "Invoke"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3)))
     def _invoke(self, **kwargs):
         args = {}
         for para in self._param.variables:

agent/component/llm.py

@@ -26,8 +26,7 @@ from api.db.services.llm_service import LLMBundle
 from api.db.services.tenant_llm_service import TenantLLMService
 from agent.component.base import ComponentBase, ComponentParamBase
 from api.utils.api_utils import timeout
-from rag.prompts import message_fit_in, citation_prompt
-from rag.prompts.prompts import tool_call_summary
+from rag.prompts.generator import tool_call_summary, message_fit_in, citation_prompt
 
 
 class LLMParam(ComponentParamBase):
@@ -82,9 +81,9 @@ class LLMParam(ComponentParamBase):
 
 class LLM(ComponentBase):
     component_name = "LLM"
-    
-    def __init__(self, canvas, id, param: ComponentParamBase):
-        super().__init__(canvas, id, param)
+
+    def __init__(self, canvas, component_id, param: ComponentParamBase):
+        super().__init__(canvas, component_id, param)
         self.chat_mdl = LLMBundle(self._canvas.get_tenant_id(), TenantLLMService.llm_id2llm_type(self._param.llm_id),
                                   self._param.llm_id, max_retries=self._param.max_retries,
                                   retry_interval=self._param.delay_after_error
@@ -102,6 +101,8 @@ class LLM(ComponentBase):
 
     def get_input_elements(self) -> dict[str, Any]:
         res = self.get_input_elements_from_text(self._param.sys_prompt)
+        if isinstance(self._param.prompts, str):
+            self._param.prompts = [{"role": "user", "content": self._param.prompts}]
         for prompt in self._param.prompts:
             d = self.get_input_elements_from_text(prompt["content"])
             res.update(d)
@@ -113,6 +114,17 @@ class LLM(ComponentBase):
     def add2system_prompt(self, txt):
         self._param.sys_prompt += txt
 
+    def _sys_prompt_and_msg(self, msg, args):
+        if isinstance(self._param.prompts, str):
+            self._param.prompts = [{"role": "user", "content": self._param.prompts}]
+        for p in self._param.prompts:
+            if msg and msg[-1]["role"] == p["role"]:
+                continue
+            p = deepcopy(p)
+            p["content"] = self.string_format(p["content"], args)
+            msg.append(p)
+        return msg, self.string_format(self._param.sys_prompt, args)
+
     def _prepare_prompt_variables(self):
         if self._param.visual_files_var:
             self.imgs = self._canvas.get_variable_value(self._param.visual_files_var)
@@ -128,7 +140,6 @@ class LLM(ComponentBase):
 
         args = {}
         vars = self.get_input_elements() if not self._param.debug_inputs else self._param.debug_inputs
-        sys_prompt = self._param.sys_prompt
         for k, o in vars.items():
             args[k] = o["value"]
             if not isinstance(args[k], str):
@@ -138,16 +149,8 @@ class LLM(ComponentBase):
                     args[k] = str(args[k])
             self.set_input_value(k, args[k])
 
-        msg = self._canvas.get_history(self._param.message_history_window_size)[:-1]
-        for p in self._param.prompts:
-            if msg and msg[-1]["role"] == p["role"]:
-                continue
-            msg.append(deepcopy(p))
-
-        sys_prompt = self.string_format(sys_prompt, args)
+        msg, sys_prompt = self._sys_prompt_and_msg(self._canvas.get_history(self._param.message_history_window_size)[:-1], args)
         user_defined_prompt, sys_prompt = self._extract_prompts(sys_prompt)
-        for m in msg:
-            m["content"] = self.string_format(m["content"], args)
         if self._param.cite and self._canvas.get_reference()["chunks"]:
             sys_prompt += citation_prompt(user_defined_prompt)
 
@@ -202,7 +205,7 @@ class LLM(ComponentBase):
             for txt in self.chat_mdl.chat_streamly(msg[0]["content"], msg[1:], self._param.gen_conf(), images=self.imgs, **kwargs):
                 yield delta(txt)
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         def clean_formated_answer(ans: str) -> str:
             ans = re.sub(r"^.*</think>", "", ans, flags=re.DOTALL)
@@ -210,7 +213,7 @@ class LLM(ComponentBase):
             return re.sub(r"```\n*$", "", ans, flags=re.DOTALL)
 
         prompt, msg, _ = self._prepare_prompt_variables()
-        error = ""
+        error: str = ""
 
         if self._param.output_structure:
             prompt += "\nThe output MUST follow this JSON format:\n"+json.dumps(self._param.output_structure, ensure_ascii=False, indent=2)

agent/component/message.py

@@ -49,7 +49,7 @@ class MessageParam(ComponentParamBase):
 class Message(ComponentBase):
     component_name = "Message"
 
-    def get_kwargs(self, script:str, kwargs:dict = {}, delimeter:str=None) -> tuple[str, dict[str, str | list | Any]]:
+    def get_kwargs(self, script:str, kwargs:dict = {}, delimiter:str=None) -> tuple[str, dict[str, str | list | Any]]:
         for k,v in self.get_input_elements_from_text(script).items():
             if k in kwargs:
                 continue
@@ -60,8 +60,8 @@ class Message(ComponentBase):
             if isinstance(v, partial):
                 for t in v():
                     ans += t
-            elif isinstance(v, list) and delimeter:
-                ans = delimeter.join([str(vv) for vv in v])
+            elif isinstance(v, list) and delimiter:
+                ans = delimiter.join([str(vv) for vv in v])
             elif not isinstance(v, str):
                 try:
                     ans = json.dumps(v, ensure_ascii=False)
@@ -127,7 +127,7 @@ class Message(ComponentBase):
         ]
         return any([re.search(p, content) for p in patt])
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         rand_cnt = random.choice(self._param.content)
         if self._param.stream and not self._is_jinjia2(rand_cnt):

agent/component/string_transform.py

@@ -56,7 +56,7 @@ class StringTransform(Message, ABC):
             "type": "line"
         } for k, o in self.get_input_elements_from_text(self._param.script).items()}
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         if self._param.method == "split":
             self._split(kwargs.get("line"))
@@ -90,7 +90,7 @@ class StringTransform(Message, ABC):
         for k,v in kwargs.items():
             if not v:
                 v = ""
-            script = re.sub(k, v, script)
+            script = re.sub(k, lambda match: v, script)
 
         self.set_output("result", script)
 

agent/component/switch.py

@@ -61,7 +61,7 @@ class SwitchParam(ComponentParamBase):
 class Switch(ComponentBase, ABC):
     component_name = "Switch"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3)))
     def _invoke(self, **kwargs):
         for cond in self._param.conditions:
             res = []

agent/tools/arxiv.py

@@ -61,7 +61,7 @@ class ArXivParam(ToolParamBase):
 class ArXiv(ToolBase, ABC):
     component_name = "ArXiv"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
@@ -97,6 +97,6 @@ class ArXiv(ToolBase, ABC):
 
     def thoughts(self) -> str:
         return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
-                """.format(self.get_input().get("query", "-_-!"))
+                """.format(self.get_input().get("query", "-_-!"))

agent/tools/base.py

@@ -22,7 +22,7 @@ from typing import TypedDict, List, Any
 from agent.component.base import ComponentParamBase, ComponentBase
 from api.utils import hash_str2int
 from rag.llm.chat_model import ToolCallSession
-from rag.prompts.prompts import kb_prompt
+from rag.prompts.generator import kb_prompt
 from rag.utils.mcp_tool_call_conn import MCPToolCallSession
 from timeit import default_timer as timer
 

agent/tools/code_exec.py

@@ -129,7 +129,7 @@ module.exports = { main };
 class CodeExec(ToolBase, ABC):
     component_name = "CodeExec"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         lang = kwargs.get("lang", self._param.lang)
         script = kwargs.get("script", self._param.script)
@@ -157,7 +157,7 @@ class CodeExec(ToolBase, ABC):
 
         try:
             resp = requests.post(url=f"http://{settings.SANDBOX_HOST}:9385/run", json=code_req, timeout=os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
-            logging.info(f"http://{settings.SANDBOX_HOST}:9385/run,  code_req: {code_req}, resp.status_code {resp.status_code}:")
+            logging.info(f"http://{settings.SANDBOX_HOST}:9385/run", code_req, resp.status_code)
             if resp.status_code != 200:
                 resp.raise_for_status()
             body = resp.json()

agent/tools/duckduckgo.py

@@ -73,7 +73,7 @@ class DuckDuckGoParam(ToolParamBase):
 class DuckDuckGo(ToolBase, ABC):
     component_name = "DuckDuckGo"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
@@ -115,6 +115,6 @@ class DuckDuckGo(ToolBase, ABC):
 
     def thoughts(self) -> str:
         return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
-                """.format(self.get_input().get("query", "-_-!"))
+                """.format(self.get_input().get("query", "-_-!"))

agent/tools/email.py

@@ -98,8 +98,8 @@ class EmailParam(ToolParamBase):
 
 class Email(ToolBase, ABC):
     component_name = "Email"
-    
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60))
+
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
     def _invoke(self, **kwargs):
         if not kwargs.get("to_email"):
             self.set_output("success", False)
@@ -212,4 +212,4 @@ class Email(ToolBase, ABC):
 To: {}
 Subject: {}
 Your email is on its way—sit tight!
-""".format(inputs.get("to_email", "-_-!"), inputs.get("subject", "-_-!"))
+""".format(inputs.get("to_email", "-_-!"), inputs.get("subject", "-_-!"))

agent/tools/exesql.py

@@ -53,7 +53,7 @@ class ExeSQLParam(ToolParamBase):
         self.max_records = 1024
 
     def check(self):
-        self.check_valid_value(self.db_type, "Choose DB type", ['mysql', 'postgres', 'mariadb', 'mssql'])
+        self.check_valid_value(self.db_type, "Choose DB type", ['mysql', 'postgres', 'mariadb', 'mssql', 'IBM DB2'])
         self.check_empty(self.database, "Database name")
         self.check_empty(self.username, "database username")
         self.check_empty(self.host, "IP Address")
@@ -78,7 +78,7 @@ class ExeSQLParam(ToolParamBase):
 class ExeSQL(ToolBase, ABC):
     component_name = "ExeSQL"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
     def _invoke(self, **kwargs):
 
         def convert_decimals(obj):
@@ -123,6 +123,55 @@ class ExeSQL(ToolBase, ABC):
                     r'PWD=' + self._param.password
             )
             db = pyodbc.connect(conn_str)
+        elif self._param.db_type == 'IBM DB2':
+            import ibm_db
+            conn_str = (
+                f"DATABASE={self._param.database};"
+                f"HOSTNAME={self._param.host};"
+                f"PORT={self._param.port};"
+                f"PROTOCOL=TCPIP;"
+                f"UID={self._param.username};"
+                f"PWD={self._param.password};"
+            )
+            try:
+                conn = ibm_db.connect(conn_str, "", "")
+            except Exception as e:
+                raise Exception("Database Connection Failed! \n" + str(e))
+
+            sql_res = []
+            formalized_content = []
+            for single_sql in sqls:
+                single_sql = single_sql.replace("```", "").strip()
+                if not single_sql:
+                    continue
+                single_sql = re.sub(r"\[ID:[0-9]+\]", "", single_sql)
+
+                stmt = ibm_db.exec_immediate(conn, single_sql)
+                rows = []
+                row = ibm_db.fetch_assoc(stmt)
+                while row and len(rows) < self._param.max_records:
+                    rows.append(row)
+                    row = ibm_db.fetch_assoc(stmt)
+
+                if not rows:
+                    sql_res.append({"content": "No record in the database!"})
+                    continue
+
+                df = pd.DataFrame(rows)
+                for col in df.columns:
+                    if pd.api.types.is_datetime64_any_dtype(df[col]):
+                        df[col] = df[col].dt.strftime("%Y-%m-%d")
+
+                df = df.where(pd.notnull(df), None)
+
+                sql_res.append(convert_decimals(df.to_dict(orient="records")))
+                formalized_content.append(df.to_markdown(index=False, floatfmt=".6f"))
+
+            ibm_db.close(conn)
+
+            self.set_output("json", sql_res)
+            self.set_output("formalized_content", "\n\n".join(formalized_content))
+            return self.output("formalized_content")
         try:
             cursor = db.cursor()
         except Exception as e:
@@ -150,6 +199,8 @@ class ExeSQL(ToolBase, ABC):
                 if pd.api.types.is_datetime64_any_dtype(single_res[col]):
                     single_res[col] = single_res[col].dt.strftime('%Y-%m-%d')
 
+            single_res = single_res.where(pd.notnull(single_res), None)
+
             sql_res.append(convert_decimals(single_res.to_dict(orient='records')))
             formalized_content.append(single_res.to_markdown(index=False, floatfmt=".6f"))
 

agent/tools/github.py

@@ -57,7 +57,7 @@ class GitHubParam(ToolParamBase):
 class GitHub(ToolBase, ABC):
     component_name = "GitHub"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
@@ -88,4 +88,4 @@ class GitHub(ToolBase, ABC):
         assert False, self.output()
 
     def thoughts(self) -> str:
-        return "Scanning GitHub repos related to `{}`.".format(self.get_input().get("query", "-_-!"))
+        return "Scanning GitHub repos related to `{}`.".format(self.get_input().get("query", "-_-!"))

agent/tools/google.py

@@ -116,7 +116,7 @@ class GoogleParam(ToolParamBase):
 class Google(ToolBase, ABC):
     component_name = "Google"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("q"):
             self.set_output("formalized_content", "")
@@ -154,6 +154,6 @@ class Google(ToolBase, ABC):
 
     def thoughts(self) -> str:
         return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
-        """.format(self.get_input().get("query", "-_-!"))
+        """.format(self.get_input().get("query", "-_-!"))

agent/tools/googlescholar.py

@@ -63,7 +63,7 @@ class GoogleScholarParam(ToolParamBase):
 class GoogleScholar(ToolBase, ABC):
     component_name = "GoogleScholar"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
@@ -93,4 +93,4 @@ class GoogleScholar(ToolBase, ABC):
         assert False, self.output()
 
     def thoughts(self) -> str:
-        return "Looking for scholarly papers on `{}`,” prioritising reputable sources.".format(self.get_input().get("query", "-_-!"))
+        return "Looking for scholarly papers on `{}`,” prioritising reputable sources.".format(self.get_input().get("query", "-_-!"))

agent/tools/pubmed.py

@@ -33,7 +33,7 @@ class PubMedParam(ToolParamBase):
         self.meta:ToolMeta = {
             "name": "pubmed_search",
             "description": """
-PubMed is an openly accessible, free database which includes primarily the MEDLINE database of references and abstracts on life sciences and biomedical topics. 
+PubMed is an openly accessible, free database which includes primarily the MEDLINE database of references and abstracts on life sciences and biomedical topics.
 In addition to MEDLINE, PubMed provides access to:
  - older references from the print version of Index Medicus, back to 1951 and earlier
  - references to some journals before they were indexed in Index Medicus and MEDLINE, for instance Science, BMJ, and Annals of Surgery
@@ -69,7 +69,7 @@ In addition to MEDLINE, PubMed provides access to:
 class PubMed(ToolBase, ABC):
     component_name = "PubMed"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
@@ -105,4 +105,4 @@ class PubMed(ToolBase, ABC):
         assert False, self.output()
 
     def thoughts(self) -> str:
-        return "Looking for scholarly papers on `{}`,” prioritising reputable sources.".format(self.get_input().get("query", "-_-!"))
+        return "Looking for scholarly papers on `{}`,” prioritising reputable sources.".format(self.get_input().get("query", "-_-!"))

agent/tools/retrieval.py

@@ -23,8 +23,7 @@ from api.db.services.llm_service import LLMBundle
 from api import settings
 from api.utils.api_utils import timeout
 from rag.app.tag import label_question
-from rag.prompts import kb_prompt
-from rag.prompts.prompts import cross_languages
+from rag.prompts.generator import cross_languages, kb_prompt
 
 
 class RetrievalParam(ToolParamBase):
@@ -75,7 +74,7 @@ class RetrievalParam(ToolParamBase):
 class Retrieval(ToolBase, ABC):
     component_name = "Retrieval"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("formalized_content", self._param.empty_response)
@@ -163,13 +162,20 @@ class Retrieval(ToolBase, ABC):
             self.set_output("formalized_content", self._param.empty_response)
             return
 
+        # Format the chunks for JSON output (similar to how other tools do it)
+        json_output = kbinfos["chunks"].copy()
+
         self._canvas.add_reference(kbinfos["chunks"], kbinfos["doc_aggs"])
         form_cnt = "\n".join(kb_prompt(kbinfos, 200000, True))
+
+        # Set both formalized content and JSON output
         self.set_output("formalized_content", form_cnt)
+        self.set_output("json", json_output)
+
         return form_cnt
 
     def thoughts(self) -> str:
         return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
-        """.format(self.get_input().get("query", "-_-!"))
+        """.format(self.get_input().get("query", "-_-!"))

agent/tools/searxng.py

@@ -77,7 +77,7 @@ class SearXNGParam(ToolParamBase):
 class SearXNG(ToolBase, ABC):
     component_name = "SearXNG"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         # Gracefully handle try-run without inputs
         query = kwargs.get("query")
@@ -94,7 +94,6 @@ class SearXNG(ToolBase, ABC):
         last_e = ""
         for _ in range(self._param.max_retries+1):
             try:
-                # 构建搜索参数
                 search_params = {
                     'q': query,
                     'format': 'json',
@@ -104,33 +103,29 @@ class SearXNG(ToolBase, ABC):
                     'pageno': 1
                 }
 
-                # 发送搜索请求
                 response = requests.get(
                     f"{searxng_url}/search",
                     params=search_params,
                     timeout=10
                 )
                 response.raise_for_status()
-                
+
                 data = response.json()
-                
-                # 验证响应数据
+
                 if not data or not isinstance(data, dict):
                     raise ValueError("Invalid response from SearXNG")
-                
+
                 results = data.get("results", [])
                 if not isinstance(results, list):
                     raise ValueError("Invalid results format from SearXNG")
-                
-                # 限制结果数量
+
                 results = results[:self._param.top_n]
-                
-                # 处理搜索结果
+
                 self._retrieve_chunks(results,
                                       get_title=lambda r: r.get("title", ""),
                                       get_url=lambda r: r.get("url", ""),
                                       get_content=lambda r: r.get("content", ""))
-                
+
                 self.set_output("json", results)
                 return self.output("formalized_content")
 
@@ -151,6 +146,6 @@ class SearXNG(ToolBase, ABC):
 
     def thoughts(self) -> str:
         return """
-Keywords: {} 
+Keywords: {}
 Searching with SearXNG for relevant results...
                 """.format(self.get_input().get("query", "-_-!"))

agent/tools/tavily.py

@@ -31,7 +31,7 @@ class TavilySearchParam(ToolParamBase):
         self.meta:ToolMeta = {
             "name": "tavily_search",
             "description": """
-Tavily is a search engine optimized for LLMs, aimed at efficient, quick and persistent search results. 
+Tavily is a search engine optimized for LLMs, aimed at efficient, quick and persistent search results.
 When searching:
    - Start with specific query which should focus on just a single aspect.
    - Number of keywords in query should be less than 5.
@@ -101,7 +101,7 @@ When searching:
 class TavilySearch(ToolBase, ABC):
     component_name = "TavilySearch"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
@@ -136,7 +136,7 @@ class TavilySearch(ToolBase, ABC):
 
     def thoughts(self) -> str:
         return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
                 """.format(self.get_input().get("query", "-_-!"))
 
@@ -199,7 +199,7 @@ class TavilyExtractParam(ToolParamBase):
 class TavilyExtract(ToolBase, ABC):
     component_name = "TavilyExtract"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         self.tavily_client = TavilyClient(api_key=self._param.api_key)
         last_e = None
@@ -224,4 +224,4 @@ class TavilyExtract(ToolBase, ABC):
         assert False, self.output()
 
     def thoughts(self) -> str:
-        return "Opened {}—pulling out the main text…".format(self.get_input().get("urls", "-_-!"))
+        return "Opened {}—pulling out the main text…".format(self.get_input().get("urls", "-_-!"))

agent/tools/wencai.py

@@ -68,7 +68,7 @@ fund selection platform: through AI technology, is committed to providing excell
 class WenCai(ToolBase, ABC):
     component_name = "WenCai"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 12)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("report", "")
@@ -111,4 +111,4 @@ class WenCai(ToolBase, ABC):
         assert False, self.output()
 
     def thoughts(self) -> str:
-        return "Pulling live financial data for `{}`.".format(self.get_input().get("query", "-_-!"))
+        return "Pulling live financial data for `{}`.".format(self.get_input().get("query", "-_-!"))

agent/tools/wikipedia.py

@@ -64,7 +64,7 @@ class WikipediaParam(ToolParamBase):
 class Wikipedia(ToolBase, ABC):
     component_name = "Wikipedia"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
     def _invoke(self, **kwargs):
         if not kwargs.get("query"):
             self.set_output("formalized_content", "")
@@ -99,6 +99,6 @@ class Wikipedia(ToolBase, ABC):
 
     def thoughts(self) -> str:
         return """
-Keywords: {} 
+Keywords: {}
 Looking for the most relevant articles.
-        """.format(self.get_input().get("query", "-_-!"))
+        """.format(self.get_input().get("query", "-_-!"))

agent/tools/yahoofinance.py

@@ -72,7 +72,7 @@ class YahooFinanceParam(ToolParamBase):
 class YahooFinance(ToolBase, ABC):
     component_name = "YahooFinance"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 60)))
     def _invoke(self, **kwargs):
         if not kwargs.get("stock_code"):
             self.set_output("report", "")
@@ -111,4 +111,4 @@ class YahooFinance(ToolBase, ABC):
         assert False, self.output()
 
     def thoughts(self) -> str:
-        return "Pulling live financial data for `{}`.".format(self.get_input().get("stock_code", "-_-!"))
+        return "Pulling live financial data for `{}`.".format(self.get_input().get("stock_code", "-_-!"))

api/apps/__init__.py

@@ -27,7 +27,8 @@ from itsdangerous.url_safe import URLSafeTimedSerializer as Serializer
 from api.db import StatusEnum
 from api.db.db_models import close_connection
 from api.db.services import UserService
-from api.utils import CustomJSONEncoder, commands
+from api.utils.json import CustomJSONEncoder
+from api.utils import commands
 
 from flask_mail import Mail
 from flask_session import Session

api/apps/api_app.py

@@ -39,7 +39,7 @@ from api.utils.api_utils import server_error_response, get_data_error_result, ge
 
 from api.utils.file_utils import filename_type, thumbnail
 from rag.app.tag import label_question
-from rag.prompts import keyword_extraction
+from rag.prompts.generator import keyword_extraction
 from rag.utils.storage_factory import STORAGE_IMPL
 
 from api.db.services.canvas_service import UserCanvasService

api/apps/canvas_app.py

@@ -19,15 +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, TaskService
 from api.db.services.user_service import TenantService
 from api.db.services.user_canvas_version import UserCanvasVersionService
 from api.settings import RetCode
@@ -35,10 +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
 
 
@@ -48,14 +54,6 @@ def templates():
     return get_json_result(data=[c.to_dict() for c in CanvasTemplateService.query(canvas_category=CanvasCategory.Agent)])
 
 
-@manager.route('/list', methods=['GET'])  # noqa: F821
-@login_required
-def canvas_list():
-    return get_json_result(data=sorted([c.to_dict() for c in \
-                                 UserCanvasService.query(user_id=current_user.id, canvas_category=CanvasCategory.Agent)], key=lambda x: x["update_time"]*-1)
-                           )
-
-
 @manager.route('/rm', methods=['POST'])  # noqa: F821
 @validate_request("canvas_ids")
 @login_required
@@ -77,9 +75,10 @@ def save():
     if not isinstance(req["dsl"], str):
         req["dsl"] = json.dumps(req["dsl"], ensure_ascii=False)
     req["dsl"] = json.loads(req["dsl"])
+    cate = req.get("canvas_category", CanvasCategory.Agent)
     if "id" not in req:
         req["user_id"] = current_user.id
-        if UserCanvasService.query(user_id=current_user.id, title=req["title"].strip(), canvas_category=CanvasCategory.Agent):
+        if UserCanvasService.query(user_id=current_user.id, title=req["title"].strip(), canvas_category=cate):
             return get_data_error_result(message=f"{req['title'].strip()} already exists.")
         req["id"] = get_uuid()
         if not UserCanvasService.save(**req):
@@ -101,7 +100,7 @@ def save():
 def get(canvas_id):
     if not UserCanvasService.accessible(canvas_id, current_user.id):
         return get_data_error_result(message="canvas not found.")
-    e, c = UserCanvasService.get_by_tenant_id(canvas_id)
+    e, c = UserCanvasService.get_by_canvas_id(canvas_id)
     return get_json_result(data=c)
 
 
@@ -148,6 +147,14 @@ def run():
     if not isinstance(cvs.dsl, str):
         cvs.dsl = json.dumps(cvs.dsl, ensure_ascii=False)
 
+    if cvs.canvas_category == CanvasCategory.DataFlow:
+        task_id = get_uuid()
+        Pipeline(cvs.dsl, tenant_id=current_user.id, doc_id=CANVAS_DEBUG_DOC_ID, task_id=task_id, flow_id=req["id"])
+        ok, error_message = queue_dataflow(tenant_id=user_id, flow_id=req["id"], task_id=task_id, file=files[0], priority=0)
+        if not ok:
+            return get_data_error_result(message=error_message)
+        return get_json_result(data={"message_id": task_id})
+
     try:
         canvas = Canvas(cvs.dsl, current_user.id, req["id"])
     except Exception as e:
@@ -173,6 +180,44 @@ def run():
     return resp
 
 
+@manager.route('/rerun', methods=['POST'])  # noqa: F821
+@validate_request("id", "dsl", "component_id")
+@login_required
+def rerun():
+    req = request.json
+    doc = PipelineOperationLogService.get_documents_info(req["id"])
+    if not doc:
+        return get_data_error_result(message="Document not found.")
+    doc = doc[0]
+    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})
+    queue_dataflow(tenant_id=current_user.id, flow_id=req["id"], task_id=get_uuid(), doc_id=doc["id"], priority=0, rerun=True)
+    return get_json_result(data=True)
+
+
+@manager.route('/cancel/<task_id>', methods=['PUT'])  # noqa: F821
+@login_required
+def cancel(task_id):
+    try:
+        REDIS_CONN.set(f"{task_id}-cancel", "x")
+    except Exception as e:
+        logging.exception(e)
+    return get_json_result(data=True)
+
+
 @manager.route('/reset', methods=['POST'])  # noqa: F821
 @validate_request("id")
 @login_required
@@ -198,7 +243,7 @@ def reset():
 
 @manager.route("/upload/<canvas_id>", methods=["POST"])  # noqa: F821
 def upload(canvas_id):
-    e, cvs = UserCanvasService.get_by_tenant_id(canvas_id)
+    e, cvs = UserCanvasService.get_by_canvas_id(canvas_id)
     if not e:
         return get_data_error_result(message="canvas not found.")
 
@@ -348,6 +393,22 @@ def test_db_connect():
             cursor = db.cursor()
             cursor.execute("SELECT 1")
             cursor.close()
+        elif req["db_type"] == 'IBM DB2':
+            import ibm_db
+            conn_str = (
+                f"DATABASE={req['database']};"
+                f"HOSTNAME={req['host']};"
+                f"PORT={req['port']};"
+                f"PROTOCOL=TCPIP;"
+                f"UID={req['username']};"
+                f"PWD={req['password']};"
+            )
+            logging.info(conn_str)
+            conn = ibm_db.connect(conn_str, "", "")
+            stmt = ibm_db.exec_immediate(conn, "SELECT 1 FROM sysibm.sysdummy1")
+            ibm_db.fetch_assoc(stmt)
+            ibm_db.close(conn)
+            return get_json_result(data="Database Connection Successful!")
         else:
             return server_error_response("Unsupported database type.")
         if req["db_type"] != 'mssql':
@@ -383,22 +444,32 @@ def getversion( version_id):
         return get_json_result(data=f"Error getting history file: {e}")
 
 
-@manager.route('/listteam', methods=['GET'])  # noqa: F821
+@manager.route('/list', methods=['GET'])  # noqa: F821
 @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")
-    desc = request.args.get("desc", True)
-    try:
+    canvas_category = request.args.get("canvas_category")
+    if request.args.get("desc", "true").lower() == "false":
+        desc = False
+    else:
+        desc = True
+    owner_ids = [id for id in request.args.get("owner_ids", "").strip().split(",") if id]
+    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(
-            [m["tenant_id"] for m in tenants], current_user.id, page_number,
-            items_per_page, orderby, desc, keywords, canvas_category=CanvasCategory.Agent)
-        return get_json_result(data={"canvas": canvas, "total": total})
-    except Exception as e:
-        return server_error_response(e)
+            tenants, current_user.id, page_number,
+            items_per_page, orderby, desc, keywords, canvas_category)
+    else:
+        tenants = owner_ids
+        canvas, total = UserCanvasService.get_by_tenant_ids(
+            tenants, current_user.id, 0,
+            0, orderby, desc, keywords, canvas_category)
+    return get_json_result(data={"canvas": canvas, "total": total})
 
 
 @manager.route('/setting', methods=['POST'])  # noqa: F821
@@ -474,7 +545,7 @@ def sessions(canvas_id):
 @manager.route('/prompts', methods=['GET'])  # noqa: F821
 @login_required
 def prompts():
-    from rag.prompts.prompts import ANALYZE_TASK_SYSTEM, ANALYZE_TASK_USER, NEXT_STEP, REFLECT, CITATION_PROMPT_TEMPLATE
+    from rag.prompts.generator import ANALYZE_TASK_SYSTEM, ANALYZE_TASK_USER, NEXT_STEP, REFLECT, CITATION_PROMPT_TEMPLATE
     return get_json_result(data={
         "task_analysis": ANALYZE_TASK_SYSTEM +"\n\n"+ ANALYZE_TASK_USER,
         "plan_generation": NEXT_STEP,
@@ -483,3 +554,11 @@ 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)

api/apps/chunk_app.py

@@ -33,8 +33,7 @@ from api.utils.api_utils import get_data_error_result, get_json_result, server_e
 from rag.app.qa import beAdoc, rmPrefix
 from rag.app.tag import label_question
 from rag.nlp import rag_tokenizer, search
-from rag.prompts import cross_languages, keyword_extraction
-from rag.prompts.prompts import gen_meta_filter
+from rag.prompts.generator import gen_meta_filter, cross_languages, keyword_extraction
 from rag.settings import PAGERANK_FLD
 from rag.utils import rmSpace
 

api/apps/conversation_app.py

@@ -15,7 +15,7 @@
 #
 import json
 import re
-import traceback
+import logging
 from copy import deepcopy
 from flask import Response, request
 from flask_login import current_user, login_required
@@ -29,8 +29,8 @@ from api.db.services.search_service import SearchService
 from api.db.services.tenant_llm_service import TenantLLMService
 from api.db.services.user_service import TenantService, UserTenantService
 from api.utils.api_utils import get_data_error_result, get_json_result, server_error_response, validate_request
-from rag.prompts.prompt_template import load_prompt
-from rag.prompts.prompts import chunks_format
+from rag.prompts.template import load_prompt
+from rag.prompts.generator import chunks_format
 
 
 @manager.route("/set", methods=["POST"])  # noqa: F821
@@ -226,7 +226,7 @@ def completion():
                 if not is_embedded:
                     ConversationService.update_by_id(conv.id, conv.to_dict())
             except Exception as e:
-                traceback.print_exc()
+                logging.exception(e)
                 yield "data:" + json.dumps({"code": 500, "message": str(e), "data": {"answer": "**ERROR**: " + str(e), "reference": []}}, ensure_ascii=False) + "\n\n"
             yield "data:" + json.dumps({"code": 0, "message": "", "data": True}, ensure_ascii=False) + "\n\n"
 

api/apps/dataflow_app.py

@@ -1,353 +0,0 @@
-#
-#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
-#
-#  Licensed under the Apache License, Version 2.0 (the "License");
-#  you may not use this file except in compliance with the License.
-#  You may obtain a copy of the License at
-#
-#      http://www.apache.org/licenses/LICENSE-2.0
-#
-#  Unless required by applicable law or agreed to in writing, software
-#  distributed under the License is distributed on an "AS IS" BASIS,
-#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-#  See the License for the specific language governing permissions and
-#  limitations under the License.
-#
-import json
-import re
-import sys
-import time
-from functools import partial
-
-import trio
-from flask import request
-from flask_login import current_user, login_required
-
-from agent.canvas import Canvas
-from agent.component import LLM
-from api.db import CanvasCategory, FileType
-from api.db.services.canvas_service import CanvasTemplateService, UserCanvasService
-from api.db.services.document_service import DocumentService
-from api.db.services.file_service import FileService
-from api.db.services.task_service import queue_dataflow
-from api.db.services.user_canvas_version import UserCanvasVersionService
-from api.db.services.user_service import TenantService
-from api.settings import RetCode
-from api.utils import get_uuid
-from api.utils.api_utils import get_data_error_result, get_json_result, server_error_response, validate_request
-from api.utils.file_utils import filename_type, read_potential_broken_pdf
-from rag.flow.pipeline import Pipeline
-
-
-@manager.route("/templates", methods=["GET"])  # noqa: F821
-@login_required
-def templates():
-    return get_json_result(data=[c.to_dict() for c in CanvasTemplateService.query(canvas_category=CanvasCategory.DataFlow)])
-
-
-@manager.route("/list", methods=["GET"])  # noqa: F821
-@login_required
-def canvas_list():
-    return get_json_result(data=sorted([c.to_dict() for c in UserCanvasService.query(user_id=current_user.id, canvas_category=CanvasCategory.DataFlow)], key=lambda x: x["update_time"] * -1))
-
-
-@manager.route("/rm", methods=["POST"])  # noqa: F821
-@validate_request("canvas_ids")
-@login_required
-def rm():
-    for i in request.json["canvas_ids"]:
-        if not UserCanvasService.accessible(i, current_user.id):
-            return get_json_result(data=False, message="Only owner of canvas authorized for this operation.", code=RetCode.OPERATING_ERROR)
-        UserCanvasService.delete_by_id(i)
-    return get_json_result(data=True)
-
-
-@manager.route("/set", methods=["POST"])  # noqa: F821
-@validate_request("dsl", "title")
-@login_required
-def save():
-    req = request.json
-    if not isinstance(req["dsl"], str):
-        req["dsl"] = json.dumps(req["dsl"], ensure_ascii=False)
-    req["dsl"] = json.loads(req["dsl"])
-    req["canvas_category"] = CanvasCategory.DataFlow
-    if "id" not in req:
-        req["user_id"] = current_user.id
-        if UserCanvasService.query(user_id=current_user.id, title=req["title"].strip(), canvas_category=CanvasCategory.DataFlow):
-            return get_data_error_result(message=f"{req['title'].strip()} already exists.")
-        req["id"] = get_uuid()
-
-        if not UserCanvasService.save(**req):
-            return get_data_error_result(message="Fail to save canvas.")
-    else:
-        if not UserCanvasService.accessible(req["id"], current_user.id):
-            return get_json_result(data=False, message="Only owner of canvas authorized for this operation.", code=RetCode.OPERATING_ERROR)
-        UserCanvasService.update_by_id(req["id"], req)
-    # save version
-    UserCanvasVersionService.insert(user_canvas_id=req["id"], dsl=req["dsl"], title="{0}_{1}".format(req["title"], time.strftime("%Y_%m_%d_%H_%M_%S")))
-    UserCanvasVersionService.delete_all_versions(req["id"])
-    return get_json_result(data=req)
-
-
-@manager.route("/get/<canvas_id>", methods=["GET"])  # noqa: F821
-@login_required
-def get(canvas_id):
-    if not UserCanvasService.accessible(canvas_id, current_user.id):
-        return get_data_error_result(message="canvas not found.")
-    e, c = UserCanvasService.get_by_tenant_id(canvas_id)
-    return get_json_result(data=c)
-
-
-@manager.route("/run", methods=["POST"])  # noqa: F821
-@validate_request("id")
-@login_required
-def run():
-    req = request.json
-    flow_id = req.get("id", "")
-    doc_id = req.get("doc_id", "")
-    if not all([flow_id, doc_id]):
-        return get_data_error_result(message="id and doc_id are required.")
-
-    if not DocumentService.get_by_id(doc_id):
-        return get_data_error_result(message=f"Document for {doc_id} not found.")
-
-    user_id = req.get("user_id", current_user.id)
-    if not UserCanvasService.accessible(flow_id, current_user.id):
-        return get_json_result(data=False, message="Only owner of canvas authorized for this operation.", code=RetCode.OPERATING_ERROR)
-
-    e, cvs = UserCanvasService.get_by_id(flow_id)
-    if not e:
-        return get_data_error_result(message="canvas not found.")
-
-    if not isinstance(cvs.dsl, str):
-        cvs.dsl = json.dumps(cvs.dsl, ensure_ascii=False)
-
-    task_id = get_uuid()
-
-    ok, error_message = queue_dataflow(dsl=cvs.dsl, tenant_id=user_id, doc_id=doc_id, task_id=task_id, flow_id=flow_id, priority=0)
-    if not ok:
-        return server_error_response(error_message)
-
-    return get_json_result(data={"task_id": task_id, "flow_id": flow_id})
-
-
-@manager.route("/reset", methods=["POST"])  # noqa: F821
-@validate_request("id")
-@login_required
-def reset():
-    req = request.json
-    flow_id = req.get("id", "")
-    if not flow_id:
-        return get_data_error_result(message="id is required.")
-
-    if not UserCanvasService.accessible(flow_id, current_user.id):
-        return get_json_result(data=False, message="Only owner of canvas authorized for this operation.", code=RetCode.OPERATING_ERROR)
-
-    task_id = req.get("task_id", "")
-
-    try:
-        e, user_canvas = UserCanvasService.get_by_id(req["id"])
-        if not e:
-            return get_data_error_result(message="canvas not found.")
-
-        dataflow = Pipeline(dsl=json.dumps(user_canvas.dsl), tenant_id=current_user.id, flow_id=flow_id, task_id=task_id)
-        dataflow.reset()
-        req["dsl"] = json.loads(str(dataflow))
-        UserCanvasService.update_by_id(req["id"], {"dsl": req["dsl"]})
-        return get_json_result(data=req["dsl"])
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route("/upload/<canvas_id>", methods=["POST"])  # noqa: F821
-def upload(canvas_id):
-    e, cvs = UserCanvasService.get_by_tenant_id(canvas_id)
-    if not e:
-        return get_data_error_result(message="canvas not found.")
-
-    user_id = cvs["user_id"]
-
-    def structured(filename, filetype, blob, content_type):
-        nonlocal user_id
-        if filetype == FileType.PDF.value:
-            blob = read_potential_broken_pdf(blob)
-
-        location = get_uuid()
-        FileService.put_blob(user_id, location, blob)
-
-        return {
-            "id": location,
-            "name": filename,
-            "size": sys.getsizeof(blob),
-            "extension": filename.split(".")[-1].lower(),
-            "mime_type": content_type,
-            "created_by": user_id,
-            "created_at": time.time(),
-            "preview_url": None,
-        }
-
-    if request.args.get("url"):
-        from crawl4ai import AsyncWebCrawler, BrowserConfig, CrawlerRunConfig, CrawlResult, DefaultMarkdownGenerator, PruningContentFilter
-
-        try:
-            url = request.args.get("url")
-            filename = re.sub(r"\?.*", "", url.split("/")[-1])
-
-            async def adownload():
-                browser_config = BrowserConfig(
-                    headless=True,
-                    verbose=False,
-                )
-                async with AsyncWebCrawler(config=browser_config) as crawler:
-                    crawler_config = CrawlerRunConfig(markdown_generator=DefaultMarkdownGenerator(content_filter=PruningContentFilter()), pdf=True, screenshot=False)
-                    result: CrawlResult = await crawler.arun(url=url, config=crawler_config)
-                    return result
-
-            page = trio.run(adownload())
-            if page.pdf:
-                if filename.split(".")[-1].lower() != "pdf":
-                    filename += ".pdf"
-                return get_json_result(data=structured(filename, "pdf", page.pdf, page.response_headers["content-type"]))
-
-            return get_json_result(data=structured(filename, "html", str(page.markdown).encode("utf-8"), page.response_headers["content-type"], user_id))
-
-        except Exception as e:
-            return server_error_response(e)
-
-    file = request.files["file"]
-    try:
-        DocumentService.check_doc_health(user_id, file.filename)
-        return get_json_result(data=structured(file.filename, filename_type(file.filename), file.read(), file.content_type))
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route("/input_form", methods=["GET"])  # noqa: F821
-@login_required
-def input_form():
-    flow_id = request.args.get("id")
-    cpn_id = request.args.get("component_id")
-    try:
-        e, user_canvas = UserCanvasService.get_by_id(flow_id)
-        if not e:
-            return get_data_error_result(message="canvas not found.")
-        if not UserCanvasService.query(user_id=current_user.id, id=flow_id):
-            return get_json_result(data=False, message="Only owner of canvas authorized for this operation.", code=RetCode.OPERATING_ERROR)
-
-        dataflow = Pipeline(dsl=json.dumps(user_canvas.dsl), tenant_id=current_user.id, flow_id=flow_id, task_id="")
-
-        return get_json_result(data=dataflow.get_component_input_form(cpn_id))
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route("/debug", methods=["POST"])  # noqa: F821
-@validate_request("id", "component_id", "params")
-@login_required
-def debug():
-    req = request.json
-    if not UserCanvasService.accessible(req["id"], current_user.id):
-        return get_json_result(data=False, message="Only owner of canvas authorized for this operation.", code=RetCode.OPERATING_ERROR)
-    try:
-        e, user_canvas = UserCanvasService.get_by_id(req["id"])
-        canvas = Canvas(json.dumps(user_canvas.dsl), current_user.id)
-        canvas.reset()
-        canvas.message_id = get_uuid()
-        component = canvas.get_component(req["component_id"])["obj"]
-        component.reset()
-
-        if isinstance(component, LLM):
-            component.set_debug_inputs(req["params"])
-        component.invoke(**{k: o["value"] for k, o in req["params"].items()})
-        outputs = component.output()
-        for k in outputs.keys():
-            if isinstance(outputs[k], partial):
-                txt = ""
-                for c in outputs[k]():
-                    txt += c
-                outputs[k] = txt
-        return get_json_result(data=outputs)
-    except Exception as e:
-        return server_error_response(e)
-
-
-# api get list version dsl of canvas
-@manager.route("/getlistversion/<canvas_id>", methods=["GET"])  # noqa: F821
-@login_required
-def getlistversion(canvas_id):
-    try:
-        list = sorted([c.to_dict() for c in UserCanvasVersionService.list_by_canvas_id(canvas_id)], key=lambda x: x["update_time"] * -1)
-        return get_json_result(data=list)
-    except Exception as e:
-        return get_data_error_result(message=f"Error getting history files: {e}")
-
-
-# api get version dsl of canvas
-@manager.route("/getversion/<version_id>", methods=["GET"])  # noqa: F821
-@login_required
-def getversion(version_id):
-    try:
-        e, version = UserCanvasVersionService.get_by_id(version_id)
-        if version:
-            return get_json_result(data=version.to_dict())
-    except Exception as e:
-        return get_json_result(data=f"Error getting history file: {e}")
-
-
-@manager.route("/listteam", methods=["GET"])  # noqa: F821
-@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))
-    orderby = request.args.get("orderby", "create_time")
-    desc = request.args.get("desc", True)
-    try:
-        tenants = TenantService.get_joined_tenants_by_user_id(current_user.id)
-        canvas, total = UserCanvasService.get_by_tenant_ids(
-            [m["tenant_id"] for m in tenants], current_user.id, page_number, items_per_page, orderby, desc, keywords, canvas_category=CanvasCategory.DataFlow
-        )
-        return get_json_result(data={"canvas": canvas, "total": total})
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route("/setting", methods=["POST"])  # noqa: F821
-@validate_request("id", "title", "permission")
-@login_required
-def setting():
-    req = request.json
-    req["user_id"] = current_user.id
-
-    if not UserCanvasService.accessible(req["id"], current_user.id):
-        return get_json_result(data=False, message="Only owner of canvas authorized for this operation.", code=RetCode.OPERATING_ERROR)
-
-    e, flow = UserCanvasService.get_by_id(req["id"])
-    if not e:
-        return get_data_error_result(message="canvas not found.")
-    flow = flow.to_dict()
-    flow["title"] = req["title"]
-    for key in ("description", "permission", "avatar"):
-        if value := req.get(key):
-            flow[key] = value
-
-    num = UserCanvasService.update_by_id(req["id"], flow)
-    return get_json_result(data=num)
-
-
-@manager.route("/trace", methods=["GET"])  # noqa: F821
-def trace():
-    dataflow_id = request.args.get("dataflow_id")
-    task_id = request.args.get("task_id")
-    if not all([dataflow_id, task_id]):
-        return get_data_error_result(message="dataflow_id and task_id are required.")
-
-    e, dataflow_canvas = UserCanvasService.get_by_id(dataflow_id)
-    if not e:
-        return get_data_error_result(message="dataflow not found.")
-
-    dsl_str = json.dumps(dataflow_canvas.dsl, ensure_ascii=False)
-    dataflow = Pipeline(dsl=dsl_str, tenant_id=dataflow_canvas.user_id, flow_id=dataflow_id, task_id=task_id)
-    log = dataflow.fetch_logs()
-
-    return get_json_result(data=log)

api/apps/document_app.py

@@ -32,7 +32,7 @@ from api.db.services.document_service import DocumentService, doc_upload_and_par
 from api.db.services.file2document_service import File2DocumentService
 from api.db.services.file_service import FileService
 from api.db.services.knowledgebase_service import KnowledgebaseService
-from api.db.services.task_service import TaskService, cancel_all_task_of, queue_tasks
+from api.db.services.task_service import TaskService, cancel_all_task_of, queue_tasks, queue_dataflow
 from api.db.services.user_service import UserTenantService
 from api.utils import get_uuid
 from api.utils.api_utils import (
@@ -182,6 +182,7 @@ def create():
                 "id": get_uuid(),
                 "kb_id": kb.id,
                 "parser_id": kb.parser_id,
+                "pipeline_id": kb.pipeline_id,
                 "parser_config": kb.parser_config,
                 "created_by": current_user.id,
                 "type": FileType.VIRTUAL,
@@ -479,8 +480,11 @@ def run():
                         kb_table_num_map[kb_id] = count
                         if kb_table_num_map[kb_id] <= 0:
                             KnowledgebaseService.delete_field_map(kb_id)
-                bucket, name = File2DocumentService.get_storage_address(doc_id=doc["id"])
-                queue_tasks(doc, bucket, name, 0)
+                if doc.get("pipeline_id", ""):
+                    queue_dataflow(tenant_id, flow_id=doc["pipeline_id"], task_id=get_uuid(), doc_id=id)
+                else:
+                    bucket, name = File2DocumentService.get_storage_address(doc_id=doc["id"])
+                    queue_tasks(doc, bucket, name, 0)
 
         return get_json_result(data=True)
     except Exception as e:
@@ -546,31 +550,22 @@ def get(doc_id):
 
 @manager.route("/change_parser", methods=["POST"])  # noqa: F821
 @login_required
-@validate_request("doc_id", "parser_id")
+@validate_request("doc_id")
 def change_parser():
     req = request.json
 
     if not DocumentService.accessible(req["doc_id"], current_user.id):
         return get_json_result(data=False, message="No authorization.", code=settings.RetCode.AUTHENTICATION_ERROR)
-    try:
-        e, doc = DocumentService.get_by_id(req["doc_id"])
-        if not e:
-            return get_data_error_result(message="Document not found!")
-        if doc.parser_id.lower() == req["parser_id"].lower():
-            if "parser_config" in req:
-                if req["parser_config"] == doc.parser_config:
-                    return get_json_result(data=True)
-            else:
-                return get_json_result(data=True)
 
-        if (doc.type == FileType.VISUAL and req["parser_id"] != "picture") or (re.search(r"\.(ppt|pptx|pages)$", doc.name) and req["parser_id"] != "presentation"):
-            return get_data_error_result(message="Not supported yet!")
+    e, doc = DocumentService.get_by_id(req["doc_id"])
+    if not e:
+        return get_data_error_result(message="Document not found!")
 
+    def reset_doc():
+        nonlocal doc
         e = DocumentService.update_by_id(doc.id, {"parser_id": req["parser_id"], "progress": 0, "progress_msg": "", "run": TaskStatus.UNSTART.value})
         if not e:
             return get_data_error_result(message="Document not found!")
-        if "parser_config" in req:
-            DocumentService.update_parser_config(doc.id, req["parser_config"])
         if doc.token_num > 0:
             e = DocumentService.increment_chunk_num(doc.id, doc.kb_id, doc.token_num * -1, doc.chunk_num * -1, doc.process_duration * -1)
             if not e:
@@ -581,6 +576,26 @@ def change_parser():
             if settings.docStoreConn.indexExist(search.index_name(tenant_id), doc.kb_id):
                 settings.docStoreConn.delete({"doc_id": doc.id}, search.index_name(tenant_id), doc.kb_id)
 
+    try:
+        if "pipeline_id" in req:
+            if doc.pipeline_id == req["pipeline_id"]:
+                return get_json_result(data=True)
+            DocumentService.update_by_id(doc.id, {"pipeline_id": req["pipeline_id"]})
+            reset_doc()
+            return get_json_result(data=True)
+
+        if doc.parser_id.lower() == req["parser_id"].lower():
+            if "parser_config" in req:
+                if req["parser_config"] == doc.parser_config:
+                    return get_json_result(data=True)
+            else:
+                return get_json_result(data=True)
+
+        if (doc.type == FileType.VISUAL and req["parser_id"] != "picture") or (re.search(r"\.(ppt|pptx|pages)$", doc.name) and req["parser_id"] != "presentation"):
+            return get_data_error_result(message="Not supported yet!")
+        if "parser_config" in req:
+            DocumentService.update_parser_config(doc.id, req["parser_config"])
+        reset_doc()
         return get_json_result(data=True)
     except Exception as e:
         return server_error_response(e)

api/apps/file_app.py

@@ -246,6 +246,8 @@ def rm():
                 return get_data_error_result(message="File or Folder not found!")
             if not file.tenant_id:
                 return get_data_error_result(message="Tenant not found!")
+            if file.tenant_id != current_user.id:
+                return get_json_result(data=False, message='No authorization.', code=settings.RetCode.AUTHENTICATION_ERROR)
             if file.source_type == FileSource.KNOWLEDGEBASE:
                 continue
 
@@ -292,6 +294,8 @@ def rename():
         e, file = FileService.get_by_id(req["file_id"])
         if not e:
             return get_data_error_result(message="File not found!")
+        if file.tenant_id != current_user.id:
+            return get_json_result(data=False, message='No authorization.', code=settings.RetCode.AUTHENTICATION_ERROR)
         if file.type != FileType.FOLDER.value \
             and pathlib.Path(req["name"].lower()).suffix != pathlib.Path(
                 file.name.lower()).suffix:
@@ -328,6 +332,8 @@ def get(file_id):
         e, file = FileService.get_by_id(file_id)
         if not e:
             return get_data_error_result(message="Document not found!")
+        if file.tenant_id != current_user.id:
+            return get_json_result(data=False, message='No authorization.', code=settings.RetCode.AUTHENTICATION_ERROR)
 
         blob = STORAGE_IMPL.get(file.parent_id, file.location)
         if not blob:
@@ -367,6 +373,8 @@ def move():
                 return get_data_error_result(message="File or Folder not found!")
             if not file.tenant_id:
                 return get_data_error_result(message="Tenant not found!")
+            if file.tenant_id != current_user.id:
+                return get_json_result(data=False, message='No authorization.', code=settings.RetCode.AUTHENTICATION_ERROR)
         fe, _ = FileService.get_by_id(parent_id)
         if not fe:
             return get_data_error_result(message="Parent Folder not found!")

api/apps/kb_app.py

@@ -14,18 +14,21 @@
 #  limitations under the License.
 #
 import json
+import logging
 
 from flask import request
 from flask_login import login_required, current_user
 
 from api.db.services import duplicate_name
-from api.db.services.document_service import DocumentService
+from api.db.services.document_service import DocumentService, queue_raptor_o_graphrag_tasks
 from api.db.services.file2document_service import File2DocumentService
 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 TaskService, GRAPH_RAPTOR_FAKE_DOC_ID
 from api.db.services.user_service import TenantService, UserTenantService
-from api.utils.api_utils import server_error_response, get_data_error_result, validate_request, not_allowed_parameters
+from api.utils.api_utils import get_error_data_result, server_error_response, get_data_error_result, validate_request, not_allowed_parameters
 from api.utils import get_uuid
-from api.db import StatusEnum, FileSource
+from api.db import PipelineTaskType, StatusEnum, FileSource, VALID_FILE_TYPES, VALID_TASK_STATUS
 from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.db_models import File
 from api.utils.api_utils import get_json_result
@@ -35,7 +38,6 @@ from api.constants import DATASET_NAME_LIMIT
 from rag.settings import PAGERANK_FLD
 from rag.utils.storage_factory import STORAGE_IMPL
 
-
 @manager.route('/create', methods=['post'])  # noqa: F821
 @login_required
 @validate_request("name")
@@ -61,10 +63,39 @@ def create():
         req["name"] = dataset_name
         req["tenant_id"] = current_user.id
         req["created_by"] = current_user.id
+        if not req.get("parser_id"):
+            req["parser_id"] = "naive"
         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"]})
@@ -395,3 +426,352 @@ def get_basic_info():
     basic_info = DocumentService.knowledgebase_basic_info(kb_id)
 
     return get_json_result(data=basic_info)
+
+
+@manager.route("/list_pipeline_logs", methods=["POST"])  # noqa: F821
+@login_required
+def list_pipeline_logs():
+    kb_id = request.args.get("kb_id")
+    if not kb_id:
+        return get_json_result(data=False, message='Lack of "KB ID"', code=settings.RetCode.ARGUMENT_ERROR)
+
+    keywords = request.args.get("keywords", "")
+
+    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")
+    if request.args.get("desc", "true").lower() == "false":
+        desc = False
+    else:
+        desc = True
+    create_date_from = request.args.get("create_date_from", "")
+    create_date_to = request.args.get("create_date_to", "")
+    if create_date_to > create_date_from:
+        return get_data_error_result(message="Create data filter is abnormal.")
+
+    req = request.get_json()
+
+    operation_status = req.get("operation_status", [])
+    if operation_status:
+        invalid_status = {s for s in operation_status if s not in VALID_TASK_STATUS}
+        if invalid_status:
+            return get_data_error_result(message=f"Invalid filter operation_status status conditions: {', '.join(invalid_status)}")
+
+    types = req.get("types", [])
+    if types:
+        invalid_types = {t for t in types if t not in VALID_FILE_TYPES}
+        if invalid_types:
+            return get_data_error_result(message=f"Invalid filter conditions: {', '.join(invalid_types)} type{'s' if len(invalid_types) > 1 else ''}")
+
+    suffix = req.get("suffix", [])
+
+    try:
+        logs, tol = PipelineOperationLogService.get_file_logs_by_kb_id(kb_id, page_number, items_per_page, orderby, desc, keywords, operation_status, types, suffix, create_date_from, create_date_to)
+        return get_json_result(data={"total": tol, "logs": logs})
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route("/list_pipeline_dataset_logs", methods=["POST"])  # noqa: F821
+@login_required
+def list_pipeline_dataset_logs():
+    kb_id = request.args.get("kb_id")
+    if not kb_id:
+        return get_json_result(data=False, message='Lack of "KB ID"', code=settings.RetCode.ARGUMENT_ERROR)
+
+    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")
+    if request.args.get("desc", "true").lower() == "false":
+        desc = False
+    else:
+        desc = True
+    create_date_from = request.args.get("create_date_from", "")
+    create_date_to = request.args.get("create_date_to", "")
+    if create_date_to > create_date_from:
+        return get_data_error_result(message="Create data filter is abnormal.")
+
+    req = request.get_json()
+
+    operation_status = req.get("operation_status", [])
+    if operation_status:
+        invalid_status = {s for s in operation_status if s not in VALID_TASK_STATUS}
+        if invalid_status:
+            return get_data_error_result(message=f"Invalid filter operation_status status conditions: {', '.join(invalid_status)}")
+
+    try:
+        logs, tol = PipelineOperationLogService.get_dataset_logs_by_kb_id(kb_id, page_number, items_per_page, orderby, desc, operation_status, create_date_from, create_date_to)
+        return get_json_result(data={"total": tol, "logs": logs})
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route("/delete_pipeline_logs", methods=["POST"])  # noqa: F821
+@login_required
+def delete_pipeline_logs():
+    kb_id = request.args.get("kb_id")
+    if not kb_id:
+        return get_json_result(data=False, message='Lack of "KB ID"', code=settings.RetCode.ARGUMENT_ERROR)
+
+    req = request.get_json()
+    log_ids = req.get("log_ids", [])
+
+    PipelineOperationLogService.delete_by_ids(log_ids)
+
+    return get_json_result(data=True)
+
+
+@manager.route("/pipeline_log_detail", methods=["GET"])  # noqa: F821
+@login_required
+def pipeline_log_detail():
+    log_id = request.args.get("log_id")
+    if not log_id:
+        return get_json_result(data=False, message='Lack of "Pipeline log ID"', code=settings.RetCode.ARGUMENT_ERROR)
+
+    ok, log = PipelineOperationLogService.get_by_id(log_id)
+    if not ok:
+        return get_data_error_result(message="Invalid pipeline log ID")
+
+    return get_json_result(data=log.to_dict())
+
+
+@manager.route("/run_graphrag", methods=["POST"])  # noqa: F821
+@login_required
+def run_graphrag():
+    req = request.json
+
+    kb_id = req.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.graphrag_task_id
+    if task_id:
+        ok, task = TaskService.get_by_id(task_id)
+        if not ok:
+            logging.warning(f"A valid GraphRAG task id is expected for kb {kb_id}")
+
+        if task and task.progress not in [-1, 1]:
+            return get_error_data_result(message=f"Task {task_id} in progress with status {task.progress}. A Graph Task is already running.")
+
+    documents, _ = DocumentService.get_by_kb_id(
+        kb_id=kb_id,
+        page_number=0,
+        items_per_page=0,
+        orderby="create_time",
+        desc=False,
+        keywords="",
+        run_status=[],
+        types=[],
+        suffix=[],
+    )
+    if not documents:
+        return get_error_data_result(message=f"No documents in Knowledgebase {kb_id}")
+
+    sample_document = documents[0]
+    document_ids = [document["id"] for document in documents]
+
+    task_id = queue_raptor_o_graphrag_tasks(doc=sample_document, ty="graphrag", priority=0, fake_doc_id=GRAPH_RAPTOR_FAKE_DOC_ID, doc_ids=list(document_ids))
+
+    if not KnowledgebaseService.update_by_id(kb.id, {"graphrag_task_id": task_id}):
+        logging.warning(f"Cannot save graphrag_task_id for kb {kb_id}")
+
+    return get_json_result(data={"graphrag_task_id": task_id})
+
+
+@manager.route("/trace_graphrag", methods=["GET"])  # noqa: F821
+@login_required
+def trace_graphrag():
+    kb_id = request.args.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.graphrag_task_id
+    if not task_id:
+        return get_json_result(data={})
+
+    ok, task = TaskService.get_by_id(task_id)
+    if not ok:
+        return get_error_data_result(message="GraphRAG Task Not Found or Error Occurred")
+
+    return get_json_result(data=task.to_dict())
+
+
+@manager.route("/run_raptor", methods=["POST"])  # noqa: F821
+@login_required
+def run_raptor():
+    req = request.json
+
+    kb_id = req.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.raptor_task_id
+    if task_id:
+        ok, task = TaskService.get_by_id(task_id)
+        if not ok:
+            logging.warning(f"A valid RAPTOR task id is expected for kb {kb_id}")
+
+        if task and task.progress not in [-1, 1]:
+            return get_error_data_result(message=f"Task {task_id} in progress with status {task.progress}. A RAPTOR Task is already running.")
+
+    documents, _ = DocumentService.get_by_kb_id(
+        kb_id=kb_id,
+        page_number=0,
+        items_per_page=0,
+        orderby="create_time",
+        desc=False,
+        keywords="",
+        run_status=[],
+        types=[],
+        suffix=[],
+    )
+    if not documents:
+        return get_error_data_result(message=f"No documents in Knowledgebase {kb_id}")
+
+    sample_document = documents[0]
+    document_ids = [document["id"] for document in documents]
+
+    task_id = queue_raptor_o_graphrag_tasks(doc=sample_document, ty="raptor", priority=0, fake_doc_id=GRAPH_RAPTOR_FAKE_DOC_ID, doc_ids=list(document_ids))
+
+    if not KnowledgebaseService.update_by_id(kb.id, {"raptor_task_id": task_id}):
+        logging.warning(f"Cannot save raptor_task_id for kb {kb_id}")
+
+    return get_json_result(data={"raptor_task_id": task_id})
+
+
+@manager.route("/trace_raptor", methods=["GET"])  # noqa: F821
+@login_required
+def trace_raptor():
+    kb_id = request.args.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.raptor_task_id
+    if not task_id:
+        return get_json_result(data={})
+
+    ok, task = TaskService.get_by_id(task_id)
+    if not ok:
+        return get_error_data_result(message="RAPTOR Task Not Found or Error Occurred")
+
+    return get_json_result(data=task.to_dict())
+
+
+@manager.route("/run_mindmap", methods=["POST"])  # noqa: F821
+@login_required
+def run_mindmap():
+    req = request.json
+
+    kb_id = req.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.mindmap_task_id
+    if task_id:
+        ok, task = TaskService.get_by_id(task_id)
+        if not ok:
+            logging.warning(f"A valid Mindmap task id is expected for kb {kb_id}")
+
+        if task and task.progress not in [-1, 1]:
+            return get_error_data_result(message=f"Task {task_id} in progress with status {task.progress}. A Mindmap Task is already running.")
+
+    documents, _ = DocumentService.get_by_kb_id(
+        kb_id=kb_id,
+        page_number=0,
+        items_per_page=0,
+        orderby="create_time",
+        desc=False,
+        keywords="",
+        run_status=[],
+        types=[],
+        suffix=[],
+    )
+    if not documents:
+        return get_error_data_result(message=f"No documents in Knowledgebase {kb_id}")
+
+    sample_document = documents[0]
+    document_ids = [document["id"] for document in documents]
+
+    task_id = queue_raptor_o_graphrag_tasks(doc=sample_document, ty="mindmap", priority=0, fake_doc_id=GRAPH_RAPTOR_FAKE_DOC_ID, doc_ids=list(document_ids))
+
+    if not KnowledgebaseService.update_by_id(kb.id, {"mindmap_task_id": task_id}):
+        logging.warning(f"Cannot save mindmap_task_id for kb {kb_id}")
+
+    return get_json_result(data={"mindmap_task_id": task_id})
+
+
+@manager.route("/trace_mindmap", methods=["GET"])  # noqa: F821
+@login_required
+def trace_mindmap():
+    kb_id = request.args.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.mindmap_task_id
+    if not task_id:
+        return get_json_result(data={})
+
+    ok, task = TaskService.get_by_id(task_id)
+    if not ok:
+        return get_error_data_result(message="Mindmap Task Not Found or Error Occurred")
+
+    return get_json_result(data=task.to_dict())
+
+
+@manager.route("/unbind_task", methods=["DELETE"])  # noqa: F821
+@login_required
+def delete_kb_task():
+    kb_id = request.args.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_json_result(data=True)
+
+    pipeline_task_type = request.args.get("pipeline_task_type", "")
+    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")
+
+    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: "", kb_task_finish_at: None})
+    if not ok:
+        return server_error_response(f"Internal error: cannot delete task {pipeline_task_type}")
+
+    return get_json_result(data=True)

api/apps/sdk/doc.py

@@ -40,7 +40,7 @@ from api.utils.api_utils import check_duplicate_ids, construct_json_result, get_
 from rag.app.qa import beAdoc, rmPrefix
 from rag.app.tag import label_question
 from rag.nlp import rag_tokenizer, search
-from rag.prompts import cross_languages, keyword_extraction
+from rag.prompts.generator import cross_languages, keyword_extraction
 from rag.utils import rmSpace
 from rag.utils.storage_factory import STORAGE_IMPL
 

api/apps/sdk/files.py

@@ -83,16 +83,16 @@ def upload(tenant_id):
             return get_json_result(data=False, message="Can't find this folder!", code=404)
 
         for file_obj in file_objs:
-            # 文件路径处理
+            # Handle file path
             full_path = '/' + file_obj.filename
             file_obj_names = full_path.split('/')
             file_len = len(file_obj_names)
 
-            # 获取文件夹路径ID
+            # Get folder path ID
             file_id_list = FileService.get_id_list_by_id(pf_id, file_obj_names, 1, [pf_id])
             len_id_list = len(file_id_list)
 
-            # 创建文件夹结构
+            # Crete file folder
             if file_len != len_id_list:
                 e, file = FileService.get_by_id(file_id_list[len_id_list - 1])
                 if not e:

api/apps/sdk/session.py

@@ -38,9 +38,8 @@ from api.db.services.user_service import UserTenantService
 from api.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
 from rag.app.tag import label_question
-from rag.prompts import chunks_format
-from rag.prompts.prompt_template import load_prompt
-from rag.prompts.prompts import cross_languages, gen_meta_filter, keyword_extraction
+from rag.prompts.template import load_prompt
+from rag.prompts.generator import cross_languages, gen_meta_filter, keyword_extraction, chunks_format
 
 
 @manager.route("/chats/<chat_id>/sessions", methods=["POST"])  # noqa: F821

api/apps/system_app.py

@@ -39,6 +39,7 @@ from rag.utils.redis_conn import REDIS_CONN
 from flask import jsonify
 from api.utils.health_utils import run_health_checks
 
+
 @manager.route("/version", methods=["GET"])  # noqa: F821
 @login_required
 def version():

api/apps/user_app.py

@@ -34,7 +34,6 @@ from api.db.services.user_service import TenantService, UserService, UserTenantS
 from api.utils import (
     current_timestamp,
     datetime_format,
-    decrypt,
     download_img,
     get_format_time,
     get_uuid,
@@ -46,6 +45,7 @@ from api.utils.api_utils import (
     server_error_response,
     validate_request,
 )
+from api.utils.crypt import decrypt
 
 
 @manager.route("/login", methods=["POST", "GET"])  # noqa: F821
@@ -98,7 +98,14 @@ def login():
         return get_json_result(data=False, code=settings.RetCode.SERVER_ERROR, message="Fail to crypt password")
 
     user = UserService.query_user(email, password)
-    if user:
+
+    if user and hasattr(user, 'is_active') and user.is_active == "0":
+        return get_json_result(
+            data=False,
+            code=settings.RetCode.FORBIDDEN,
+            message="This account has been disabled, please contact the administrator!",
+        )
+    elif user:
         response_data = user.to_json()
         user.access_token = get_uuid()
         login_user(user)
@@ -227,6 +234,9 @@ def oauth_callback(channel):
         # User exists, try to log in
         user = users[0]
         user.access_token = get_uuid()
+        if user and hasattr(user, 'is_active') and user.is_active == "0":
+            return redirect("/?error=user_inactive")
+
         login_user(user)
         user.save()
         return redirect(f"/?auth={user.get_id()}")
@@ -317,6 +327,8 @@ def github_callback():
     # User has already registered, try to log in
     user = users[0]
     user.access_token = get_uuid()
+    if user and hasattr(user, 'is_active') and user.is_active == "0":
+        return redirect("/?error=user_inactive")
     login_user(user)
     user.save()
     return redirect("/?auth=%s" % user.get_id())
@@ -418,6 +430,8 @@ def feishu_callback():
 
     # User has already registered, try to log in
     user = users[0]
+    if user and hasattr(user, 'is_active') and user.is_active == "0":
+        return redirect("/?error=user_inactive")
     user.access_token = get_uuid()
     login_user(user)
     user.save()

api/common/README.md

@@ -0,0 +1,2 @@
+The python files in this directory are shared between service. They contain common utilities, models, and functions that can be used across various
+services to ensure consistency and reduce code duplication.

api/common/base64.py

@@ -0,0 +1,21 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+import base64
+
+def encode_to_base64(input_string):
+    base64_encoded = base64.b64encode(input_string.encode('utf-8'))
+    return base64_encoded.decode('utf-8')

api/db/__init__.py

@@ -23,6 +23,11 @@ class StatusEnum(Enum):
     INVALID = "0"
 
 
+class ActiveEnum(Enum):
+    ACTIVE = "1"
+    INACTIVE = "0"
+
+
 class UserTenantRole(StrEnum):
     OWNER = 'owner'
     ADMIN = 'admin'
@@ -111,7 +116,7 @@ class CanvasCategory(StrEnum):
     Agent = "agent_canvas"
     DataFlow = "dataflow_canvas"
 
-VALID_CAVAS_CATEGORIES = {CanvasCategory.Agent, CanvasCategory.DataFlow}
+VALID_CANVAS_CATEGORIES = {CanvasCategory.Agent, CanvasCategory.DataFlow}
 
 
 class MCPServerType(StrEnum):
@@ -122,4 +127,15 @@ class MCPServerType(StrEnum):
 VALID_MCP_SERVER_TYPES = {MCPServerType.SSE, MCPServerType.STREAMABLE_HTTP}
 
 
+class PipelineTaskType(StrEnum):
+    PARSE = "Parse"
+    DOWNLOAD = "Download"
+    RAPTOR = "RAPTOR"
+    GRAPH_RAG = "GraphRAG"
+    MINDMAP = "Mindmap"
+
+
+VALID_PIPELINE_TASK_TYPES = {PipelineTaskType.PARSE, PipelineTaskType.DOWNLOAD, PipelineTaskType.RAPTOR, PipelineTaskType.GRAPH_RAG, PipelineTaskType.MINDMAP}
+
+
 KNOWLEDGEBASE_FOLDER_NAME=".knowledgebase"

api/db/db_models.py

@@ -26,12 +26,14 @@ from functools import wraps
 
 from flask_login import UserMixin
 from itsdangerous.url_safe import URLSafeTimedSerializer as Serializer
-from peewee import BigIntegerField, BooleanField, CharField, CompositeKey, DateTimeField, Field, FloatField, IntegerField, Metadata, Model, TextField
+from peewee import InterfaceError, OperationalError, BigIntegerField, BooleanField, CharField, CompositeKey, DateTimeField, Field, FloatField, IntegerField, Metadata, Model, TextField
 from playhouse.migrate import MySQLMigrator, PostgresqlMigrator, migrate
 from playhouse.pool import PooledMySQLDatabase, PooledPostgresqlDatabase
 
 from api import settings, utils
 from api.db import ParserType, SerializedType
+from api.utils.json import json_dumps, json_loads
+from api.utils.configs import deserialize_b64, serialize_b64
 
 
 def singleton(cls, *args, **kw):
@@ -70,12 +72,12 @@ class JSONField(LongTextField):
     def db_value(self, value):
         if value is None:
             value = self.default_value
-        return utils.json_dumps(value)
+        return json_dumps(value)
 
     def python_value(self, value):
         if not value:
             return self.default_value
-        return utils.json_loads(value, object_hook=self._object_hook, object_pairs_hook=self._object_pairs_hook)
+        return json_loads(value, object_hook=self._object_hook, object_pairs_hook=self._object_pairs_hook)
 
 
 class ListField(JSONField):
@@ -91,21 +93,21 @@ class SerializedField(LongTextField):
 
     def db_value(self, value):
         if self._serialized_type == SerializedType.PICKLE:
-            return utils.serialize_b64(value, to_str=True)
+            return serialize_b64(value, to_str=True)
         elif self._serialized_type == SerializedType.JSON:
             if value is None:
                 return None
-            return utils.json_dumps(value, with_type=True)
+            return json_dumps(value, with_type=True)
         else:
             raise ValueError(f"the serialized type {self._serialized_type} is not supported")
 
     def python_value(self, value):
         if self._serialized_type == SerializedType.PICKLE:
-            return utils.deserialize_b64(value)
+            return deserialize_b64(value)
         elif self._serialized_type == SerializedType.JSON:
             if value is None:
                 return {}
-            return utils.json_loads(value, object_hook=self._object_hook, object_pairs_hook=self._object_pairs_hook)
+            return json_loads(value, object_hook=self._object_hook, object_pairs_hook=self._object_pairs_hook)
         else:
             raise ValueError(f"the serialized type {self._serialized_type} is not supported")
 
@@ -250,36 +252,63 @@ class RetryingPooledMySQLDatabase(PooledMySQLDatabase):
         super().__init__(*args, **kwargs)
 
     def execute_sql(self, sql, params=None, commit=True):
-        from peewee import OperationalError
-
         for attempt in range(self.max_retries + 1):
             try:
                 return super().execute_sql(sql, params, commit)
-            except OperationalError as e:
-                if e.args[0] in (2013, 2006) and attempt < self.max_retries:
-                    logging.warning(f"Lost connection (attempt {attempt + 1}/{self.max_retries}): {e}")
+            except (OperationalError, InterfaceError) as e:
+                error_codes = [2013, 2006]
+                error_messages = ['', 'Lost connection']
+                should_retry = (
+                    (hasattr(e, 'args') and e.args and e.args[0] in error_codes) or
+                    (str(e) in error_messages) or
+                    (hasattr(e, '__class__') and e.__class__.__name__ == 'InterfaceError')
+                )
+
+                if should_retry and attempt < self.max_retries:
+                    logging.warning(
+                        f"Database connection issue (attempt {attempt+1}/{self.max_retries}): {e}"
+                    )
                     self._handle_connection_loss()
-                    time.sleep(self.retry_delay * (2**attempt))
+                    time.sleep(self.retry_delay * (2 ** attempt))
                 else:
                     logging.error(f"DB execution failure: {e}")
                     raise
         return None
 
     def _handle_connection_loss(self):
-        self.close_all()
-        self.connect()
+        # self.close_all()
+        # self.connect()
+        try:
+            self.close()
+        except Exception:
+            pass
+        try:
+            self.connect()
+        except Exception as e:
+            logging.error(f"Failed to reconnect: {e}")
+            time.sleep(0.1)
+            self.connect()
 
     def begin(self):
-        from peewee import OperationalError
-
         for attempt in range(self.max_retries + 1):
             try:
                 return super().begin()
-            except OperationalError as e:
-                if e.args[0] in (2013, 2006) and attempt < self.max_retries:
-                    logging.warning(f"Lost connection during transaction (attempt {attempt + 1}/{self.max_retries})")
+            except (OperationalError, InterfaceError) as e:
+                error_codes = [2013, 2006]
+                error_messages = ['', 'Lost connection']
+
+                should_retry = (
+                    (hasattr(e, 'args') and e.args and e.args[0] in error_codes) or
+                    (str(e) in error_messages) or
+                    (hasattr(e, '__class__') and e.__class__.__name__ == 'InterfaceError')
+                )
+
+                if should_retry and attempt < self.max_retries:
+                    logging.warning(
+                        f"Lost connection during transaction (attempt {attempt+1}/{self.max_retries})"
+                    )
                     self._handle_connection_loss()
-                    time.sleep(self.retry_delay * (2**attempt))
+                    time.sleep(self.retry_delay * (2 ** attempt))
                 else:
                     raise
 
@@ -299,7 +328,16 @@ class BaseDataBase:
     def __init__(self):
         database_config = settings.DATABASE.copy()
         db_name = database_config.pop("name")
-        self.database_connection = PooledDatabase[settings.DATABASE_TYPE.upper()].value(db_name, **database_config)
+        
+        pool_config = {
+            'max_retries': 5,
+            'retry_delay': 1,
+        }
+        database_config.update(pool_config)
+        self.database_connection = PooledDatabase[settings.DATABASE_TYPE.upper()].value(
+            db_name, **database_config
+        )
+        # self.database_connection = PooledDatabase[settings.DATABASE_TYPE.upper()].value(db_name, **database_config)
         logging.info("init database on cluster mode successfully")
 
 
@@ -646,8 +684,17 @@ class Knowledgebase(DataBaseModel):
     vector_similarity_weight = FloatField(default=0.3, index=True)
 
     parser_id = CharField(max_length=32, null=False, help_text="default parser ID", default=ParserType.NAIVE.value, index=True)
+    pipeline_id = CharField(max_length=32, null=True, help_text="Pipeline ID", index=True)
     parser_config = JSONField(null=False, default={"pages": [[1, 1000000]]})
     pagerank = IntegerField(default=0, index=False)
+
+    graphrag_task_id = CharField(max_length=32, null=True, help_text="Graph RAG task ID", index=True)
+    graphrag_task_finish_at = DateTimeField(null=True)
+    raptor_task_id = CharField(max_length=32, null=True, help_text="RAPTOR task ID", index=True)
+    raptor_task_finish_at = DateTimeField(null=True)
+    mindmap_task_id = CharField(max_length=32, null=True, help_text="Mindmap task ID", index=True)
+    mindmap_task_finish_at = DateTimeField(null=True)
+
     status = CharField(max_length=1, null=True, help_text="is it validate(0: wasted, 1: validate)", default="1", index=True)
 
     def __str__(self):
@@ -662,6 +709,7 @@ class Document(DataBaseModel):
     thumbnail = TextField(null=True, help_text="thumbnail base64 string")
     kb_id = CharField(max_length=256, null=False, index=True)
     parser_id = CharField(max_length=32, null=False, help_text="default parser ID", index=True)
+    pipeline_id = CharField(max_length=32, null=True, help_text="pipleline ID", index=True)
     parser_config = JSONField(null=False, default={"pages": [[1, 1000000]]})
     source_type = CharField(max_length=128, null=False, default="local", help_text="where dose this document come from", index=True)
     type = CharField(max_length=32, null=False, help_text="file extension", index=True)
@@ -904,6 +952,32 @@ class Search(DataBaseModel):
         db_table = "search"
 
 
+class PipelineOperationLog(DataBaseModel):
+    id = CharField(max_length=32, primary_key=True)
+    document_id = CharField(max_length=32, index=True)
+    tenant_id = CharField(max_length=32, null=False, index=True)
+    kb_id = CharField(max_length=32, null=False, index=True)
+    pipeline_id = CharField(max_length=32, null=True, help_text="Pipeline ID", index=True)
+    pipeline_title = CharField(max_length=32, null=True, help_text="Pipeline title", index=True)
+    parser_id = CharField(max_length=32, null=False, help_text="Parser ID", index=True)
+    document_name = CharField(max_length=255, null=False, help_text="File name")
+    document_suffix = CharField(max_length=255, null=False, help_text="File suffix")
+    document_type = CharField(max_length=255, null=False, help_text="Document type")
+    source_from = CharField(max_length=255, null=False, help_text="Source")
+    progress = FloatField(default=0, index=True)
+    progress_msg = TextField(null=True, help_text="process message", default="")
+    process_begin_at = DateTimeField(null=True, index=True)
+    process_duration = FloatField(default=0)
+    dsl = JSONField(null=True, default=dict)
+    task_type = CharField(max_length=32, null=False, default="")
+    operation_status = CharField(max_length=32, null=False, help_text="Operation status")
+    avatar = TextField(null=True, help_text="avatar base64 string")
+    status = CharField(max_length=1, null=True, help_text="is it validate(0: wasted, 1: validate)", default="1", index=True)
+
+    class Meta:
+        db_table = "pipeline_operation_log"
+
+
 def migrate_db():
     logging.disable(logging.ERROR)
     migrator = DatabaseMigrator[settings.DATABASE_TYPE.upper()].value(DB)
@@ -1020,7 +1094,6 @@ def migrate_db():
         migrate(migrator.add_column("dialog", "meta_data_filter", JSONField(null=True, default={})))
     except Exception:
         pass
-
     try:
         migrate(migrator.alter_column_type("canvas_template", "title", JSONField(null=True, default=dict, help_text="Canvas title")))
     except Exception:
@@ -1037,4 +1110,36 @@ def migrate_db():
         migrate(migrator.add_column("canvas_template", "canvas_category", CharField(max_length=32, null=False, default="agent_canvas", help_text="agent_canvas|dataflow_canvas", index=True)))
     except Exception:
         pass
+    try:
+        migrate(migrator.add_column("knowledgebase", "pipeline_id", CharField(max_length=32, null=True, help_text="Pipeline ID", index=True)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.add_column("document", "pipeline_id", CharField(max_length=32, null=True, help_text="Pipeline ID", index=True)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.add_column("knowledgebase", "graphrag_task_id", CharField(max_length=32, null=True, help_text="Gragh RAG task ID", index=True)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.add_column("knowledgebase", "raptor_task_id", CharField(max_length=32, null=True, help_text="RAPTOR task ID", index=True)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.add_column("knowledgebase", "graphrag_task_finish_at", DateTimeField(null=True)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.add_column("knowledgebase", "raptor_task_finish_at", CharField(null=True)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.add_column("knowledgebase", "mindmap_task_id", CharField(max_length=32, null=True, help_text="Mindmap task ID", index=True)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.add_column("knowledgebase", "mindmap_task_finish_at", CharField(null=True)))
+    except Exception:
+        pass
     logging.disable(logging.NOTSET)

api/db/init_data.py

@@ -14,7 +14,6 @@
 #  limitations under the License.
 #
 import logging
-import base64
 import json
 import os
 import time
@@ -32,11 +31,7 @@ from api.db.services.llm_service import LLMService, LLMBundle, get_init_tenant_l
 from api.db.services.user_service import TenantService, UserTenantService
 from api import settings
 from api.utils.file_utils import get_project_base_directory
-
-
-def encode_to_base64(input_string):
-    base64_encoded = base64.b64encode(input_string.encode('utf-8'))
-    return base64_encoded.decode('utf-8')
+from api.common.base64 import encode_to_base64
 
 
 def init_superuser():

api/db/joint_services/user_account_service.py

@@ -0,0 +1,327 @@
+#
+#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+import logging
+import uuid
+
+from api import settings
+from api.utils.api_utils import group_by
+from api.db import FileType, UserTenantRole, ActiveEnum
+from api.db.services.api_service import APITokenService, API4ConversationService
+from api.db.services.canvas_service import UserCanvasService
+from api.db.services.conversation_service import ConversationService
+from api.db.services.dialog_service import DialogService
+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.db.services.langfuse_service import TenantLangfuseService
+from api.db.services.llm_service import get_init_tenant_llm
+from api.db.services.file_service import FileService
+from api.db.services.mcp_server_service import MCPServerService
+from api.db.services.search_service import SearchService
+from api.db.services.task_service import TaskService
+from api.db.services.tenant_llm_service import TenantLLMService
+from api.db.services.user_canvas_version import UserCanvasVersionService
+from api.db.services.user_service import TenantService, UserService, UserTenantService
+from rag.utils.storage_factory import STORAGE_IMPL
+from rag.nlp import search
+
+
+def create_new_user(user_info: dict) -> dict:
+    """
+    Add a new user, and create tenant, tenant llm, file folder for new user.
+    :param user_info: {
+        "email": <example@example.com>,
+        "nickname": <str, "name">,
+        "password": <decrypted password>,
+        "login_channel": <enum, "password">,
+        "is_superuser": <bool, role == "admin">,
+    }
+    :return: {
+        "success": <bool>,
+        "user_info": <dict>, # if true, return user_info
+    }
+    """
+    # generate user_id and access_token for user
+    user_id = uuid.uuid1().hex
+    user_info['id'] = user_id
+    user_info['access_token'] = uuid.uuid1().hex
+    # construct tenant info
+    tenant = {
+        "id": user_id,
+        "name": user_info["nickname"] + "‘s Kingdom",
+        "llm_id": settings.CHAT_MDL,
+        "embd_id": settings.EMBEDDING_MDL,
+        "asr_id": settings.ASR_MDL,
+        "parser_ids": settings.PARSERS,
+        "img2txt_id": settings.IMAGE2TEXT_MDL,
+        "rerank_id": settings.RERANK_MDL,
+    }
+    usr_tenant = {
+        "tenant_id": user_id,
+        "user_id": user_id,
+        "invited_by": user_id,
+        "role": UserTenantRole.OWNER,
+    }
+    # construct file folder info
+    file_id = uuid.uuid1().hex
+    file = {
+        "id": file_id,
+        "parent_id": file_id,
+        "tenant_id": user_id,
+        "created_by": user_id,
+        "name": "/",
+        "type": FileType.FOLDER.value,
+        "size": 0,
+        "location": "",
+    }
+    try:
+        tenant_llm = get_init_tenant_llm(user_id)
+
+        if not UserService.save(**user_info):
+            return {"success": False}
+
+        TenantService.insert(**tenant)
+        UserTenantService.insert(**usr_tenant)
+        TenantLLMService.insert_many(tenant_llm)
+        FileService.insert(file)
+
+        return {
+            "success": True,
+            "user_info": user_info,
+        }
+
+    except Exception as create_error:
+        logging.exception(create_error)
+        # rollback
+        try:
+            TenantService.delete_by_id(user_id)
+        except Exception as e:
+            logging.exception(e)
+        try:
+            u = UserTenantService.query(tenant_id=user_id)
+            if u:
+                UserTenantService.delete_by_id(u[0].id)
+        except Exception as e:
+            logging.exception(e)
+        try:
+            TenantLLMService.delete_by_tenant_id(user_id)
+        except Exception as e:
+            logging.exception(e)
+        try:
+            FileService.delete_by_id(file["id"])
+        except Exception as e:
+            logging.exception(e)
+        # delete user row finally
+        try:
+            UserService.delete_by_id(user_id)
+        except Exception as e:
+            logging.exception(e)
+        # reraise
+        raise create_error
+
+
+def delete_user_data(user_id: str) -> dict:
+    # use user_id to delete
+    usr = UserService.filter_by_id(user_id)
+    if not usr:
+        return {"success": False, "message": f"{user_id} can't be found."}
+    # check is inactive and not admin
+    if usr.is_active == ActiveEnum.ACTIVE.value:
+        return {"success": False, "message": f"{user_id} is active and can't be deleted."}
+    if usr.is_superuser:
+        return {"success": False, "message": "Can't delete the super user."}
+    # tenant info
+    tenants = UserTenantService.get_user_tenant_relation_by_user_id(usr.id)
+    owned_tenant = [t for t in tenants if t["role"] == UserTenantRole.OWNER.value]
+
+    done_msg = ''
+    try:
+        # step1. delete owned tenant info
+        if owned_tenant:
+            done_msg += "Start to delete owned tenant.\n"
+            tenant_id = owned_tenant[0]["tenant_id"]
+            kb_ids = KnowledgebaseService.get_kb_ids(usr.id)
+            # step1.1 delete knowledgebase related file and info
+            if kb_ids:
+                # step1.1.1 delete files in storage, remove bucket
+                for kb_id in kb_ids:
+                    if STORAGE_IMPL.bucket_exists(kb_id):
+                        STORAGE_IMPL.remove_bucket(kb_id)
+                done_msg += f"- Removed {len(kb_ids)} dataset's buckets.\n"
+                # step1.1.2 delete file and document info in db
+                doc_ids = DocumentService.get_all_doc_ids_by_kb_ids(kb_ids)
+                if doc_ids:
+                    doc_delete_res = DocumentService.delete_by_ids([i["id"] for i in doc_ids])
+                    done_msg += f"- Deleted {doc_delete_res} document records.\n"
+                    task_delete_res = TaskService.delete_by_doc_ids([i["id"] for i in doc_ids])
+                    done_msg += f"- Deleted {task_delete_res} task records.\n"
+                file_ids = FileService.get_all_file_ids_by_tenant_id(usr.id)
+                if file_ids:
+                    file_delete_res = FileService.delete_by_ids([f["id"] for f in file_ids])
+                    done_msg += f"- Deleted {file_delete_res} file records.\n"
+                if doc_ids or file_ids:
+                    file2doc_delete_res = File2DocumentService.delete_by_document_ids_or_file_ids(
+                        [i["id"] for i in doc_ids],
+                        [f["id"] for f in file_ids]
+                    )
+                    done_msg += f"- Deleted {file2doc_delete_res} document-file relation records.\n"
+                # step1.1.3 delete chunk in es
+                r = settings.docStoreConn.delete({"kb_id": kb_ids},
+                                         search.index_name(tenant_id), kb_ids)
+                done_msg += f"- Deleted {r} chunk records.\n"
+                kb_delete_res = KnowledgebaseService.delete_by_ids(kb_ids)
+                done_msg += f"- Deleted {kb_delete_res} knowledgebase records.\n"
+                # step1.1.4 delete agents
+                agent_delete_res = delete_user_agents(usr.id)
+                done_msg += f"- Deleted {agent_delete_res['agents_deleted_count']} agent, {agent_delete_res['version_deleted_count']} versions records.\n"
+                # step1.1.5 delete dialogs
+                dialog_delete_res = delete_user_dialogs(usr.id)
+                done_msg += f"- Deleted {dialog_delete_res['dialogs_deleted_count']} dialogs, {dialog_delete_res['conversations_deleted_count']} conversations, {dialog_delete_res['api_token_deleted_count']} api tokens, {dialog_delete_res['api4conversation_deleted_count']} api4conversations.\n"
+                # step1.1.6 delete mcp server
+                mcp_delete_res = MCPServerService.delete_by_tenant_id(usr.id)
+                done_msg += f"- Deleted {mcp_delete_res} MCP server.\n"
+                # step1.1.7 delete search
+                search_delete_res = SearchService.delete_by_tenant_id(usr.id)
+                done_msg += f"- Deleted {search_delete_res} search records.\n"
+            # step1.2 delete tenant_llm and tenant_langfuse
+            llm_delete_res = TenantLLMService.delete_by_tenant_id(tenant_id)
+            done_msg += f"- Deleted {llm_delete_res} tenant-LLM records.\n"
+            langfuse_delete_res = TenantLangfuseService.delete_ty_tenant_id(tenant_id)
+            done_msg += f"- Deleted {langfuse_delete_res} langfuse records.\n"
+            # step1.3 delete own tenant
+            tenant_delete_res = TenantService.delete_by_id(tenant_id)
+            done_msg += f"- Deleted {tenant_delete_res} tenant.\n"
+        # step2 delete user-tenant relation
+        if tenants:
+            # step2.1 delete docs and files in joined team
+            joined_tenants = [t for t in tenants if t["role"] == UserTenantRole.NORMAL.value]
+            if joined_tenants:
+                done_msg += "Start to delete data in joined tenants.\n"
+                created_documents = DocumentService.get_all_docs_by_creator_id(usr.id)
+                if created_documents:
+                    # step2.1.1 delete files
+                    doc_file_info = File2DocumentService.get_by_document_ids([d['id'] for d in created_documents])
+                    created_files = FileService.get_by_ids([f['file_id'] for f in doc_file_info])
+                    if created_files:
+                        # step2.1.1.1 delete file in storage
+                        for f in created_files:
+                            STORAGE_IMPL.rm(f.parent_id, f.location)
+                        done_msg += f"- Deleted {len(created_files)} uploaded file.\n"
+                        # step2.1.1.2 delete file record
+                        file_delete_res = FileService.delete_by_ids([f.id for f in created_files])
+                        done_msg += f"- Deleted {file_delete_res} file records.\n"
+                    # step2.1.2 delete document-file relation record
+                    file2doc_delete_res = File2DocumentService.delete_by_document_ids_or_file_ids(
+                        [d['id'] for d in created_documents],
+                        [f.id for f in created_files]
+                    )
+                    done_msg += f"- Deleted {file2doc_delete_res} document-file relation records.\n"
+                    # step2.1.3 delete chunks
+                    doc_groups = group_by(created_documents, "tenant_id")
+                    kb_grouped_doc = {k: group_by(v, "kb_id") for k, v in doc_groups.items()}
+                    # chunks in {'tenant_id': {'kb_id': [{'id': doc_id}]}} structure
+                    chunk_delete_res = 0
+                    kb_doc_info = {}
+                    for _tenant_id, kb_doc in kb_grouped_doc.items():
+                        for _kb_id, docs in kb_doc.items():
+                            chunk_delete_res += settings.docStoreConn.delete(
+                                {"doc_id": [d["id"] for d in docs]},
+                                search.index_name(_tenant_id), _kb_id
+                            )
+                            # record doc info
+                            if _kb_id in kb_doc_info.keys():
+                                kb_doc_info[_kb_id]['doc_num'] += 1
+                                kb_doc_info[_kb_id]['token_num'] += sum([d["token_num"] for d in docs])
+                                kb_doc_info[_kb_id]['chunk_num'] += sum([d["chunk_num"] for d in docs])
+                            else:
+                                kb_doc_info[_kb_id] = {
+                                    'doc_num': 1,
+                                    'token_num': sum([d["token_num"] for d in docs]),
+                                    'chunk_num': sum([d["chunk_num"] for d in docs])
+                                }
+                    done_msg += f"- Deleted {chunk_delete_res} chunks.\n"
+                    # step2.1.4 delete tasks
+                    task_delete_res = TaskService.delete_by_doc_ids([d['id'] for d in created_documents])
+                    done_msg += f"- Deleted {task_delete_res} tasks.\n"
+                    # step2.1.5 delete document record
+                    doc_delete_res = DocumentService.delete_by_ids([d['id'] for d in created_documents])
+                    done_msg += f"- Deleted {doc_delete_res} documents.\n"
+                    # step2.1.6 update knowledge base doc&chunk&token cnt
+                    for kb_id, doc_num in kb_doc_info.items():
+                        KnowledgebaseService.decrease_document_num_in_delete(kb_id, doc_num)
+
+            # step2.2 delete relation
+            user_tenant_delete_res = UserTenantService.delete_by_ids([t["id"] for t in tenants])
+            done_msg += f"- Deleted {user_tenant_delete_res} user-tenant records.\n"
+        # step3 finally delete user
+        user_delete_res = UserService.delete_by_id(usr.id)
+        done_msg += f"- Deleted {user_delete_res} user.\nDelete done!"
+
+        return {"success": True, "message": f"Successfully deleted user. Details:\n{done_msg}"}
+
+    except Exception as e:
+        logging.exception(e)
+        return {"success": False, "message": f"Error: {str(e)}. Already done:\n{done_msg}"}
+
+
+def delete_user_agents(user_id: str) -> dict:
+    """
+    use user_id to delete
+    :return: {
+        "agents_deleted_count": 1,
+        "version_deleted_count": 2
+    }
+    """
+    agents_deleted_count, agents_version_deleted_count = 0, 0
+    user_agents = UserCanvasService.get_all_agents_by_tenant_ids([user_id], user_id)
+    if user_agents:
+        agents_version = UserCanvasVersionService.get_all_canvas_version_by_canvas_ids([a['id'] for a in user_agents])
+        agents_version_deleted_count = UserCanvasVersionService.delete_by_ids([v['id'] for v in agents_version])
+        agents_deleted_count = UserCanvasService.delete_by_ids([a['id'] for a in user_agents])
+    return {
+        "agents_deleted_count": agents_deleted_count,
+        "version_deleted_count": agents_version_deleted_count
+    }
+
+
+def delete_user_dialogs(user_id: str) -> dict:
+    """
+    use user_id to delete
+    :return: {
+        "dialogs_deleted_count": 1,
+        "conversations_deleted_count": 1,
+        "api_token_deleted_count": 2,
+        "api4conversation_deleted_count": 2
+    }
+    """
+    dialog_deleted_count, conversations_deleted_count, api_token_deleted_count, api4conversation_deleted_count = 0, 0, 0, 0
+    user_dialogs = DialogService.get_all_dialogs_by_tenant_id(user_id)
+    if user_dialogs:
+        # delete conversation
+        conversations = ConversationService.get_all_conversation_by_dialog_ids([ud['id'] for ud in user_dialogs])
+        conversations_deleted_count = ConversationService.delete_by_ids([c['id'] for c in conversations])
+        # delete api token
+        api_token_deleted_count = APITokenService.delete_by_tenant_id(user_id)
+        # delete api for conversation
+        api4conversation_deleted_count = API4ConversationService.delete_by_dialog_ids([ud['id'] for ud in user_dialogs])
+        # delete dialog at last
+        dialog_deleted_count = DialogService.delete_by_ids([ud['id'] for ud in user_dialogs])
+    return {
+        "dialogs_deleted_count": dialog_deleted_count,
+        "conversations_deleted_count": conversations_deleted_count,
+        "api_token_deleted_count": api_token_deleted_count,
+        "api4conversation_deleted_count": api4conversation_deleted_count
+    }

api/db/services/__init__.py

@@ -19,7 +19,7 @@ from pathlib import PurePath
 from .user_service import UserService as UserService
 
 
-def split_name_counter(filename: str) -> tuple[str, int | None]:
+def _split_name_counter(filename: str) -> tuple[str, int | None]:
     """
     Splits a filename into main part and counter (if present in parentheses).
 
@@ -87,7 +87,7 @@ def duplicate_name(query_func, **kwargs) -> str:
         stem = path.stem
         suffix = path.suffix
 
-        main_part, counter = split_name_counter(stem)
+        main_part, counter = _split_name_counter(stem)
         counter = counter + 1 if counter else 1
 
         new_name = f"{main_part}({counter}){suffix}"

api/db/services/api_service.py

@@ -35,6 +35,11 @@ class APITokenService(CommonService):
             cls.model.token == token
         )
 
+    @classmethod
+    @DB.connection_context()
+    def delete_by_tenant_id(cls, tenant_id):
+        return cls.model.delete().where(cls.model.tenant_id == tenant_id).execute()
+
 
 class API4ConversationService(CommonService):
     model = API4Conversation
@@ -100,3 +105,8 @@ class API4ConversationService(CommonService):
             cls.model.create_date <= to_date,
             cls.model.source == source
         ).group_by(cls.model.create_date.truncate("day")).dicts()
+
+    @classmethod
+    @DB.connection_context()
+    def delete_by_dialog_ids(cls, dialog_ids):
+        return cls.model.delete().where(cls.model.dialog_id.in_(dialog_ids)).execute()

api/db/services/canvas_service.py

@@ -63,7 +63,38 @@ class UserCanvasService(CommonService):
 
     @classmethod
     @DB.connection_context()
-    def get_by_tenant_id(cls, pid):
+    def get_all_agents_by_tenant_ids(cls, tenant_ids, user_id):
+        # will get all permitted agents, be cautious
+        fields = [
+            cls.model.id,
+            cls.model.title,
+            cls.model.permission,
+            cls.model.canvas_type,
+            cls.model.canvas_category
+        ]
+        # find team agents and owned agents
+        agents = cls.model.select(*fields).where(
+            (cls.model.user_id.in_(tenant_ids) & (cls.model.permission == TenantPermission.TEAM.value)) | (
+                cls.model.user_id == user_id
+            )
+        )
+        # sort by create_time, asc
+        agents.order_by(cls.model.create_time.asc())
+        # maybe cause slow query by deep paginate, optimize later
+        offset, limit = 0, 50
+        res = []
+        while True:
+            ag_batch = agents.offset(offset).limit(limit)
+            _temp = list(ag_batch.dicts())
+            if not _temp:
+                break
+            res.extend(_temp)
+            offset += limit
+        return res
+
+    @classmethod
+    @DB.connection_context()
+    def get_by_canvas_id(cls, pid):
         try:
 
             fields = [
@@ -95,7 +126,7 @@ class UserCanvasService(CommonService):
     @DB.connection_context()
     def get_by_tenant_ids(cls, joined_tenant_ids, user_id,
                           page_number, items_per_page,
-                          orderby, desc, keywords, canvas_category=CanvasCategory.Agent,
+                          orderby, desc, keywords, canvas_category=None
                           ):
         fields = [
             cls.model.id,
@@ -104,6 +135,7 @@ 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,
@@ -111,31 +143,33 @@ 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))
             )
-        agents = agents.where(cls.model.canvas_category == canvas_category)
+        if canvas_category:
+            agents = agents.where(cls.model.canvas_category == canvas_category)
         if desc:
             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
     @DB.connection_context()
     def accessible(cls, canvas_id, tenant_id):
         from api.db.services.user_service import UserTenantService
-        e, c = UserCanvasService.get_by_tenant_id(canvas_id)
+        e, c = UserCanvasService.get_by_canvas_id(canvas_id)
         if not e:
             return False
 

api/db/services/common_service.py

@@ -14,12 +14,24 @@
 #  limitations under the License.
 #
 from datetime import datetime
-
+from tenacity import retry, stop_after_attempt, wait_exponential, retry_if_exception_type
 import peewee
+from peewee import InterfaceError, OperationalError
 
 from api.db.db_models import DB
 from api.utils import current_timestamp, datetime_format, get_uuid
 
+def retry_db_operation(func):
+    @retry(
+        stop=stop_after_attempt(3),
+        wait=wait_exponential(multiplier=1, min=1, max=5),
+        retry=retry_if_exception_type((InterfaceError, OperationalError)),
+        before_sleep=lambda retry_state: print(f"RETRY {retry_state.attempt_number} TIMES"),
+        reraise=True,
+    )
+    def wrapper(*args, **kwargs):
+        return func(*args, **kwargs)
+    return wrapper
 
 class CommonService:
     """Base service class that provides common database operations.
@@ -202,6 +214,7 @@ class CommonService:
 
     @classmethod
     @DB.connection_context()
+    @retry_db_operation
     def update_by_id(cls, pid, data):
         # Update a single record by ID
         # Args:

api/db/services/conversation_service.py

@@ -23,7 +23,7 @@ from api.db.services.dialog_service import DialogService, chat
 from api.utils import get_uuid
 import json
 
-from rag.prompts import chunks_format
+from rag.prompts.generator import chunks_format
 
 
 class ConversationService(CommonService):
@@ -48,6 +48,21 @@ class ConversationService(CommonService):
 
         return list(sessions.dicts())
 
+    @classmethod
+    @DB.connection_context()
+    def get_all_conversation_by_dialog_ids(cls, dialog_ids):
+        sessions = cls.model.select().where(cls.model.dialog_id.in_(dialog_ids))
+        sessions.order_by(cls.model.create_time.asc())
+        offset, limit = 0, 100
+        res = []
+        while True:
+            s_batch = sessions.offset(offset).limit(limit)
+            _temp = list(s_batch.dicts())
+            if not _temp:
+                break
+            res.extend(_temp)
+            offset += limit
+        return res
 
 def structure_answer(conv, ans, message_id, session_id):
     reference = ans["reference"]

api/db/services/dialog_service.py

@@ -39,8 +39,8 @@ from graphrag.general.mind_map_extractor import MindMapExtractor
 from rag.app.resume import forbidden_select_fields4resume
 from rag.app.tag import label_question
 from rag.nlp.search import index_name
-from rag.prompts import chunks_format, citation_prompt, cross_languages, full_question, kb_prompt, keyword_extraction, message_fit_in
-from rag.prompts.prompts import gen_meta_filter, PROMPT_JINJA_ENV, ASK_SUMMARY
+from rag.prompts.generator import chunks_format, citation_prompt, cross_languages, full_question, kb_prompt, keyword_extraction, message_fit_in, \
+    gen_meta_filter, PROMPT_JINJA_ENV, ASK_SUMMARY
 from rag.utils import num_tokens_from_string, rmSpace
 from rag.utils.tavily_conn import Tavily
 
@@ -159,6 +159,22 @@ class DialogService(CommonService):
 
         return list(dialogs.dicts()), count
 
+    @classmethod
+    @DB.connection_context()
+    def get_all_dialogs_by_tenant_id(cls, tenant_id):
+        fields = [cls.model.id]
+        dialogs = cls.model.select(*fields).where(cls.model.tenant_id == tenant_id)
+        dialogs.order_by(cls.model.create_time.asc())
+        offset, limit = 0, 100
+        res = []
+        while True:
+            d_batch = dialogs.offset(offset).limit(limit)
+            _temp = list(d_batch.dicts())
+            if not _temp:
+                break
+            res.extend(_temp)
+            offset += limit
+        return res
 
 def chat_solo(dialog, messages, stream=True):
     if TenantLLMService.llm_id2llm_type(dialog.llm_id) == "image2text":
@@ -176,7 +192,7 @@ def chat_solo(dialog, messages, stream=True):
         delta_ans = ""
         for ans in chat_mdl.chat_streamly(prompt_config.get("system", ""), msg, dialog.llm_setting):
             answer = ans
-            delta_ans = ans[len(last_ans) :]
+            delta_ans = ans[len(last_ans):]
             if num_tokens_from_string(delta_ans) < 16:
                 continue
             last_ans = answer
@@ -261,13 +277,13 @@ def convert_conditions(metadata_condition):
         "not is": "≠"
     }
     return [
-    {
-        "op": op_mapping.get(cond["comparison_operator"], cond["comparison_operator"]),
-        "key": cond["name"],
-        "value": cond["value"]
-    }
-    for cond in metadata_condition.get("conditions", [])
-]
+        {
+            "op": op_mapping.get(cond["comparison_operator"], cond["comparison_operator"]),
+            "key": cond["name"],
+            "value": cond["value"]
+        }
+        for cond in metadata_condition.get("conditions", [])
+    ]
 
 
 def meta_filter(metas: dict, filters: list[dict]):
@@ -284,19 +300,19 @@ def meta_filter(metas: dict, filters: list[dict]):
                 value = str(value)
 
             for conds in [
-                    (operator == "contains", str(value).lower() in str(input).lower()),
-                    (operator == "not contains", str(value).lower() not in str(input).lower()),
-                    (operator == "start with", str(input).lower().startswith(str(value).lower())),
-                    (operator == "end with", str(input).lower().endswith(str(value).lower())),
-                    (operator == "empty", not input),
-                    (operator == "not empty", input),
-                    (operator == "=", input == value),
-                    (operator == "≠", input != value),
-                    (operator == ">", input > value),
-                    (operator == "<", input < value),
-                    (operator == "≥", input >= value),
-                    (operator == "≤", input <= value),
-                ]:
+                (operator == "contains", str(value).lower() in str(input).lower()),
+                (operator == "not contains", str(value).lower() not in str(input).lower()),
+                (operator == "start with", str(input).lower().startswith(str(value).lower())),
+                (operator == "end with", str(input).lower().endswith(str(value).lower())),
+                (operator == "empty", not input),
+                (operator == "not empty", input),
+                (operator == "=", input == value),
+                (operator == "≠", input != value),
+                (operator == ">", input > value),
+                (operator == "<", input < value),
+                (operator == "≥", input >= value),
+                (operator == "≤", input <= value),
+            ]:
                 try:
                     if all(conds):
                         ids.extend(docids)
@@ -456,7 +472,8 @@ def chat(dialog, messages, stream=True, **kwargs):
                 kbinfos["chunks"].extend(tav_res["chunks"])
                 kbinfos["doc_aggs"].extend(tav_res["doc_aggs"])
             if prompt_config.get("use_kg"):
-                ck = settings.kg_retrievaler.retrieval(" ".join(questions), tenant_ids, dialog.kb_ids, embd_mdl, LLMBundle(dialog.tenant_id, LLMType.CHAT))
+                ck = settings.kg_retrievaler.retrieval(" ".join(questions), tenant_ids, dialog.kb_ids, embd_mdl,
+                                                       LLMBundle(dialog.tenant_id, LLMType.CHAT))
                 if ck["content_with_weight"]:
                     kbinfos["chunks"].insert(0, ck)
 
@@ -467,7 +484,8 @@ def chat(dialog, messages, stream=True, **kwargs):
     retrieval_ts = timer()
     if not knowledges and prompt_config.get("empty_response"):
         empty_res = prompt_config["empty_response"]
-        yield {"answer": empty_res, "reference": kbinfos, "prompt": "\n\n### Query:\n%s" % " ".join(questions), "audio_binary": tts(tts_mdl, empty_res)}
+        yield {"answer": empty_res, "reference": kbinfos, "prompt": "\n\n### Query:\n%s" % " ".join(questions),
+               "audio_binary": tts(tts_mdl, empty_res)}
         return {"answer": prompt_config["empty_response"], "reference": kbinfos}
 
     kwargs["knowledge"] = "\n------\n" + "\n\n------\n\n".join(knowledges)
@@ -565,7 +583,8 @@ def chat(dialog, messages, stream=True, **kwargs):
 
     if langfuse_tracer:
         langfuse_generation = langfuse_tracer.start_generation(
-            trace_context=trace_context, name="chat", model=llm_model_config["llm_name"], input={"prompt": prompt, "prompt4citation": prompt4citation, "messages": msg}
+            trace_context=trace_context, name="chat", model=llm_model_config["llm_name"],
+            input={"prompt": prompt, "prompt4citation": prompt4citation, "messages": msg}
         )
 
     if stream:
@@ -575,12 +594,12 @@ def chat(dialog, messages, stream=True, **kwargs):
             if thought:
                 ans = re.sub(r"^.*</think>", "", ans, flags=re.DOTALL)
             answer = ans
-            delta_ans = ans[len(last_ans) :]
+            delta_ans = ans[len(last_ans):]
             if num_tokens_from_string(delta_ans) < 16:
                 continue
             last_ans = answer
             yield {"answer": thought + answer, "reference": {}, "audio_binary": tts(tts_mdl, delta_ans)}
-        delta_ans = answer[len(last_ans) :]
+        delta_ans = answer[len(last_ans):]
         if delta_ans:
             yield {"answer": thought + answer, "reference": {}, "audio_binary": tts(tts_mdl, delta_ans)}
         yield decorate_answer(thought + answer)
@@ -676,7 +695,9 @@ Please write the SQL, only SQL, without any other explanations or text.
 
     # compose Markdown table
     columns = (
-        "|" + "|".join([re.sub(r"(/.*|（[^（）]+）)", "", field_map.get(tbl["columns"][i]["name"], tbl["columns"][i]["name"])) for i in column_idx]) + ("|Source|" if docid_idx and docid_idx else "|")
+            "|" + "|".join(
+        [re.sub(r"(/.*|（[^（）]+）)", "", field_map.get(tbl["columns"][i]["name"], tbl["columns"][i]["name"])) for i in column_idx]) + (
+                "|Source|" if docid_idx and docid_idx else "|")
     )
 
     line = "|" + "|".join(["------" for _ in range(len(column_idx))]) + ("|------|" if docid_idx and docid_idx else "")
@@ -753,7 +774,7 @@ def ask(question, kb_ids, tenant_id, chat_llm_name=None, search_config={}):
                 doc_ids = None
 
     kbinfos = retriever.retrieval(
-        question = question,
+        question=question,
         embd_mdl=embd_mdl,
         tenant_ids=tenant_ids,
         kb_ids=kb_ids,
@@ -775,7 +796,8 @@ def ask(question, kb_ids, tenant_id, chat_llm_name=None, search_config={}):
 
     def decorate_answer(answer):
         nonlocal knowledges, kbinfos, sys_prompt
-        answer, idx = retriever.insert_citations(answer, [ck["content_ltks"] for ck in kbinfos["chunks"]], [ck["vector"] for ck in kbinfos["chunks"]], embd_mdl, tkweight=0.7, vtweight=0.3)
+        answer, idx = retriever.insert_citations(answer, [ck["content_ltks"] for ck in kbinfos["chunks"]], [ck["vector"] for ck in kbinfos["chunks"]],
+                                                 embd_mdl, tkweight=0.7, vtweight=0.3)
         idx = set([kbinfos["chunks"][int(i)]["doc_id"] for i in idx])
         recall_docs = [d for d in kbinfos["doc_aggs"] if d["doc_id"] in idx]
         if not recall_docs:

api/db/services/document_service.py

@@ -24,12 +24,13 @@ from io import BytesIO
 
 import trio
 import xxhash
-from peewee import fn, Case
+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
-from api.db.db_models import DB, Document, Knowledgebase, Task, Tenant, UserTenant, File2Document, File
+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, \
+    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
@@ -51,6 +52,7 @@ class DocumentService(CommonService):
             cls.model.thumbnail,
             cls.model.kb_id,
             cls.model.parser_id,
+            cls.model.pipeline_id,
             cls.model.parser_config,
             cls.model.source_type,
             cls.model.type,
@@ -79,7 +81,10 @@ class DocumentService(CommonService):
     def get_list(cls, kb_id, page_number, items_per_page,
                  orderby, desc, keywords, id, name):
         fields = cls.get_cls_model_fields()
-        docs = cls.model.select(*fields).join(File2Document, on = (File2Document.document_id == cls.model.id)).join(File, on = (File.id == File2Document.file_id)).where(cls.model.kb_id == kb_id)
+        docs = cls.model.select(*[*fields, UserCanvas.title]).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) & (UserCanvas.canvas_category == CanvasCategory.DataFlow.value)), join_type=JOIN.LEFT_OUTER)\
+            .where(cls.model.kb_id == kb_id)
         if id:
             docs = docs.where(
                 cls.model.id == id)
@@ -117,12 +122,22 @@ class DocumentService(CommonService):
                      orderby, desc, keywords, run_status, types, suffix):
         fields = cls.get_cls_model_fields()
         if keywords:
-            docs = cls.model.select(*fields).join(File2Document, on=(File2Document.document_id == cls.model.id)).join(File, on=(File.id == File2Document.file_id)).where(
-                (cls.model.kb_id == kb_id),
-                (fn.LOWER(cls.model.name).contains(keywords.lower()))
-            )
+            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).join(File2Document, on=(File2Document.document_id == cls.model.id)).join(File, on=(File.id == File2Document.file_id)).where(cls.model.kb_id == kb_id)
+            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:
             docs = docs.where(cls.model.run.in_(run_status))
@@ -228,6 +243,46 @@ class DocumentService(CommonService):
 
         return int(query.scalar()) or 0
 
+    @classmethod
+    @DB.connection_context()
+    def get_all_doc_ids_by_kb_ids(cls, kb_ids):
+        fields = [cls.model.id]
+        docs = cls.model.select(*fields).where(cls.model.kb_id.in_(kb_ids))
+        docs.order_by(cls.model.create_time.asc())
+        # maybe cause slow query by deep paginate, optimize later
+        offset, limit = 0, 100
+        res = []
+        while True:
+            doc_batch = docs.offset(offset).limit(limit)
+            _temp = list(doc_batch.dicts())
+            if not _temp:
+                break
+            res.extend(_temp)
+            offset += limit
+        return res
+
+    @classmethod
+    @DB.connection_context()
+    def get_all_docs_by_creator_id(cls, creator_id):
+        fields = [
+            cls.model.id, cls.model.kb_id, cls.model.token_num, cls.model.chunk_num, Knowledgebase.tenant_id
+        ]
+        docs = cls.model.select(*fields).join(Knowledgebase, on=(Knowledgebase.id == cls.model.kb_id)).where(
+            cls.model.created_by == creator_id
+        )
+        docs.order_by(cls.model.create_time.asc())
+        # maybe cause slow query by deep paginate, optimize later
+        offset, limit = 0, 100
+        res = []
+        while True:
+            doc_batch = docs.offset(offset).limit(limit)
+            _temp = list(doc_batch.dicts())
+            if not _temp:
+                break
+            res.extend(_temp)
+            offset += limit
+        return res
+
     @classmethod
     @DB.connection_context()
     def insert(cls, doc):
@@ -330,8 +385,7 @@ class DocumentService(CommonService):
                                process_duration=cls.model.process_duration + duration).where(
             cls.model.id == doc_id).execute()
         if num == 0:
-            raise LookupError(
-                "Document not found which is supposed to be there")
+            logging.warning("Document not found which is supposed to be there")
         num = Knowledgebase.update(
             token_num=Knowledgebase.token_num +
                       token_num,
@@ -597,6 +651,22 @@ class DocumentService(CommonService):
     @DB.connection_context()
     def update_progress(cls):
         docs = cls.get_unfinished_docs()
+
+        cls._sync_progress(docs)
+
+
+    @classmethod
+    @DB.connection_context()
+    def update_progress_immediately(cls, docs:list[dict]):
+        if not docs:
+            return
+
+        cls._sync_progress(docs)
+
+
+    @classmethod
+    @DB.connection_context()
+    def _sync_progress(cls, docs:list[dict]):
         for d in docs:
             try:
                 tsks = Task.query(doc_id=d["id"], order_by=Task.create_time)
@@ -606,8 +676,6 @@ class DocumentService(CommonService):
                 prg = 0
                 finished = True
                 bad = 0
-                has_raptor = False
-                has_graphrag = False
                 e, doc = DocumentService.get_by_id(d["id"])
                 status = doc.run  # TaskStatus.RUNNING.value
                 priority = 0
@@ -619,24 +687,14 @@ class DocumentService(CommonService):
                     prg += t.progress if t.progress >= 0 else 0
                     if t.progress_msg.strip():
                         msg.append(t.progress_msg)
-                    if t.task_type == "raptor":
-                        has_raptor = True
-                    elif t.task_type == "graphrag":
-                        has_graphrag = True
                     priority = max(priority, t.priority)
                 prg /= len(tsks)
                 if finished and bad:
                     prg = -1
                     status = TaskStatus.FAIL.value
                 elif finished:
-                    if (d["parser_config"].get("raptor") or {}).get("use_raptor") and not has_raptor:
-                        queue_raptor_o_graphrag_tasks(d, "raptor", priority)
-                        prg = 0.98 * len(tsks) / (len(tsks) + 1)
-                    elif (d["parser_config"].get("graphrag") or {}).get("use_graphrag") and not has_graphrag:
-                        queue_raptor_o_graphrag_tasks(d, "graphrag", priority)
-                        prg = 0.98 * len(tsks) / (len(tsks) + 1)
-                    else:
-                        status = TaskStatus.DONE.value
+                    prg = 1
+                    status = TaskStatus.DONE.value
 
                 msg = "\n".join(sorted(msg))
                 info = {
@@ -648,7 +706,7 @@ class DocumentService(CommonService):
                     info["progress"] = prg
                 if msg:
                     info["progress_msg"] = msg
-                    if msg.endswith("created task graphrag") or msg.endswith("created task raptor"):
+                    if msg.endswith("created task graphrag") or msg.endswith("created task raptor") or msg.endswith("created task mindmap"):
                         info["progress_msg"] += "\n%d tasks are ahead in the queue..."%get_queue_length(priority)
                 else:
                     info["progress_msg"] = "%d tasks are ahead in the queue..."%get_queue_length(priority)
@@ -729,7 +787,11 @@ class DocumentService(CommonService):
             "cancelled": int(cancelled),
         }
 
-def queue_raptor_o_graphrag_tasks(doc, ty, priority):
+def queue_raptor_o_graphrag_tasks(doc, ty, priority, fake_doc_id="", doc_ids=[]):
+    """
+    You can provide a fake_doc_id to bypass the restriction of tasks at the knowledgebase level.
+    Optionally, specify a list of doc_ids to determine which documents participate in the task.
+    """
     chunking_config = DocumentService.get_chunking_config(doc["id"])
     hasher = xxhash.xxh64()
     for field in sorted(chunking_config.keys()):
@@ -739,11 +801,12 @@ def queue_raptor_o_graphrag_tasks(doc, ty, priority):
         nonlocal doc
         return {
             "id": get_uuid(),
-            "doc_id": doc["id"],
+            "doc_id": fake_doc_id if fake_doc_id else doc["id"],
             "from_page": 100000000,
             "to_page": 100000000,
             "task_type": ty,
-            "progress_msg":  datetime.now().strftime("%H:%M:%S") + " created task " + ty
+            "progress_msg":  datetime.now().strftime("%H:%M:%S") + " created task " + ty,
+            "begin_at": datetime.now(),
         }
 
     task = new_task()
@@ -752,7 +815,12 @@ def queue_raptor_o_graphrag_tasks(doc, ty, priority):
     hasher.update(ty.encode("utf-8"))
     task["digest"] = hasher.hexdigest()
     bulk_insert_into_db(Task, [task], True)
+
+    if ty in ["graphrag", "raptor", "mindmap"]:
+        task["doc_ids"] = doc_ids
+        DocumentService.begin2parse(doc["id"])
     assert REDIS_CONN.queue_product(get_svr_queue_name(priority), message=task), "Can't access Redis. Please check the Redis' status."
+    return task["id"]
 
 
 def get_queue_length(priority):

api/db/services/file2document_service.py

@@ -38,6 +38,12 @@ class File2DocumentService(CommonService):
         objs = cls.model.select().where(cls.model.document_id == document_id)
         return objs
 
+    @classmethod
+    @DB.connection_context()
+    def get_by_document_ids(cls, document_ids):
+        objs = cls.model.select().where(cls.model.document_id.in_(document_ids))
+        return list(objs.dicts())
+
     @classmethod
     @DB.connection_context()
     def insert(cls, obj):
@@ -50,6 +56,15 @@ class File2DocumentService(CommonService):
     def delete_by_file_id(cls, file_id):
         return cls.model.delete().where(cls.model.file_id == file_id).execute()
 
+    @classmethod
+    @DB.connection_context()
+    def delete_by_document_ids_or_file_ids(cls, document_ids, file_ids):
+        if not document_ids:
+            return cls.model.delete().where(cls.model.file_id.in_(file_ids)).execute()
+        elif not file_ids:
+            return cls.model.delete().where(cls.model.document_id.in_(document_ids)).execute()
+        return cls.model.delete().where(cls.model.document_id.in_(document_ids) | cls.model.file_id.in_(file_ids)).execute()
+
     @classmethod
     @DB.connection_context()
     def delete_by_document_id(cls, doc_id):

api/db/services/file_service.py

@@ -161,6 +161,23 @@ class FileService(CommonService):
             result_ids.append(folder_id)
         return result_ids
 
+    @classmethod
+    @DB.connection_context()
+    def get_all_file_ids_by_tenant_id(cls, tenant_id):
+        fields = [cls.model.id]
+        files = cls.model.select(*fields).where(cls.model.tenant_id == tenant_id)
+        files.order_by(cls.model.create_time.asc())
+        offset, limit = 0, 100
+        res = []
+        while True:
+            file_batch = files.offset(offset).limit(limit)
+            _temp = list(file_batch.dicts())
+            if not _temp:
+                break
+            res.extend(_temp)
+            offset += limit
+        return res
+
     @classmethod
     @DB.connection_context()
     def create_folder(cls, file, parent_id, name, count):
@@ -440,6 +457,7 @@ class FileService(CommonService):
                     "id": doc_id,
                     "kb_id": kb.id,
                     "parser_id": self.get_parser(filetype, filename, kb.parser_id),
+                    "pipeline_id": kb.pipeline_id,
                     "parser_config": kb.parser_config,
                     "created_by": user_id,
                     "type": filetype,
@@ -495,7 +513,7 @@ class FileService(CommonService):
             return ParserType.AUDIO.value
         if re.search(r"\.(ppt|pptx|pages)$", filename):
             return ParserType.PRESENTATION.value
-        if re.search(r"\.(eml)$", filename):
+        if re.search(r"\.(msg|eml)$", filename):
             return ParserType.EMAIL.value
         return default
 

api/db/services/knowledgebase_service.py

@@ -15,10 +15,10 @@
 #
 from datetime import datetime
 
-from peewee import fn
+from peewee import fn, JOIN
 
 from api.db import StatusEnum, TenantPermission
-from api.db.db_models import DB, Document, Knowledgebase, Tenant, User, UserTenant
+from api.db.db_models import DB, Document, Knowledgebase, User, UserTenant, UserCanvas
 from api.db.services.common_service import CommonService
 from api.utils import current_timestamp, datetime_format
 
@@ -190,6 +190,41 @@ class KnowledgebaseService(CommonService):
 
         return list(kbs.dicts()), count
 
+    @classmethod
+    @DB.connection_context()
+    def get_all_kb_by_tenant_ids(cls, tenant_ids, user_id):
+        # will get all permitted kb, be cautious.
+        fields = [
+            cls.model.name,
+            cls.model.language,
+            cls.model.permission,
+            cls.model.doc_num,
+            cls.model.token_num,
+            cls.model.chunk_num,
+            cls.model.status,
+            cls.model.create_date,
+            cls.model.update_date
+        ]
+        # find team kb and owned kb
+        kbs = cls.model.select(*fields).where(
+            (cls.model.tenant_id.in_(tenant_ids) & (cls.model.permission ==TenantPermission.TEAM.value)) | (
+                cls.model.tenant_id == user_id
+            )
+        )
+        # sort by create_time asc
+        kbs.order_by(cls.model.create_time.asc())
+        # maybe cause slow query by deep paginate, optimize later.
+        offset, limit = 0, 50
+        res = []
+        while True:
+            kb_batch = kbs.offset(offset).limit(limit)
+            _temp = list(kb_batch.dicts())
+            if not _temp:
+                break
+            res.extend(_temp)
+            offset += limit
+        return res
+
     @classmethod
     @DB.connection_context()
     def get_kb_ids(cls, tenant_id):
@@ -225,20 +260,29 @@ class KnowledgebaseService(CommonService):
             cls.model.token_num,
             cls.model.chunk_num,
             cls.model.parser_id,
+            cls.model.pipeline_id,
+            UserCanvas.title.alias("pipeline_name"),
+            UserCanvas.avatar.alias("pipeline_avatar"),
             cls.model.parser_config,
             cls.model.pagerank,
+            cls.model.graphrag_task_id,
+            cls.model.graphrag_task_finish_at,
+            cls.model.raptor_task_id,
+            cls.model.raptor_task_finish_at,
+            cls.model.mindmap_task_id,
+            cls.model.mindmap_task_finish_at,
             cls.model.create_time,
             cls.model.update_time
             ]
-        kbs = cls.model.select(*fields).join(Tenant, on=(
-            (Tenant.id == cls.model.tenant_id) & (Tenant.status == StatusEnum.VALID.value))).where(
+        kbs = cls.model.select(*fields)\
+                .join(UserCanvas, on=(cls.model.pipeline_id == UserCanvas.id), join_type=JOIN.LEFT_OUTER)\
+            .where(
             (cls.model.id == kb_id),
             (cls.model.status == StatusEnum.VALID.value)
-        )
+        ).dicts()
         if not kbs:
             return
-        d = kbs[0].to_dict()
-        return d
+        return kbs[0]
 
     @classmethod
     @DB.connection_context()
@@ -436,3 +480,17 @@ class KnowledgebaseService(CommonService):
             else:
                 raise e
 
+    @classmethod
+    @DB.connection_context()
+    def decrease_document_num_in_delete(cls, kb_id, doc_num_info: dict):
+        kb_row = cls.model.get_by_id(kb_id)
+        if not kb_row:
+            raise RuntimeError(f"kb_id {kb_id} does not exist")
+        update_dict = {
+            'doc_num': kb_row.doc_num - doc_num_info['doc_num'],
+            'chunk_num': kb_row.chunk_num - doc_num_info['chunk_num'],
+            'token_num': kb_row.token_num - doc_num_info['token_num'],
+            'update_time': current_timestamp(),
+            'update_date': datetime_format(datetime.now())
+        }
+        return cls.model.update(update_dict).where(cls.model.id == kb_id).execute()

api/db/services/langfuse_service.py

@@ -51,6 +51,11 @@ class TenantLangfuseService(CommonService):
         except peewee.DoesNotExist:
             return None
 
+    @classmethod
+    @DB.connection_context()
+    def delete_ty_tenant_id(cls, tenant_id):
+        return cls.model.delete().where(cls.model.tenant_id == tenant_id).execute()
+
     @classmethod
     def update_by_tenant(cls, tenant_id, langfuse_keys):
         langfuse_keys["update_time"] = current_timestamp()

api/db/services/mcp_server_service.py

@@ -84,3 +84,8 @@ class MCPServerService(CommonService):
             return bool(mcp_server), mcp_server
         except Exception:
             return False, None
+
+    @classmethod
+    @DB.connection_context()
+    def delete_by_tenant_id(cls, tenant_id: str):
+        return cls.model.delete().where(cls.model.tenant_id == tenant_id).execute()

api/db/services/pipeline_operation_log_service.py

@@ -0,0 +1,263 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+import json
+import logging
+import os
+from datetime import datetime, timedelta
+
+from peewee import fn
+
+from api.db import VALID_PIPELINE_TASK_TYPES, PipelineTaskType
+from api.db.db_models import DB, Document, PipelineOperationLog
+from api.db.services.canvas_service import UserCanvasService
+from api.db.services.common_service import CommonService
+from api.db.services.document_service import DocumentService
+from api.db.services.knowledgebase_service import KnowledgebaseService
+from api.db.services.task_service import GRAPH_RAPTOR_FAKE_DOC_ID
+from api.utils import current_timestamp, datetime_format, get_uuid
+
+
+class PipelineOperationLogService(CommonService):
+    model = PipelineOperationLog
+
+    @classmethod
+    def get_file_logs_fields(cls):
+        return [
+            cls.model.id,
+            cls.model.document_id,
+            cls.model.tenant_id,
+            cls.model.kb_id,
+            cls.model.pipeline_id,
+            cls.model.pipeline_title,
+            cls.model.parser_id,
+            cls.model.document_name,
+            cls.model.document_suffix,
+            cls.model.document_type,
+            cls.model.source_from,
+            cls.model.progress,
+            cls.model.progress_msg,
+            cls.model.process_begin_at,
+            cls.model.process_duration,
+            cls.model.dsl,
+            cls.model.task_type,
+            cls.model.operation_status,
+            cls.model.avatar,
+            cls.model.status,
+            cls.model.create_time,
+            cls.model.create_date,
+            cls.model.update_time,
+            cls.model.update_date,
+        ]
+
+    @classmethod
+    def get_dataset_logs_fields(cls):
+        return [
+            cls.model.id,
+            cls.model.tenant_id,
+            cls.model.kb_id,
+            cls.model.progress,
+            cls.model.progress_msg,
+            cls.model.process_begin_at,
+            cls.model.process_duration,
+            cls.model.task_type,
+            cls.model.operation_status,
+            cls.model.avatar,
+            cls.model.status,
+            cls.model.create_time,
+            cls.model.create_date,
+            cls.model.update_time,
+            cls.model.update_date,
+        ]
+
+    @classmethod
+    def save(cls, **kwargs):
+        """
+        wrap this function in a transaction
+        """
+        sample_obj = cls.model(**kwargs).save(force_insert=True)
+        return sample_obj
+
+    @classmethod
+    @DB.connection_context()
+    def create(cls, document_id, pipeline_id, task_type, fake_document_ids=[], dsl: str = "{}"):
+        referred_document_id = document_id
+
+        if referred_document_id == GRAPH_RAPTOR_FAKE_DOC_ID and fake_document_ids:
+            referred_document_id = fake_document_ids[0]
+        ok, document = DocumentService.get_by_id(referred_document_id)
+        if not ok:
+            logging.warning(f"Document for referred_document_id {referred_document_id} not found")
+            return
+        DocumentService.update_progress_immediately([document.to_dict()])
+        ok, document = DocumentService.get_by_id(referred_document_id)
+        if not ok:
+            logging.warning(f"Document for referred_document_id {referred_document_id} not found")
+            return
+        if document.progress not in [1, -1]:
+            return
+        operation_status = document.run
+
+        if pipeline_id:
+            ok, user_pipeline = UserCanvasService.get_by_id(pipeline_id)
+            if not ok:
+                raise RuntimeError(f"Pipeline {pipeline_id} not found")
+            tenant_id = user_pipeline.user_id
+            title = user_pipeline.title
+            avatar = user_pipeline.avatar
+        else:
+            ok, kb_info = KnowledgebaseService.get_by_id(document.kb_id)
+            if not ok:
+                raise RuntimeError(f"Cannot find knowledge base {document.kb_id} for referred_document {referred_document_id}")
+
+            tenant_id = kb_info.tenant_id
+            title = document.parser_id
+            avatar = document.thumbnail
+
+        if task_type not in VALID_PIPELINE_TASK_TYPES:
+            raise ValueError(f"Invalid task type: {task_type}")
+
+        if task_type in [PipelineTaskType.GRAPH_RAG, PipelineTaskType.RAPTOR, PipelineTaskType.MINDMAP]:
+            finish_at = document.process_begin_at + timedelta(seconds=document.process_duration)
+            if task_type == PipelineTaskType.GRAPH_RAG:
+                KnowledgebaseService.update_by_id(
+                    document.kb_id,
+                    {"graphrag_task_finish_at": finish_at},
+                )
+            elif task_type == PipelineTaskType.RAPTOR:
+                KnowledgebaseService.update_by_id(
+                    document.kb_id,
+                    {"raptor_task_finish_at": finish_at},
+                )
+            elif task_type == PipelineTaskType.MINDMAP:
+                KnowledgebaseService.update_by_id(
+                    document.kb_id,
+                    {"mindmap_task_finish_at": finish_at},
+                )
+
+        log = dict(
+            id=get_uuid(),
+            document_id=document_id,  # GRAPH_RAPTOR_FAKE_DOC_ID or real document_id
+            tenant_id=tenant_id,
+            kb_id=document.kb_id,
+            pipeline_id=pipeline_id,
+            pipeline_title=title,
+            parser_id=document.parser_id,
+            document_name=document.name,
+            document_suffix=document.suffix,
+            document_type=document.type,
+            source_from="",  # TODO: add in the future
+            progress=document.progress,
+            progress_msg=document.progress_msg,
+            process_begin_at=document.process_begin_at,
+            process_duration=document.process_duration,
+            dsl=json.loads(dsl),
+            task_type=task_type,
+            operation_status=operation_status,
+            avatar=avatar,
+        )
+        log["create_time"] = current_timestamp()
+        log["create_date"] = datetime_format(datetime.now())
+        log["update_time"] = current_timestamp()
+        log["update_date"] = datetime_format(datetime.now())
+
+        with DB.atomic():
+            obj = cls.save(**log)
+
+            limit = int(os.getenv("PIPELINE_OPERATION_LOG_LIMIT", 1000))
+            total = cls.model.select().where(cls.model.kb_id == document.kb_id).count()
+
+            if total > limit:
+                keep_ids = [m.id for m in cls.model.select(cls.model.id).where(cls.model.kb_id == document.kb_id).order_by(cls.model.create_time.desc()).limit(limit)]
+
+                deleted = cls.model.delete().where(cls.model.kb_id == document.kb_id, cls.model.id.not_in(keep_ids)).execute()
+                logging.info(f"[PipelineOperationLogService] Cleaned {deleted} old logs, kept latest {limit} for {document.kb_id}")
+
+        return obj
+
+    @classmethod
+    @DB.connection_context()
+    def record_pipeline_operation(cls, document_id, pipeline_id, task_type, fake_document_ids=[]):
+        return cls.create(document_id=document_id, pipeline_id=pipeline_id, task_type=task_type, fake_document_ids=fake_document_ids)
+
+    @classmethod
+    @DB.connection_context()
+    def get_file_logs_by_kb_id(cls, kb_id, page_number, items_per_page, orderby, desc, keywords, operation_status, types, suffix, create_date_from=None, create_date_to=None):
+        fields = cls.get_file_logs_fields()
+        if keywords:
+            logs = cls.model.select(*fields).where((cls.model.kb_id == kb_id), (fn.LOWER(cls.model.document_name).contains(keywords.lower())))
+        else:
+            logs = cls.model.select(*fields).where(cls.model.kb_id == kb_id)
+
+        logs = logs.where(cls.model.document_id != GRAPH_RAPTOR_FAKE_DOC_ID)
+
+        if operation_status:
+            logs = logs.where(cls.model.operation_status.in_(operation_status))
+        if types:
+            logs = logs.where(cls.model.document_type.in_(types))
+        if suffix:
+            logs = logs.where(cls.model.document_suffix.in_(suffix))
+        if create_date_from:
+            logs = logs.where(cls.model.create_date >= create_date_from)
+        if create_date_to:
+            logs = logs.where(cls.model.create_date <= create_date_to)
+
+        count = logs.count()
+        if desc:
+            logs = logs.order_by(cls.model.getter_by(orderby).desc())
+        else:
+            logs = logs.order_by(cls.model.getter_by(orderby).asc())
+
+        if page_number and items_per_page:
+            logs = logs.paginate(page_number, items_per_page)
+
+        return list(logs.dicts()), count
+
+    @classmethod
+    @DB.connection_context()
+    def get_documents_info(cls, id):
+        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
+            )
+            .dicts()
+        )
+
+    @classmethod
+    @DB.connection_context()
+    def get_dataset_logs_by_kb_id(cls, kb_id, page_number, items_per_page, orderby, desc, operation_status, create_date_from=None, create_date_to=None):
+        fields = cls.get_dataset_logs_fields()
+        logs = cls.model.select(*fields).where((cls.model.kb_id == kb_id), (cls.model.document_id == GRAPH_RAPTOR_FAKE_DOC_ID))
+
+        if operation_status:
+            logs = logs.where(cls.model.operation_status.in_(operation_status))
+        if create_date_from:
+            logs = logs.where(cls.model.create_date >= create_date_from)
+        if create_date_to:
+            logs = logs.where(cls.model.create_date <= create_date_to)
+
+        count = logs.count()
+        if desc:
+            logs = logs.order_by(cls.model.getter_by(orderby).desc())
+        else:
+            logs = logs.order_by(cls.model.getter_by(orderby).asc())
+
+        if page_number and items_per_page:
+            logs = logs.paginate(page_number, items_per_page)
+
+        return list(logs.dicts()), count

api/db/services/search_service.py

@@ -110,3 +110,8 @@ class SearchService(CommonService):
             query = query.paginate(page_number, items_per_page)
 
         return list(query.dicts()), count
+
+    @classmethod
+    @DB.connection_context()
+    def delete_by_tenant_id(cls, tenant_id):
+        return cls.model.delete().where(cls.model.tenant_id == tenant_id).execute()

api/db/services/task_service.py

@@ -35,6 +35,8 @@ from rag.utils.redis_conn import REDIS_CONN
 from api import settings
 from rag.nlp import search
 
+CANVAS_DEBUG_DOC_ID = "dataflow_x"
+GRAPH_RAPTOR_FAKE_DOC_ID = "graph_raptor_x"
 
 def trim_header_by_lines(text: str, max_length) -> str:
     # Trim header text to maximum length while preserving line breaks
@@ -70,7 +72,7 @@ class TaskService(CommonService):
 
     @classmethod
     @DB.connection_context()
-    def get_task(cls, task_id):
+    def get_task(cls, task_id, doc_ids=[]):
         """Retrieve detailed task information by task ID.
 
         This method fetches comprehensive task details including associated document,
@@ -84,6 +86,10 @@ class TaskService(CommonService):
             dict: Task details dictionary containing all task information and related metadata.
                  Returns None if task is not found or has exceeded retry limit.
         """
+        doc_id = cls.model.doc_id
+        if doc_id == CANVAS_DEBUG_DOC_ID and doc_ids:
+            doc_id = doc_ids[0]
+
         fields = [
             cls.model.id,
             cls.model.doc_id,
@@ -109,7 +115,7 @@ class TaskService(CommonService):
         ]
         docs = (
             cls.model.select(*fields)
-                .join(Document, on=(cls.model.doc_id == Document.id))
+                .join(Document, on=(doc_id == Document.id))
                 .join(Knowledgebase, on=(Document.kb_id == Knowledgebase.id))
                 .join(Tenant, on=(Knowledgebase.tenant_id == Tenant.id))
                 .where(cls.model.id == task_id)
@@ -292,21 +298,29 @@ class TaskService(CommonService):
                         ((prog == -1) | (prog > cls.model.progress))
                     )
                 ).execute()
-            return
+        else:
+            with DB.lock("update_progress", -1):
+                if info["progress_msg"]:
+                    progress_msg = trim_header_by_lines(task.progress_msg + "\n" + info["progress_msg"], 3000)
+                    cls.model.update(progress_msg=progress_msg).where(cls.model.id == id).execute()
+                if "progress" in info:
+                    prog = info["progress"]
+                    cls.model.update(progress=prog).where(
+                        (cls.model.id == id) &
+                        (
+                            (cls.model.progress != -1) &
+                            ((prog == -1) | (prog > cls.model.progress))
+                        )
+                    ).execute()
 
-        with DB.lock("update_progress", -1):
-            if info["progress_msg"]:
-                progress_msg = trim_header_by_lines(task.progress_msg + "\n" + info["progress_msg"], 3000)
-                cls.model.update(progress_msg=progress_msg).where(cls.model.id == id).execute()
-            if "progress" in info:
-                prog = info["progress"]
-                cls.model.update(progress=prog).where(
-                    (cls.model.id == id) &
-                    (
-                        (cls.model.progress != -1) &
-                        ((prog == -1) | (prog > cls.model.progress))
-                    )
-                ).execute()
+        process_duration = (datetime.now() - task.begin_at).total_seconds()
+        cls.model.update(process_duration=process_duration).where(cls.model.id == id).execute()
+
+    @classmethod
+    @DB.connection_context()
+    def delete_by_doc_ids(cls, doc_ids):
+        """Delete task associated with a document."""
+        return cls.model.delete().where(cls.model.doc_id.in_(doc_ids)).execute()
 
 
 def queue_tasks(doc: dict, bucket: str, name: str, priority: int):
@@ -330,7 +344,14 @@ def queue_tasks(doc: dict, bucket: str, name: str, priority: int):
         - Previous task chunks may be reused if available
     """
     def new_task():
-        return {"id": get_uuid(), "doc_id": doc["id"], "progress": 0.0, "from_page": 0, "to_page": 100000000}
+        return {
+            "id": get_uuid(),
+            "doc_id": doc["id"],
+            "progress": 0.0,
+            "from_page": 0,
+            "to_page": 100000000,
+            "begin_at": datetime.now(),
+        }
 
     parse_task_array = []
 
@@ -343,7 +364,7 @@ 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:
@@ -472,33 +493,26 @@ def has_canceled(task_id):
     return False
 
 
-def queue_dataflow(dsl:str, tenant_id:str, doc_id:str, task_id:str, flow_id:str, priority: int, callback=None) -> tuple[bool, str]:
-    """
-    Returns a tuple (success: bool, error_message: str).
-    """
-    _ = callback
+def queue_dataflow(tenant_id:str, flow_id:str, task_id:str, doc_id:str=CANVAS_DEBUG_DOC_ID, file:dict=None, priority: int=0, rerun:bool=False) -> tuple[bool, str]:
 
     task = dict(
-    id=get_uuid() if not task_id else task_id,
-    doc_id=doc_id,
-    from_page=0,
-    to_page=100000000,
-    task_type="dataflow",
-    priority=priority,
+        id=task_id,
+        doc_id=doc_id,
+        from_page=0,
+        to_page=100000000,
+        task_type="dataflow" if not rerun else "dataflow_rerun",
+        priority=priority,
+        begin_at=datetime.now(),
     )
-
-    TaskService.model.delete().where(TaskService.model.id == task["id"]).execute()
+    if doc_id not in [CANVAS_DEBUG_DOC_ID, GRAPH_RAPTOR_FAKE_DOC_ID]:
+        TaskService.model.delete().where(TaskService.model.doc_id == doc_id).execute()
+        DocumentService.begin2parse(doc_id)
     bulk_insert_into_db(model=Task, data_source=[task], replace_on_conflict=True)
 
-    kb_id = DocumentService.get_knowledgebase_id(doc_id)
-    if not kb_id:
-        return False, f"Can't find KB of this document: {doc_id}"
-
-    task["kb_id"] = kb_id
+    task["kb_id"] = DocumentService.get_knowledgebase_id(doc_id)
     task["tenant_id"] = tenant_id
-    task["task_type"] = "dataflow"
-    task["dsl"] = dsl
-    task["dataflow_id"] = get_uuid() if not flow_id else flow_id
+    task["dataflow_id"] = flow_id
+    task["file"] = file
 
     if not REDIS_CONN.queue_product(
         get_svr_queue_name(priority), message=task

api/db/services/tenant_llm_service.py

@@ -209,6 +209,11 @@ class TenantLLMService(CommonService):
         objs = cls.model.select().where((cls.model.llm_factory == "OpenAI"), ~(cls.model.llm_name == "text-embedding-3-small"), ~(cls.model.llm_name == "text-embedding-3-large")).dicts()
         return list(objs)
 
+    @classmethod
+    @DB.connection_context()
+    def delete_by_tenant_id(cls, tenant_id):
+        return cls.model.delete().where(cls.model.tenant_id == tenant_id).execute()
+
     @staticmethod
     def llm_id2llm_type(llm_id: str) -> str | None:
         from api.db.services.llm_service import LLMService

api/db/services/user_canvas_version.py

@@ -24,7 +24,24 @@ class UserCanvasVersionService(CommonService):
             return None
         except Exception:
             return None
-    
+
+    @classmethod
+    @DB.connection_context()
+    def get_all_canvas_version_by_canvas_ids(cls, canvas_ids):
+        fields = [cls.model.id]
+        versions = cls.model.select(*fields).where(cls.model.user_canvas_id.in_(canvas_ids))
+        versions.order_by(cls.model.create_time.asc())
+        offset, limit = 0, 100
+        res = []
+        while True:
+            version_batch = versions.offset(offset).limit(limit)
+            _temp = list(version_batch.dicts())
+            if not _temp:
+                break
+            res.extend(_temp)
+            offset += limit
+        return res
+
     @classmethod
     @DB.connection_context()
     def delete_all_versions(cls, user_canvas_id):

api/db/services/user_service.py

@@ -100,6 +100,12 @@ class UserService(CommonService):
         else:
             return None
 
+    @classmethod
+    @DB.connection_context()
+    def query_user_by_email(cls, email):
+        users = cls.model.select().where((cls.model.email == email))
+        return list(users)
+
     @classmethod
     @DB.connection_context()
     def save(cls, **kwargs):
@@ -133,6 +139,17 @@ class UserService(CommonService):
                 cls.model.update(user_dict).where(
                     cls.model.id == user_id).execute()
 
+    @classmethod
+    @DB.connection_context()
+    def update_user_password(cls, user_id, new_password):
+        with DB.atomic():
+            update_dict = {
+                "password": generate_password_hash(str(new_password)),
+                "update_time": current_timestamp(),
+                "update_date": datetime_format(datetime.now())
+            }
+            cls.model.update(update_dict).where(cls.model.id == user_id).execute()
+
     @classmethod
     @DB.connection_context()
     def is_admin(cls, user_id):
@@ -271,6 +288,17 @@ class UserTenantService(CommonService):
                     .join(User, on=((cls.model.tenant_id == User.id) & (UserTenant.user_id == user_id) & (UserTenant.status == StatusEnum.VALID.value)))
                     .where(cls.model.status == StatusEnum.VALID.value).dicts())
 
+    @classmethod
+    @DB.connection_context()
+    def get_user_tenant_relation_by_user_id(cls, user_id):
+        fields = [
+            cls.model.id,
+            cls.model.user_id,
+            cls.model.tenant_id,
+            cls.model.role
+        ]
+        return list(cls.model.select(*fields).where(cls.model.user_id == user_id).dicts().dicts())
+
     @classmethod
     @DB.connection_context()
     def get_num_members(cls, user_id: str):

api/ragflow_server.py

@@ -41,7 +41,7 @@ from api import utils
 from api.db.db_models import init_database_tables as init_web_db
 from api.db.init_data import init_web_data
 from api.versions import get_ragflow_version
-from api.utils import show_configs
+from api.utils.configs import show_configs
 from rag.settings import print_rag_settings
 from rag.utils.mcp_tool_call_conn import shutdown_all_mcp_sessions
 from rag.utils.redis_conn import RedisDistributedLock

api/settings.py

@@ -24,7 +24,7 @@ import rag.utils.es_conn
 import rag.utils.infinity_conn
 import rag.utils.opensearch_conn
 from api.constants import RAG_FLOW_SERVICE_NAME
-from api.utils import decrypt_database_config, get_base_config
+from api.utils.configs import decrypt_database_config, get_base_config
 from api.utils.file_utils import get_project_base_directory
 from rag.nlp import search
 

api/utils/__init__.py

@@ -16,184 +16,15 @@
 import base64
 import datetime
 import hashlib
-import io
-import json
 import os
-import pickle
 import socket
 import time
 import uuid
 import requests
-import logging
-import copy
-from enum import Enum, IntEnum
+
 import importlib
-from Cryptodome.PublicKey import RSA
-from Cryptodome.Cipher import PKCS1_v1_5 as Cipher_pkcs1_v1_5
-from filelock import FileLock
-from api.constants import SERVICE_CONF
 
-from . import file_utils
-
-
-def conf_realpath(conf_name):
-    conf_path = f"conf/{conf_name}"
-    return os.path.join(file_utils.get_project_base_directory(), conf_path)
-
-
-def read_config(conf_name=SERVICE_CONF):
-    local_config = {}
-    local_path = conf_realpath(f'local.{conf_name}')
-
-    # load local config file
-    if os.path.exists(local_path):
-        local_config = file_utils.load_yaml_conf(local_path)
-        if not isinstance(local_config, dict):
-            raise ValueError(f'Invalid config file: "{local_path}".')
-
-    global_config_path = conf_realpath(conf_name)
-    global_config = file_utils.load_yaml_conf(global_config_path)
-
-    if not isinstance(global_config, dict):
-        raise ValueError(f'Invalid config file: "{global_config_path}".')
-
-    global_config.update(local_config)
-    return global_config
-
-
-CONFIGS = read_config()
-
-
-def show_configs():
-    msg = f"Current configs, from {conf_realpath(SERVICE_CONF)}:"
-    for k, v in CONFIGS.items():
-        if isinstance(v, dict):
-            if "password" in v:
-                v = copy.deepcopy(v)
-                v["password"] = "*" * 8
-            if "access_key" in v:
-                v = copy.deepcopy(v)
-                v["access_key"] = "*" * 8
-            if "secret_key" in v:
-                v = copy.deepcopy(v)
-                v["secret_key"] = "*" * 8
-            if "secret" in v:
-                v = copy.deepcopy(v)
-                v["secret"] = "*" * 8
-            if "sas_token" in v:
-                v = copy.deepcopy(v)
-                v["sas_token"] = "*" * 8
-            if "oauth" in k:
-                v =  copy.deepcopy(v)
-                for key, val in v.items():
-                  if "client_secret" in val:
-                      val["client_secret"] = "*" * 8
-            if "authentication" in k:
-                v =  copy.deepcopy(v)
-                for key, val in v.items():
-                  if "http_secret_key" in val:
-                      val["http_secret_key"] = "*" * 8
-        msg += f"\n\t{k}: {v}"
-    logging.info(msg)
-
-
-def get_base_config(key, default=None):
-    if key is None:
-        return None
-    if default is None:
-        default = os.environ.get(key.upper())
-    return CONFIGS.get(key, default)
-
-
-use_deserialize_safe_module = get_base_config(
-    'use_deserialize_safe_module', False)
-
-
-class BaseType:
-    def to_dict(self):
-        return dict([(k.lstrip("_"), v) for k, v in self.__dict__.items()])
-
-    def to_dict_with_type(self):
-        def _dict(obj):
-            module = None
-            if issubclass(obj.__class__, BaseType):
-                data = {}
-                for attr, v in obj.__dict__.items():
-                    k = attr.lstrip("_")
-                    data[k] = _dict(v)
-                module = obj.__module__
-            elif isinstance(obj, (list, tuple)):
-                data = []
-                for i, vv in enumerate(obj):
-                    data.append(_dict(vv))
-            elif isinstance(obj, dict):
-                data = {}
-                for _k, vv in obj.items():
-                    data[_k] = _dict(vv)
-            else:
-                data = obj
-            return {"type": obj.__class__.__name__,
-                    "data": data, "module": module}
-
-        return _dict(self)
-
-
-class CustomJSONEncoder(json.JSONEncoder):
-    def __init__(self, **kwargs):
-        self._with_type = kwargs.pop("with_type", False)
-        super().__init__(**kwargs)
-
-    def default(self, obj):
-        if isinstance(obj, datetime.datetime):
-            return obj.strftime('%Y-%m-%d %H:%M:%S')
-        elif isinstance(obj, datetime.date):
-            return obj.strftime('%Y-%m-%d')
-        elif isinstance(obj, datetime.timedelta):
-            return str(obj)
-        elif issubclass(type(obj), Enum) or issubclass(type(obj), IntEnum):
-            return obj.value
-        elif isinstance(obj, set):
-            return list(obj)
-        elif issubclass(type(obj), BaseType):
-            if not self._with_type:
-                return obj.to_dict()
-            else:
-                return obj.to_dict_with_type()
-        elif isinstance(obj, type):
-            return obj.__name__
-        else:
-            return json.JSONEncoder.default(self, obj)
-
-
-def rag_uuid():
-    return uuid.uuid1().hex
-
-
-def string_to_bytes(string):
-    return string if isinstance(
-        string, bytes) else string.encode(encoding="utf-8")
-
-
-def bytes_to_string(byte):
-    return byte.decode(encoding="utf-8")
-
-
-def json_dumps(src, byte=False, indent=None, with_type=False):
-    dest = json.dumps(
-        src,
-        indent=indent,
-        cls=CustomJSONEncoder,
-        with_type=with_type)
-    if byte:
-        dest = string_to_bytes(dest)
-    return dest
-
-
-def json_loads(src, object_hook=None, object_pairs_hook=None):
-    if isinstance(src, bytes):
-        src = bytes_to_string(src)
-    return json.loads(src, object_hook=object_hook,
-                      object_pairs_hook=object_pairs_hook)
+from .common import string_to_bytes
 
 
 def current_timestamp():
@@ -215,45 +46,6 @@ def date_string_to_timestamp(time_str, format_string="%Y-%m-%d %H:%M:%S"):
     return time_stamp
 
 
-def serialize_b64(src, to_str=False):
-    dest = base64.b64encode(pickle.dumps(src))
-    if not to_str:
-        return dest
-    else:
-        return bytes_to_string(dest)
-
-
-def deserialize_b64(src):
-    src = base64.b64decode(
-        string_to_bytes(src) if isinstance(
-            src, str) else src)
-    if use_deserialize_safe_module:
-        return restricted_loads(src)
-    return pickle.loads(src)
-
-
-safe_module = {
-    'numpy',
-    'rag_flow'
-}
-
-
-class RestrictedUnpickler(pickle.Unpickler):
-    def find_class(self, module, name):
-        import importlib
-        if module.split('.')[0] in safe_module:
-            _module = importlib.import_module(module)
-            return getattr(_module, name)
-        # Forbid everything else.
-        raise pickle.UnpicklingError("global '%s.%s' is forbidden" %
-                                     (module, name))
-
-
-def restricted_loads(src):
-    """Helper function analogous to pickle.loads()."""
-    return RestrictedUnpickler(io.BytesIO(src)).load()
-
-
 def get_lan_ip():
     if os.name != "nt":
         import fcntl
@@ -298,47 +90,6 @@ def from_dict_hook(in_dict: dict):
         return in_dict
 
 
-def decrypt_database_password(password):
-    encrypt_password = get_base_config("encrypt_password", False)
-    encrypt_module = get_base_config("encrypt_module", False)
-    private_key = get_base_config("private_key", None)
-
-    if not password or not encrypt_password:
-        return password
-
-    if not private_key:
-        raise ValueError("No private key")
-
-    module_fun = encrypt_module.split("#")
-    pwdecrypt_fun = getattr(
-        importlib.import_module(
-            module_fun[0]),
-        module_fun[1])
-
-    return pwdecrypt_fun(private_key, password)
-
-
-def decrypt_database_config(
-        database=None, passwd_key="password", name="database"):
-    if not database:
-        database = get_base_config(name, {})
-
-    database[passwd_key] = decrypt_database_password(database[passwd_key])
-    return database
-
-
-def update_config(key, value, conf_name=SERVICE_CONF):
-    conf_path = conf_realpath(conf_name=conf_name)
-    if not os.path.isabs(conf_path):
-        conf_path = os.path.join(
-            file_utils.get_project_base_directory(), conf_path)
-
-    with FileLock(os.path.join(os.path.dirname(conf_path), ".lock")):
-        config = file_utils.load_yaml_conf(conf_path=conf_path) or {}
-        config[key] = value
-        file_utils.rewrite_yaml_conf(conf_path=conf_path, config=config)
-
-
 def get_uuid():
     return uuid.uuid1().hex
 
@@ -363,37 +114,6 @@ def elapsed2time(elapsed):
     return '%02d:%02d:%02d' % (hour, minuter, second)
 
 
-def decrypt(line):
-    file_path = os.path.join(
-        file_utils.get_project_base_directory(),
-        "conf",
-        "private.pem")
-    rsa_key = RSA.importKey(open(file_path).read(), "Welcome")
-    cipher = Cipher_pkcs1_v1_5.new(rsa_key)
-    return cipher.decrypt(base64.b64decode(
-        line), "Fail to decrypt password!").decode('utf-8')
-
-
-def decrypt2(crypt_text):
-    from base64 import b64decode, b16decode
-    from Crypto.Cipher import PKCS1_v1_5 as Cipher_PKCS1_v1_5
-    from Crypto.PublicKey import RSA
-    decode_data = b64decode(crypt_text)
-    if len(decode_data) == 127:
-        hex_fixed = '00' + decode_data.hex()
-        decode_data = b16decode(hex_fixed.upper())
-
-    file_path = os.path.join(
-        file_utils.get_project_base_directory(),
-        "conf",
-        "private.pem")
-    pem = open(file_path).read()
-    rsa_key = RSA.importKey(pem, "Welcome")
-    cipher = Cipher_PKCS1_v1_5.new(rsa_key)
-    decrypt_text = cipher.decrypt(decode_data, None)
-    return (b64decode(decrypt_text)).decode()
-
-
 def download_img(url):
     if not url:
         return ""
@@ -408,5 +128,5 @@ def delta_seconds(date_string: str):
     return (datetime.datetime.now() - dt).total_seconds()
 
 
-def hash_str2int(line:str, mod: int=10 ** 8) -> int:
-    return int(hashlib.sha1(line.encode("utf-8")).hexdigest(), 16) % mod
+def hash_str2int(line: str, mod: int = 10 ** 8) -> int:
+    return int(hashlib.sha1(line.encode("utf-8")).hexdigest(), 16) % mod

api/utils/api_utils.py

@@ -39,6 +39,7 @@ from flask import (
     make_response,
     send_file,
 )
+from flask_login import current_user
 from flask import (
     request as flask_request,
 )
@@ -48,10 +49,13 @@ from werkzeug.http import HTTP_STATUS_CODES
 
 from api import settings
 from api.constants import REQUEST_MAX_WAIT_SEC, REQUEST_WAIT_SEC
+from api.db import ActiveEnum
 from api.db.db_models import APIToken
+from api.db.services import UserService
 from api.db.services.llm_service import LLMService
 from api.db.services.tenant_llm_service import TenantLLMService
-from api.utils import CustomJSONEncoder, get_uuid, json_dumps
+from api.utils.json import CustomJSONEncoder, json_dumps
+from api.utils import get_uuid
 from rag.utils.mcp_tool_call_conn import MCPToolCallSession, close_multiple_mcp_toolcall_sessions
 
 requests.models.complexjson.dumps = functools.partial(json.dumps, cls=CustomJSONEncoder)
@@ -226,6 +230,18 @@ def not_allowed_parameters(*params):
     return decorator
 
 
+def active_required(f):
+    @wraps(f)
+    def wrapper(*args, **kwargs):
+        user_id = current_user.id
+        usr = UserService.filter_by_id(user_id)
+        # check is_active
+        if not usr or not usr.is_active == ActiveEnum.ACTIVE.value:
+            return get_json_result(code=settings.RetCode.FORBIDDEN, message="User isn't active, please activate first.")
+        return f(*args, **kwargs)
+    return wrapper
+
+
 def is_localhost(ip):
     return ip in {"127.0.0.1", "::1", "[::1]", "localhost"}
 
@@ -643,6 +659,16 @@ def remap_dictionary_keys(source_data: dict, key_aliases: dict = None) -> dict:
     return transformed_data
 
 
+def group_by(list_of_dict, key):
+    res = {}
+    for item in list_of_dict:
+        if item[key] in res.keys():
+            res[item[key]].append(item)
+        else:
+            res[item[key]] = [item]
+    return res
+
+
 def get_mcp_tools(mcp_servers: list, timeout: float | int = 10) -> tuple[dict, str]:
     results = {}
     tool_call_sessions = []
@@ -679,7 +705,9 @@ TimeoutException = Union[Type[BaseException], BaseException]
 OnTimeoutCallback = Union[Callable[..., Any], Coroutine[Any, Any, Any]]
 
 
-def timeout(seconds: float | int = None, attempts: int = 2, *, exception: Optional[TimeoutException] = None, on_timeout: Optional[OnTimeoutCallback] = None):
+def timeout(seconds: float | int | str = None, attempts: int = 2, *, exception: Optional[TimeoutException] = None, on_timeout: Optional[OnTimeoutCallback] = None):
+    if isinstance(seconds, str):
+        seconds = float(seconds)
     def decorator(func):
         @wraps(func)
         def wrapper(*args, **kwargs):

api/utils/base64_image.py

@@ -1,3 +1,56 @@
 import base64
+import logging
+from functools import partial
+from io import BytesIO
+
+from PIL import Image
+
 test_image_base64 = "iVBORw0KGgoAAAANSUhEUgAAAGQAAABkCAIAAAD/gAIDAAAA6ElEQVR4nO3QwQ3AIBDAsIP9d25XIC+EZE8QZc18w5l9O+AlZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBT+IYAHHLHkdEgAAAABJRU5ErkJggg=="
-test_image = base64.b64decode(test_image_base64)
+test_image = base64.b64decode(test_image_base64)
+
+
+async def image2id(d: dict, storage_put_func: partial, objname:str, bucket:str="imagetemps"):
+    import logging
+    from io import BytesIO
+    import trio
+    from rag.svr.task_executor import minio_limiter
+    if not d.get("image"):
+        return
+
+    with BytesIO() as output_buffer:
+        if isinstance(d["image"], bytes):
+            output_buffer.write(d["image"])
+            output_buffer.seek(0)
+        else:
+            # If the image is in RGBA mode, convert it to RGB mode before saving it in JPEG format.
+            if d["image"].mode in ("RGBA", "P"):
+                converted_image = d["image"].convert("RGB")
+                d["image"] = converted_image
+            try:
+                d["image"].save(output_buffer, format='JPEG')
+            except OSError as e:
+                logging.warning(
+                    "Saving image exception, ignore: {}".format(str(e)))
+
+        async with minio_limiter:
+            await trio.to_thread.run_sync(lambda: storage_put_func(bucket=bucket, fnm=objname, binary=output_buffer.getvalue()))
+        d["img_id"] = f"{bucket}-{objname}"
+        if not isinstance(d["image"], bytes):
+            d["image"].close()
+        del d["image"]  # Remove image reference
+
+
+def id2image(image_id:str|None, storage_get_func: partial):
+    if not image_id:
+        return
+    arr = image_id.split("-")
+    if len(arr) != 2:
+        return
+    bkt, nm = image_id.split("-")
+    try:
+        blob = storage_get_func(bucket=bkt, filename=nm)
+        if not blob:
+            return
+        return Image.open(BytesIO(blob))
+    except Exception as e:
+        logging.exception(e)

api/utils/common.py

@@ -0,0 +1,23 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+def string_to_bytes(string):
+    return string if isinstance(
+        string, bytes) else string.encode(encoding="utf-8")
+
+
+def bytes_to_string(byte):
+    return byte.decode(encoding="utf-8")

api/utils/configs.py

@@ -0,0 +1,179 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+import os
+import io
+import copy
+import logging
+import base64
+import pickle
+import importlib
+
+from api.utils import file_utils
+from filelock import FileLock
+from api.utils.common import bytes_to_string, string_to_bytes
+from api.constants import SERVICE_CONF
+
+
+def conf_realpath(conf_name):
+    conf_path = f"conf/{conf_name}"
+    return os.path.join(file_utils.get_project_base_directory(), conf_path)
+
+
+def read_config(conf_name=SERVICE_CONF):
+    local_config = {}
+    local_path = conf_realpath(f'local.{conf_name}')
+
+    # load local config file
+    if os.path.exists(local_path):
+        local_config = file_utils.load_yaml_conf(local_path)
+        if not isinstance(local_config, dict):
+            raise ValueError(f'Invalid config file: "{local_path}".')
+
+    global_config_path = conf_realpath(conf_name)
+    global_config = file_utils.load_yaml_conf(global_config_path)
+
+    if not isinstance(global_config, dict):
+        raise ValueError(f'Invalid config file: "{global_config_path}".')
+
+    global_config.update(local_config)
+    return global_config
+
+
+CONFIGS = read_config()
+
+
+def show_configs():
+    msg = f"Current configs, from {conf_realpath(SERVICE_CONF)}:"
+    for k, v in CONFIGS.items():
+        if isinstance(v, dict):
+            if "password" in v:
+                v = copy.deepcopy(v)
+                v["password"] = "*" * 8
+            if "access_key" in v:
+                v = copy.deepcopy(v)
+                v["access_key"] = "*" * 8
+            if "secret_key" in v:
+                v = copy.deepcopy(v)
+                v["secret_key"] = "*" * 8
+            if "secret" in v:
+                v = copy.deepcopy(v)
+                v["secret"] = "*" * 8
+            if "sas_token" in v:
+                v = copy.deepcopy(v)
+                v["sas_token"] = "*" * 8
+            if "oauth" in k:
+                v = copy.deepcopy(v)
+                for key, val in v.items():
+                    if "client_secret" in val:
+                        val["client_secret"] = "*" * 8
+            if "authentication" in k:
+                v = copy.deepcopy(v)
+                for key, val in v.items():
+                    if "http_secret_key" in val:
+                        val["http_secret_key"] = "*" * 8
+        msg += f"\n\t{k}: {v}"
+    logging.info(msg)
+
+
+def get_base_config(key, default=None):
+    if key is None:
+        return None
+    if default is None:
+        default = os.environ.get(key.upper())
+    return CONFIGS.get(key, default)
+
+
+def decrypt_database_password(password):
+    encrypt_password = get_base_config("encrypt_password", False)
+    encrypt_module = get_base_config("encrypt_module", False)
+    private_key = get_base_config("private_key", None)
+
+    if not password or not encrypt_password:
+        return password
+
+    if not private_key:
+        raise ValueError("No private key")
+
+    module_fun = encrypt_module.split("#")
+    pwdecrypt_fun = getattr(
+        importlib.import_module(
+            module_fun[0]),
+        module_fun[1])
+
+    return pwdecrypt_fun(private_key, password)
+
+
+def decrypt_database_config(
+        database=None, passwd_key="password", name="database"):
+    if not database:
+        database = get_base_config(name, {})
+
+    database[passwd_key] = decrypt_database_password(database[passwd_key])
+    return database
+
+
+def update_config(key, value, conf_name=SERVICE_CONF):
+    conf_path = conf_realpath(conf_name=conf_name)
+    if not os.path.isabs(conf_path):
+        conf_path = os.path.join(
+            file_utils.get_project_base_directory(), conf_path)
+
+    with FileLock(os.path.join(os.path.dirname(conf_path), ".lock")):
+        config = file_utils.load_yaml_conf(conf_path=conf_path) or {}
+        config[key] = value
+        file_utils.rewrite_yaml_conf(conf_path=conf_path, config=config)
+
+
+safe_module = {
+    'numpy',
+    'rag_flow'
+}
+
+
+class RestrictedUnpickler(pickle.Unpickler):
+    def find_class(self, module, name):
+        import importlib
+        if module.split('.')[0] in safe_module:
+            _module = importlib.import_module(module)
+            return getattr(_module, name)
+        # Forbid everything else.
+        raise pickle.UnpicklingError("global '%s.%s' is forbidden" %
+                                     (module, name))
+
+
+def restricted_loads(src):
+    """Helper function analogous to pickle.loads()."""
+    return RestrictedUnpickler(io.BytesIO(src)).load()
+
+
+def serialize_b64(src, to_str=False):
+    dest = base64.b64encode(pickle.dumps(src))
+    if not to_str:
+        return dest
+    else:
+        return bytes_to_string(dest)
+
+
+def deserialize_b64(src):
+    src = base64.b64decode(
+        string_to_bytes(src) if isinstance(
+            src, str) else src)
+    use_deserialize_safe_module = get_base_config(
+        'use_deserialize_safe_module', False)
+    if use_deserialize_safe_module:
+        return restricted_loads(src)
+    return pickle.loads(src)

api/utils/crypt.py

@@ -0,0 +1,64 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+import base64
+import os
+import sys
+from Cryptodome.PublicKey import RSA
+from Cryptodome.Cipher import PKCS1_v1_5 as Cipher_pkcs1_v1_5
+from api.utils import file_utils
+
+
+def crypt(line):
+    """
+    decrypt(crypt(input_string)) == base64(input_string), which frontend and admin_client use.
+    """
+    file_path = os.path.join(file_utils.get_project_base_directory(), "conf", "public.pem")
+    rsa_key = RSA.importKey(open(file_path).read(), "Welcome")
+    cipher = Cipher_pkcs1_v1_5.new(rsa_key)
+    password_base64 = base64.b64encode(line.encode('utf-8')).decode("utf-8")
+    encrypted_password = cipher.encrypt(password_base64.encode())
+    return base64.b64encode(encrypted_password).decode('utf-8')
+
+
+def decrypt(line):
+    file_path = os.path.join(file_utils.get_project_base_directory(), "conf", "private.pem")
+    rsa_key = RSA.importKey(open(file_path).read(), "Welcome")
+    cipher = Cipher_pkcs1_v1_5.new(rsa_key)
+    return cipher.decrypt(base64.b64decode(line), "Fail to decrypt password!").decode('utf-8')
+
+
+def decrypt2(crypt_text):
+    from base64 import b64decode, b16decode
+    from Crypto.Cipher import PKCS1_v1_5 as Cipher_PKCS1_v1_5
+    from Crypto.PublicKey import RSA
+    decode_data = b64decode(crypt_text)
+    if len(decode_data) == 127:
+        hex_fixed = '00' + decode_data.hex()
+        decode_data = b16decode(hex_fixed.upper())
+
+    file_path = os.path.join(file_utils.get_project_base_directory(), "conf", "private.pem")
+    pem = open(file_path).read()
+    rsa_key = RSA.importKey(pem, "Welcome")
+    cipher = Cipher_PKCS1_v1_5.new(rsa_key)
+    decrypt_text = cipher.decrypt(decode_data, None)
+    return (b64decode(decrypt_text)).decode()
+
+
+if __name__ == "__main__":
+    passwd = crypt(sys.argv[1])
+    print(passwd)
+    print(decrypt(passwd))

api/utils/file_utils.py

@@ -155,7 +155,7 @@ def filename_type(filename):
     if re.match(r".*\.pdf$", filename):
         return FileType.PDF.value
 
-    if re.match(r".*\.(eml|doc|docx|ppt|pptx|yml|xml|htm|json|jsonl|ldjson|csv|txt|ini|xls|xlsx|wps|rtf|hlp|pages|numbers|key|md|py|js|java|c|cpp|h|php|go|ts|sh|cs|kt|html|sql)$", filename):
+    if re.match(r".*\.(msg|eml|doc|docx|ppt|pptx|yml|xml|htm|json|jsonl|ldjson|csv|txt|ini|xls|xlsx|wps|rtf|hlp|pages|numbers|key|md|py|js|java|c|cpp|h|php|go|ts|sh|cs|kt|html|sql)$", filename):
         return FileType.DOC.value
 
     if re.match(r".*\.(wav|flac|ape|alac|wavpack|wv|mp3|aac|ogg|vorbis|opus)$", filename):

api/utils/health.py

@@ -0,0 +1,104 @@
+from timeit import default_timer as timer
+
+from api import settings
+from api.db.db_models import DB
+from rag.utils.redis_conn import REDIS_CONN
+from rag.utils.storage_factory import STORAGE_IMPL
+
+
+def _ok_nok(ok: bool) -> str:
+    return "ok" if ok else "nok"
+
+
+def check_db() -> tuple[bool, dict]:
+    st = timer()
+    try:
+        # lightweight probe; works for MySQL/Postgres
+        DB.execute_sql("SELECT 1")
+        return True, {"elapsed": f"{(timer() - st) * 1000.0:.1f}"}
+    except Exception as e:
+        return False, {"elapsed": f"{(timer() - st) * 1000.0:.1f}", "error": str(e)}
+
+
+def check_redis() -> tuple[bool, dict]:
+    st = timer()
+    try:
+        ok = bool(REDIS_CONN.health())
+        return ok, {"elapsed": f"{(timer() - st) * 1000.0:.1f}"}
+    except Exception as e:
+        return False, {"elapsed": f"{(timer() - st) * 1000.0:.1f}", "error": str(e)}
+
+
+def check_doc_engine() -> tuple[bool, dict]:
+    st = timer()
+    try:
+        meta = settings.docStoreConn.health()
+        # treat any successful call as ok
+        return True, {"elapsed": f"{(timer() - st) * 1000.0:.1f}", **(meta or {})}
+    except Exception as e:
+        return False, {"elapsed": f"{(timer() - st) * 1000.0:.1f}", "error": str(e)}
+
+
+def check_storage() -> tuple[bool, dict]:
+    st = timer()
+    try:
+        STORAGE_IMPL.health()
+        return True, {"elapsed": f"{(timer() - st) * 1000.0:.1f}"}
+    except Exception as e:
+        return False, {"elapsed": f"{(timer() - st) * 1000.0:.1f}", "error": str(e)}
+
+
+def check_chat() -> tuple[bool, dict]:
+    st = timer()
+    try:
+        cfg = getattr(settings, "CHAT_CFG", None)
+        ok = bool(cfg and cfg.get("factory"))
+        return ok, {"elapsed": f"{(timer() - st) * 1000.0:.1f}"}
+    except Exception as e:
+        return False, {"elapsed": f"{(timer() - st) * 1000.0:.1f}", "error": str(e)}
+
+
+def run_health_checks() -> tuple[dict, bool]:
+    result: dict[str, str | dict] = {}
+
+    db_ok, db_meta = check_db()
+    chat_ok, chat_meta = check_chat()
+
+    result["db"] = _ok_nok(db_ok)
+    if not db_ok:
+        result.setdefault("_meta", {})["db"] = db_meta
+
+    result["chat"] = _ok_nok(chat_ok)
+    if not chat_ok:
+        result.setdefault("_meta", {})["chat"] = chat_meta
+
+    # Optional probes (do not change minimal contract but exposed for observability)
+    try:
+        redis_ok, redis_meta = check_redis()
+        result["redis"] = _ok_nok(redis_ok)
+        if not redis_ok:
+            result.setdefault("_meta", {})["redis"] = redis_meta
+    except Exception:
+        result["redis"] = "nok"
+
+    try:
+        doc_ok, doc_meta = check_doc_engine()
+        result["doc_engine"] = _ok_nok(doc_ok)
+        if not doc_ok:
+            result.setdefault("_meta", {})["doc_engine"] = doc_meta
+    except Exception:
+        result["doc_engine"] = "nok"
+
+    try:
+        sto_ok, sto_meta = check_storage()
+        result["storage"] = _ok_nok(sto_ok)
+        if not sto_ok:
+            result.setdefault("_meta", {})["storage"] = sto_meta
+    except Exception:
+        result["storage"] = "nok"
+
+    all_ok = (result.get("db") == "ok") and (result.get("chat") == "ok")
+    result["status"] = "ok" if all_ok else "nok"
+    return result, all_ok
+
+

api/utils/health_utils.py

@@ -1,3 +1,20 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+
 from timeit import default_timer as timer
 
 from api import settings

api/utils/json.py

@@ -0,0 +1,78 @@
+import datetime
+import json
+from enum import Enum, IntEnum
+from api.utils.common import string_to_bytes, bytes_to_string
+
+
+class BaseType:
+    def to_dict(self):
+        return dict([(k.lstrip("_"), v) for k, v in self.__dict__.items()])
+
+    def to_dict_with_type(self):
+        def _dict(obj):
+            module = None
+            if issubclass(obj.__class__, BaseType):
+                data = {}
+                for attr, v in obj.__dict__.items():
+                    k = attr.lstrip("_")
+                    data[k] = _dict(v)
+                module = obj.__module__
+            elif isinstance(obj, (list, tuple)):
+                data = []
+                for i, vv in enumerate(obj):
+                    data.append(_dict(vv))
+            elif isinstance(obj, dict):
+                data = {}
+                for _k, vv in obj.items():
+                    data[_k] = _dict(vv)
+            else:
+                data = obj
+            return {"type": obj.__class__.__name__,
+                    "data": data, "module": module}
+
+        return _dict(self)
+
+
+class CustomJSONEncoder(json.JSONEncoder):
+    def __init__(self, **kwargs):
+        self._with_type = kwargs.pop("with_type", False)
+        super().__init__(**kwargs)
+
+    def default(self, obj):
+        if isinstance(obj, datetime.datetime):
+            return obj.strftime('%Y-%m-%d %H:%M:%S')
+        elif isinstance(obj, datetime.date):
+            return obj.strftime('%Y-%m-%d')
+        elif isinstance(obj, datetime.timedelta):
+            return str(obj)
+        elif issubclass(type(obj), Enum) or issubclass(type(obj), IntEnum):
+            return obj.value
+        elif isinstance(obj, set):
+            return list(obj)
+        elif issubclass(type(obj), BaseType):
+            if not self._with_type:
+                return obj.to_dict()
+            else:
+                return obj.to_dict_with_type()
+        elif isinstance(obj, type):
+            return obj.__name__
+        else:
+            return json.JSONEncoder.default(self, obj)
+
+
+def json_dumps(src, byte=False, indent=None, with_type=False):
+    dest = json.dumps(
+        src,
+        indent=indent,
+        cls=CustomJSONEncoder,
+        with_type=with_type)
+    if byte:
+        dest = string_to_bytes(dest)
+    return dest
+
+
+def json_loads(src, object_hook=None, object_pairs_hook=None):
+    if isinstance(src, bytes):
+        src = bytes_to_string(src)
+    return json.loads(src, object_hook=object_hook,
+                      object_pairs_hook=object_pairs_hook)

api/utils/t_crypt.py

@@ -1,40 +0,0 @@
-#
-#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
-#
-#  Licensed under the Apache License, Version 2.0 (the "License");
-#  you may not use this file except in compliance with the License.
-#  You may obtain a copy of the License at
-#
-#      http://www.apache.org/licenses/LICENSE-2.0
-#
-#  Unless required by applicable law or agreed to in writing, software
-#  distributed under the License is distributed on an "AS IS" BASIS,
-#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-#  See the License for the specific language governing permissions and
-#  limitations under the License.
-#
-
-import base64
-import os
-import sys
-from Cryptodome.PublicKey import RSA
-from Cryptodome.Cipher import PKCS1_v1_5 as Cipher_pkcs1_v1_5
-from api.utils import decrypt, file_utils
-
-
-def crypt(line):
-    file_path = os.path.join(
-        file_utils.get_project_base_directory(),
-        "conf",
-        "public.pem")
-    rsa_key = RSA.importKey(open(file_path).read(),"Welcome")
-    cipher = Cipher_pkcs1_v1_5.new(rsa_key)
-    password_base64 = base64.b64encode(line.encode('utf-8')).decode("utf-8")
-    encrypted_password = cipher.encrypt(password_base64.encode())
-    return base64.b64encode(encrypted_password).decode('utf-8')
-
-
-if __name__ == "__main__":
-    passwd = crypt(sys.argv[1])
-    print(passwd)
-    print(decrypt(passwd))

conf/llm_factories.json

@@ -402,7 +402,7 @@
                     "is_tools": true
                 },
                 {
-                    "llm_name": "qwen3-max-preview",
+                    "llm_name": "qwen3-max",
                     "tags": "LLM,CHAT,256k",
                     "max_tokens": 256000,
                     "model_type": "chat",
@@ -436,6 +436,27 @@
                     "model_type": "chat",
                     "is_tools": true
                 },
+                {
+                    "llm_name": "qwen3-vl-plus",
+                    "tags": "LLM,CHAT,IMAGE2TEXT,256k",
+                    "max_tokens": 256000,
+                    "model_type": "image2text",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "qwen3-vl-235b-a22b-instruct",
+                    "tags": "LLM,CHAT,IMAGE2TEXT,128k",
+                    "max_tokens": 128000,
+                    "model_type": "image2text",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "qwen3-vl-235b-a22b-thinking",
+                    "tags": "LLM,CHAT,IMAGE2TEXT,128k",
+                    "max_tokens": 128000,
+                    "model_type": "image2text",
+                    "is_tools": true
+                },
                 {
                     "llm_name": "qwen3-235b-a22b-instruct-2507",
                     "tags": "LLM,CHAT,128k",
@@ -457,6 +478,20 @@
                     "model_type": "chat",
                     "is_tools": true
                 },
+                {
+                    "llm_name": "qwen3-next-80b-a3b-instruct",
+                    "tags": "LLM,CHAT,128k",
+                    "max_tokens": 128000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "qwen3-next-80b-a3b-thinking",
+                    "tags": "LLM,CHAT,128k",
+                    "max_tokens": 128000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
                 {
                     "llm_name": "qwen3-0.6b",
                     "tags": "LLM,CHAT,32k",
@@ -622,6 +657,13 @@
                     "tags": "SPEECH2TEXT,8k",
                     "max_tokens": 8000,
                     "model_type": "speech2text"
+                },
+                {
+                    "llm_name": "qianwen-deepresearch-30b-a3b-131k",
+                    "tags": "LLM,CHAT,1M,AGENT,DEEPRESEARCH",
+                    "max_tokens": 1000000,
+                    "model_type": "chat",
+                    "is_tools": true
                 }
             ]
         },

deepdoc/parser/figure_parser.py

@@ -19,7 +19,7 @@ from PIL import Image
 
 from api.utils.api_utils import timeout
 from rag.app.picture import vision_llm_chunk as picture_vision_llm_chunk
-from rag.prompts import vision_llm_figure_describe_prompt
+from rag.prompts.generator import vision_llm_figure_describe_prompt
 
 
 def vision_figure_parser_figure_data_wrapper(figures_data_without_positions):

deepdoc/parser/pdf_parser.py

@@ -37,7 +37,7 @@ from api.utils.file_utils import get_project_base_directory
 from deepdoc.vision import OCR, AscendLayoutRecognizer, LayoutRecognizer, Recognizer, TableStructureRecognizer
 from rag.app.picture import vision_llm_chunk as picture_vision_llm_chunk
 from rag.nlp import rag_tokenizer
-from rag.prompts import vision_llm_describe_prompt
+from rag.prompts.generator import vision_llm_describe_prompt
 from rag.settings import PARALLEL_DEVICES
 
 LOCK_KEY_pdfplumber = "global_shared_lock_pdfplumber"
@@ -1075,11 +1075,10 @@ class RAGFlowPdfParser:
         def insert_table_figures(tbls_or_figs, layout_type):
             def min_rectangle_distance(rect1, rect2):
                 import math
-
                 pn1, left1, right1, top1, bottom1 = rect1
                 pn2, left2, right2, top2, bottom2 = rect2
                 if right1 >= left2 and right2 >= left1 and bottom1 >= top2 and bottom2 >= top1:
-                    return 0 + (pn1 - pn2) * 10000
+                    return 0
                 if right1 < left2:
                     dx = left2 - right1
                 elif right2 < left1:
@@ -1092,20 +1091,27 @@ class RAGFlowPdfParser:
                     dy = top1 - bottom2
                 else:
                     dy = 0
-                return math.sqrt(dx * dx + dy * dy) + (pn1 - pn2) * 10000
+                return math.sqrt(dx*dx + dy*dy)# + (pn2-pn1)*10000
 
             for (img, txt), poss in tbls_or_figs:
                 bboxes = [(i, (b["page_number"], b["x0"], b["x1"], b["top"], b["bottom"])) for i, b in enumerate(self.boxes)]
-                dists = [(min_rectangle_distance((pn, left, right, top, bott), rect), i) for i, rect in bboxes for pn, left, right, top, bott in poss]
+                dists = [(min_rectangle_distance((pn, left, right, top+self.page_cum_height[pn], bott+self.page_cum_height[pn]), rect),i) for i, rect in bboxes for pn, left, right, top, bott in poss]
                 min_i = np.argmin(dists, axis=0)[0]
                 min_i, rect = bboxes[dists[min_i][-1]]
                 if isinstance(txt, list):
                     txt = "\n".join(txt)
-                self.boxes.insert(min_i, {"page_number": rect[0], "x0": rect[1], "x1": rect[2], "top": rect[3], "bottom": rect[4], "layout_type": layout_type, "text": txt, "image": img})
+                pn, left, right, top, bott = poss[0]
+                if self.boxes[min_i]["bottom"] < top+self.page_cum_height[pn]:
+                    min_i += 1
+                self.boxes.insert(min_i, {
+                    "page_number": pn+1, "x0": left, "x1": right, "top": top+self.page_cum_height[pn], "bottom": bott+self.page_cum_height[pn], "layout_type": layout_type, "text": txt, "image": img,
+                    "positions": [[pn+1, int(left), int(right), int(top), int(bott)]]
+                })
 
         for b in self.boxes:
             b["position_tag"] = self._line_tag(b, zoomin)
             b["image"] = self.crop(b["position_tag"], zoomin)
+            b["positions"] = [[pos[0][-1]+1, *pos[1:]] for pos in RAGFlowPdfParser.extract_positions(b["position_tag"])]
 
         insert_table_figures(tbls, "table")
         insert_table_figures(figs, "figure")
@@ -1123,7 +1129,7 @@ class RAGFlowPdfParser:
         for tag in re.findall(r"@@[0-9-]+\t[0-9.\t]+##", txt):
             pn, left, right, top, bottom = tag.strip("#").strip("@").split("\t")
             left, right, top, bottom = float(left), float(right), float(top), float(bottom)
-            poss.append(([int(p) - 1 for p in pn.split("-")], left, right, top, bottom))
+            poss.append(([int(p) - 1 for p in pn.split("-")], int(left), int(right), int(top), int(bottom)))
         return poss
 
     def crop(self, text, ZM=3, need_position=False):

deepdoc/vision/ocr.py

@@ -350,7 +350,7 @@ class TextRecognizer:
 
     def close(self):
         # close session and release manually
-        logging.info('Close TextRecognizer.')
+        logging.info('Close text recognizer.')
         if hasattr(self, "predictor"):
             del self.predictor
         gc.collect()
@@ -490,7 +490,7 @@ class TextDetector:
         return dt_boxes
 
     def close(self):
-        logging.info("Close TextDetector.")
+        logging.info("Close text detector.")
         if hasattr(self, "predictor"):
             del self.predictor
         gc.collect()

docs/develop/mcp/_category_.json

@@ -3,6 +3,6 @@
   "position": 40,
   "link": {
     "type": "generated-index",
-    "description": "Guides and references on accessing RAGFlow's knowledge bases via MCP."
+    "description": "Guides and references on accessing RAGFlow's datasets via MCP."
   }
 }

docs/develop/mcp/launch_mcp_server.md

@@ -14,9 +14,9 @@ A RAGFlow Model Context Protocol (MCP) server is designed as an independent comp
 An MCP server can start up in either self-host mode (default) or host mode: 
 
 - **Self-host mode**:  
-  When launching an MCP server in self-host mode, you must provide an API key to authenticate the MCP server with the RAGFlow server. In this mode, the MCP server can access *only* the datasets (knowledge bases) of a specified tenant on the RAGFlow server.
+  When launching an MCP server in self-host mode, you must provide an API key to authenticate the MCP server with the RAGFlow server. In this mode, the MCP server can access *only* the datasets of a specified tenant on the RAGFlow server.
 - **Host mode**:  
-  In host mode, each MCP client can access their own knowledge bases on the RAGFlow server. However, each client request must include a valid API key to authenticate the client with the RAGFlow server.
+  In host mode, each MCP client can access their own datasets on the RAGFlow server. However, each client request must include a valid API key to authenticate the client with the RAGFlow server.
 
 Once a connection is established, an MCP server communicates with its client in MCP HTTP+SSE (Server-Sent Events) mode, unidirectionally pushing responses from the RAGFlow server to its client in real time.
 

docs/faq.mdx

@@ -498,7 +498,7 @@ To switch your document engine from Elasticsearch to [Infinity](https://github.c
 
 ### Where are my uploaded files stored in RAGFlow's image?
 
-All uploaded files are stored in Minio, RAGFlow's object storage solution. For instance, if you upload your file directly to a knowledge base, it is located at `<knowledgebase_id>/filename`.
+All uploaded files are stored in Minio, RAGFlow's object storage solution. For instance, if you upload your file directly to a dataset, it is located at `<knowledgebase_id>/filename`.
 
 ---
 

docs/guides/agent/agent_component_reference/begin.mdx

@@ -67,14 +67,14 @@ You can tune document parsing and embedding efficiency by setting the environmen
 
 ## Frequently asked questions
 
-### Is the uploaded file in a knowledge base?
+### Is the uploaded file in a dataset?
 
-No. Files uploaded to an agent as input are not stored in a knowledge base and hence will not be processed using RAGFlow's built-in OCR, DLR or TSR models, or chunked using RAGFlow's built-in chunking methods.
+No. Files uploaded to an agent as input are not stored in a dataset and hence will not be processed using RAGFlow's built-in OCR, DLR or TSR models, or chunked using RAGFlow's built-in chunking methods.
 
 ### File size limit for an uploaded file
 
 There is no _specific_ file size limit for a file uploaded to an agent. However, note that model providers typically have a default or explicit maximum token setting, which can range from 8196 to 128k: The plain text part of the uploaded file will be passed in as the key value, but if the file's token count exceeds this limit, the string will be truncated and incomplete.
 
 :::tip NOTE
-The variables `MAX_CONTENT_LENGTH` in `/docker/.env` and `client_max_body_size` in `/docker/nginx/nginx.conf` set the file size limit for each upload to a knowledge base or **File Management**. These settings DO NOT apply in this scenario.
+The variables `MAX_CONTENT_LENGTH` in `/docker/.env` and `client_max_body_size` in `/docker/nginx/nginx.conf` set the file size limit for each upload to a dataset or **File Management**. These settings DO NOT apply in this scenario.
 :::

docs/guides/agent/agent_component_reference/retrieval.mdx

@@ -9,7 +9,7 @@ A component that retrieves information from specified datasets.
 
 ## Scenarios
 
-A **Retrieval** component is essential in most RAG scenarios, where information is extracted from designated knowledge bases before being sent to the LLM for content generation. A **Retrieval** component can operate either as a standalone workflow module or as a tool for an **Agent** component. In the latter role, the **Agent** component has autonomous control over when to invoke it for query and retrieval.
+A **Retrieval** component is essential in most RAG scenarios, where information is extracted from designated datasets before being sent to the LLM for content generation. A **Retrieval** component can operate either as a standalone workflow module or as a tool for an **Agent** component. In the latter role, the **Agent** component has autonomous control over when to invoke it for query and retrieval.
 
 The following screenshot shows a reference design using the **Retrieval** component, where the component serves as a tool for an **Agent** component. You can find it from the **Report Agent Using Knowledge Base** Agent template.
 
@@ -17,7 +17,7 @@ The following screenshot shows a reference design using the **Retrieval** compon
 
 ## Prerequisites
 
-Ensure you [have properly configured your target knowledge base(s)](../../dataset/configure_knowledge_base.md).
+Ensure you [have properly configured your target dataset(s)](../../dataset/configure_knowledge_base.md).
 
 ## Quickstart
 
@@ -36,9 +36,9 @@ The **Retrieval** component depends on query variables to specify its queries.
 
 By default, you can use `sys.query`, which is the user query and the default output of the **Begin** component. All global variables defined before the **Retrieval** component can also be used as query statements. Use the `(x)` button or type `/` to show all the available query variables.
 
-### 3. Select knowledge base(s) to query
+### 3. Select dataset(s) to query
 
-You can specify one or multiple knowledge bases to retrieve data from. If selecting mutiple, ensure they use the same embedding model.
+You can specify one or multiple datasets to retrieve data from. If selecting mutiple, ensure they use the same embedding model.
 
 ### 4. Expand **Advanced Settings** to configure the retrieval method
 
@@ -52,7 +52,7 @@ Using a rerank model will *significantly* increase the system's response time. I
 
 ### 5. Enable cross-language search
 
-If your user query is different from the languages of the knowledge bases, you can select the target languages in the **Cross-language search** dropdown menu. The model will then translates queries to ensure accurate matching of semantic meaning across languages.
+If your user query is different from the languages of the datasets, you can select the target languages in the **Cross-language search** dropdown menu. The model will then translates queries to ensure accurate matching of semantic meaning across languages.
 
 
 ### 6. Test retrieval results
@@ -76,10 +76,10 @@ The **Retrieval** component relies on query variables to specify its queries. Al
 
 ### Knowledge bases 
 
-Select the knowledge base(s) to retrieve data from.
+Select the dataset(s) to retrieve data from.
 
-- If no knowledge base is selected, meaning conversations with the agent will not be based on any knowledge base, ensure that the **Empty response** field is left blank to avoid an error.
-- If you select multiple knowledge bases, you must ensure that the knowledge bases (datasets) you select use the same embedding model; otherwise, an error message would occur.
+- If no dataset is selected, meaning conversations with the agent will not be based on any dataset, ensure that the **Empty response** field is left blank to avoid an error.
+- If you select multiple datasets, you must ensure that the datasets you select use the same embedding model; otherwise, an error message would occur.
 
 ### Similarity threshold
 
@@ -110,11 +110,11 @@ Using a rerank model will *significantly* increase the system's response time.
 
 ### Empty response
 
-- Set this as a response if no results are retrieved from the knowledge base(s) for your query, or 
+- Set this as a response if no results are retrieved from the dataset(s) for your query, or 
 - Leave this field blank to allow the chat model to improvise when nothing is found.
 
 :::caution WARNING
-If you do not specify a knowledge base, you must leave this field blank; otherwise, an error would occur.
+If you do not specify a dataset, you must leave this field blank; otherwise, an error would occur.
 :::
 
 ### Cross-language search
@@ -124,10 +124,10 @@ Select one or more languages for cross‑language search. If no language is sele
 ### Use knowledge graph
 
 :::caution IMPORTANT
-Before enabling this feature, ensure you have properly [constructed a knowledge graph from each target knowledge base](../../dataset/construct_knowledge_graph.md).
+Before enabling this feature, ensure you have properly [constructed a knowledge graph from each target dataset](../../dataset/construct_knowledge_graph.md).
 :::
 
-Whether to use knowledge graph(s) in the specified knowledge base(s) during retrieval for multi-hop question answering. When enabled, this would involve iterative searches across entity, relationship, and community report chunks, greatly increasing retrieval time.
+Whether to use knowledge graph(s) in the specified dataset(s) during retrieval for multi-hop question answering. When enabled, this would involve iterative searches across entity, relationship, and community report chunks, greatly increasing retrieval time.
 
 ### Output
 

docs/guides/agent/agent_introduction.md

@@ -27,7 +27,7 @@ Agents and RAG are complementary techniques, each enhancing the other’s capabi
 Before proceeding, ensure that:  
 
 1. You have properly set the LLM to use. See the guides on [Configure your API key](../models/llm_api_key_setup.md) or [Deploy a local LLM](../models/deploy_local_llm.mdx) for more information.
-2. You have a knowledge base configured and the corresponding files properly parsed. See the guide on [Configure a knowledge base](../dataset/configure_knowledge_base.md) for more information.
+2. You have a dataset configured and the corresponding files properly parsed. See the guide on [Configure a dataset](../dataset/configure_knowledge_base.md) for more information.
 
 :::
 

docs/guides/ai_search.md

@@ -22,7 +22,7 @@ When debugging your chat assistant, you can use AI search as a reference to veri
 ## Prerequisites
 
 - Ensure that you have configured the system's default models on the **Model providers** page.
-- Ensure that the intended knowledge bases are properly configured and the intended documents have finished file parsing.
+- Ensure that the intended datasets are properly configured and the intended documents have finished file parsing.
 
 ## Frequently asked questions
 

docs/guides/chat/set_chat_variables.md

@@ -25,13 +25,13 @@ In the **Variable** section, you add, remove, or update variables.
 
 ### `{knowledge}` - a reserved variable
 
-`{knowledge}` is the system's reserved variable, representing the chunks retrieved from the knowledge base(s) specified by **Knowledge bases** under the **Assistant settings** tab. If your chat assistant is associated with certain knowledge bases, you can keep it as is.
+`{knowledge}` is the system's reserved variable, representing the chunks retrieved from the dataset(s) specified by **Knowledge bases** under the **Assistant settings** tab. If your chat assistant is associated with certain datasets, you can keep it as is.
 
 :::info NOTE
 It currently makes no difference whether  `{knowledge}` is set as optional or mandatory, but please note this design will be updated in due course.
 :::
 
-From v0.17.0 onward, you can start an AI chat without specifying knowledge bases. In this case, we recommend removing the `{knowledge}` variable to prevent unnecessary reference and keeping the **Empty response** field empty to avoid errors.
+From v0.17.0 onward, you can start an AI chat without specifying datasets. In this case, we recommend removing the `{knowledge}` variable to prevent unnecessary reference and keeping the **Empty response** field empty to avoid errors.
 
 ### Custom variables
 
@@ -45,15 +45,15 @@ Besides `{knowledge}`, you can also define your own variables to pair with the s
 After you add or remove variables in the **Variable** section, ensure your changes are reflected in the system prompt to avoid inconsistencies or errors. Here's an example:
 
 ```
-You are an intelligent assistant. Please answer the question by summarizing chunks from the specified knowledge base(s)...
+You are an intelligent assistant. Please answer the question by summarizing chunks from the specified dataset(s)...
 
 Your answers should follow a professional and {style} style.
 
 ...
 
-Here is the knowledge base:
+Here is the dataset:
 {knowledge}
-The above is the knowledge base.
+The above is the dataset.
 ```
 
 :::tip NOTE

docs/guides/chat/start_chat.md

@@ -9,7 +9,7 @@ Initiate an AI-powered chat with a configured chat assistant.
 
 ---
 
-Knowledge base, hallucination-free chat, and file management are the three pillars of RAGFlow. Chats in RAGFlow are based on a particular knowledge base or multiple knowledge bases. Once you have created your knowledge base, finished file parsing, and [run a retrieval test](../dataset/run_retrieval_test.md), you can go ahead and start an AI conversation.
+Knowledge base, hallucination-free chat, and file management are the three pillars of RAGFlow. Chats in RAGFlow are based on a particular dataset or multiple datasets. Once you have created your dataset, finished file parsing, and [run a retrieval test](../dataset/run_retrieval_test.md), you can go ahead and start an AI conversation.
 
 ## Start an AI chat
 
@@ -21,12 +21,12 @@ You start an AI conversation by creating an assistant.
 
 2. Update **Assistant settings**:
 
-   - **Assistant name** is the name of your chat assistant. Each assistant corresponds to a dialogue with a unique combination of knowledge bases, prompts, hybrid search configurations, and large model settings.
+   - **Assistant name** is the name of your chat assistant. Each assistant corresponds to a dialogue with a unique combination of datasets, prompts, hybrid search configurations, and large model settings.
    - **Empty response**:
-     - If you wish to *confine* RAGFlow's answers to your knowledge bases, leave a response here. Then, when it doesn't retrieve an answer, it *uniformly* responds with what you set here.
-     - If you wish RAGFlow to *improvise* when it doesn't retrieve an answer from your knowledge bases, leave it blank, which may give rise to hallucinations.
+     - If you wish to *confine* RAGFlow's answers to your datasets, leave a response here. Then, when it doesn't retrieve an answer, it *uniformly* responds with what you set here.
+     - If you wish RAGFlow to *improvise* when it doesn't retrieve an answer from your datasets, leave it blank, which may give rise to hallucinations.
    - **Show quote**: This is a key feature of RAGFlow and enabled by default. RAGFlow does not work like a black box. Instead, it clearly shows the sources of information that its responses are based on.
-   - Select the corresponding knowledge bases. You can select one or multiple knowledge bases, but ensure that they use the same embedding model, otherwise an error would occur.
+   - Select the corresponding datasets. You can select one or multiple datasets, but ensure that they use the same embedding model, otherwise an error would occur.
 
 3. Update **Prompt engine**:
 
@@ -37,14 +37,14 @@ You start an AI conversation by creating an assistant.
      - If **Rerank model** is selected, the hybrid score system uses keyword similarity and reranker score, and the default weight assigned to the reranker score is 1-0.7=0.3.
    - **Top N** determines the *maximum* number of chunks to feed to the LLM. In other words, even if more chunks are retrieved, only the top N chunks are provided as input.
    - **Multi-turn optimization** enhances user queries using existing context in a multi-round conversation. It is enabled by default. When enabled, it will consume additional LLM tokens and significantly increase the time to generate answers.
-   - **Use knowledge graph** indicates whether to use knowledge graph(s) in the specified knowledge base(s) during retrieval for multi-hop question answering. When enabled, this would involve iterative searches across entity, relationship, and community report chunks, greatly increasing retrieval time.
+   - **Use knowledge graph** indicates whether to use knowledge graph(s) in the specified dataset(s) during retrieval for multi-hop question answering. When enabled, this would involve iterative searches across entity, relationship, and community report chunks, greatly increasing retrieval time.
    - **Reasoning** indicates whether to generate answers through reasoning processes like Deepseek-R1/OpenAI o1. Once enabled, the chat model autonomously integrates Deep Research during question answering when encountering an unknown topic. This involves the chat model dynamically searching external knowledge and generating final answers through reasoning.
    - **Rerank model** sets the reranker model to use. It is left empty by default.
      - If **Rerank model** is left empty, the hybrid score system uses keyword similarity and vector similarity, and the default weight assigned to the vector similarity component is 1-0.7=0.3.
      - If **Rerank model** is selected, the hybrid score system uses keyword similarity and reranker score, and the default weight assigned to the reranker score is 1-0.7=0.3.
    - [Cross-language search](../../references/glossary.mdx#cross-language-search): Optional  
      Select one or more target languages from the dropdown menu. The system’s default chat model will then translate your query into the selected target language(s). This translation ensures accurate semantic matching across languages, allowing you to retrieve relevant results regardless of language differences.  
-     - When selecting target languages, please ensure that these languages are present in the knowledge base to guarantee an effective search.
+     - When selecting target languages, please ensure that these languages are present in the dataset to guarantee an effective search.
      - If no target language is selected, the system will search only in the language of your query, which may cause relevant information in other languages to be missed.
    - **Variable** refers to the variables (keys) to be used in the system prompt. `{knowledge}` is a reserved variable. Click **Add** to add more variables for the system prompt.
       - If you are uncertain about the logic behind **Variable**, leave it *as-is*.

docs/guides/dataset/_category_.json

@@ -3,6 +3,6 @@
   "position": 0,
   "link": {
     "type": "generated-index",
-    "description": "Guides on configuring a knowledge base."
+    "description": "Guides on configuring a dataset."
   }
 }

docs/guides/dataset/autokeyword_autoquestion.mdx

@@ -6,7 +6,7 @@ slug: /autokeyword_autoquestion
 # Auto-keyword Auto-question
 import APITable from '@site/src/components/APITable';
 
-Use a chat model to generate keywords or questions from each chunk in the knowledge base.
+Use a chat model to generate keywords or questions from each chunk in the dataset.
 
 ---
 
@@ -18,7 +18,7 @@ Enabling this feature increases document indexing time and uses extra tokens, as
 
 ## What is Auto-keyword?
 
-Auto-keyword refers to the auto-keyword generation feature of RAGFlow. It uses a chat model to generate a set of keywords or synonyms from each chunk to correct errors and enhance retrieval accuracy. This feature is implemented as a slider under **Page rank** on the **Configuration** page of your knowledge base.
+Auto-keyword refers to the auto-keyword generation feature of RAGFlow. It uses a chat model to generate a set of keywords or synonyms from each chunk to correct errors and enhance retrieval accuracy. This feature is implemented as a slider under **Page rank** on the **Configuration** page of your dataset.
 
 **Values**:  
 
@@ -33,7 +33,7 @@ Auto-keyword refers to the auto-keyword generation feature of RAGFlow. It uses a
 
 ## What is Auto-question?
 
-Auto-question is a feature of RAGFlow that automatically generates questions from chunks of data using a chat model. These questions (e.g. who, what, and why) also help correct errors and improve the matching of user queries. The feature usually works with FAQ retrieval scenarios involving product manuals or policy documents. And you can find this feature as a slider under **Page rank** on the **Configuration** page of your knowledge base.
+Auto-question is a feature of RAGFlow that automatically generates questions from chunks of data using a chat model. These questions (e.g. who, what, and why) also help correct errors and improve the matching of user queries. The feature usually works with FAQ retrieval scenarios involving product manuals or policy documents. And you can find this feature as a slider under **Page rank** on the **Configuration** page of your dataset.
 
 **Values**:
 
@@ -48,7 +48,7 @@ Auto-question is a feature of RAGFlow that automatically generates questions fro
 
 ## Tips from the community
 
-The Auto-keyword or Auto-question values relate closely to the chunking size in your knowledge base. However, if you are new to this feature and unsure which value(s) to start with, the following are some value settings we gathered from our community. While they may not be accurate, they provide a starting point at the very least.
+The Auto-keyword or Auto-question values relate closely to the chunking size in your dataset. However, if you are new to this feature and unsure which value(s) to start with, the following are some value settings we gathered from our community. While they may not be accurate, they provide a starting point at the very least.
 
 ```mdx-code-block
 <APITable>

docs/guides/dataset/best_practices/_category_.json

@@ -3,6 +3,6 @@
   "position": 11,
   "link": {
     "type": "generated-index",
-    "description": "Best practices on configuring a knowledge base."
+    "description": "Best practices on configuring a dataset."
   }
 }

docs/guides/dataset/best_practices/accelerate_doc_indexing.mdx

@@ -13,7 +13,7 @@ A checklist to speed up document parsing and indexing.
 Please note that some of your settings may consume a significant amount of time. If you often find that document parsing is time-consuming, here is a checklist to consider:
 
 - Use GPU to reduce embedding time.
-- On the configuration page of your knowledge base, switch off **Use RAPTOR to enhance retrieval**.
+- On the configuration page of your dataset, switch off **Use RAPTOR to enhance retrieval**.
 - Extracting knowledge graph (GraphRAG) is time-consuming.
-- Disable **Auto-keyword** and **Auto-question** on the configuration page of your knowledge base, as both depend on the LLM.
-- **v0.17.0+:** If all PDFs in your knowledge base are plain text and do not require GPU-intensive processes like OCR (Optical Character Recognition), TSR (Table Structure Recognition), or DLA (Document Layout Analysis), you can choose **Naive** over **DeepDoc** or other time-consuming large model options in the **Document parser** dropdown. This will substantially reduce document parsing time.
+- Disable **Auto-keyword** and **Auto-question** on the configuration page of your dataset, as both depend on the LLM.
+- **v0.17.0+:** If all PDFs in your dataset are plain text and do not require GPU-intensive processes like OCR (Optical Character Recognition), TSR (Table Structure Recognition), or DLA (Document Layout Analysis), you can choose **Naive** over **DeepDoc** or other time-consuming large model options in the **Document parser** dropdown. This will substantially reduce document parsing time.

docs/guides/dataset/configure_knowledge_base.md

@@ -3,28 +3,28 @@ sidebar_position: -1
 slug: /configure_knowledge_base
 ---
 
-# Configure knowledge base
+# Configure dataset
 
-Knowledge base, hallucination-free chat, and file management are the three pillars of RAGFlow. RAGFlow's AI chats are based on knowledge bases. Each of RAGFlow's knowledge bases serves as a knowledge source, *parsing* files uploaded from your local machine and file references generated in **File Management** into the real 'knowledge' for future AI chats. This guide demonstrates some basic usages of the knowledge base feature, covering the following topics:
+Most of RAGFlow's chat assistants and Agents are based on datasets. Each of RAGFlow's datasets serves as a knowledge source, *parsing* files uploaded from your local machine and file references generated in **File Management** into the real 'knowledge' for future AI chats. This guide demonstrates some basic usages of the dataset feature, covering the following topics:
 
-- Create a knowledge base
-- Configure a knowledge base
-- Search for a knowledge base
-- Delete a knowledge base
+- Create a dataset
+- Configure a dataset
+- Search for a dataset
+- Delete a dataset
 
-## Create knowledge base
+## Create dataset
 
-With multiple knowledge bases, you can build more flexible, diversified question answering. To create your first knowledge base:
+With multiple datasets, you can build more flexible, diversified question answering. To create your first dataset:
 
-![create knowledge base](https://raw.githubusercontent.com/infiniflow/ragflow-docs/main/images/create_knowledge_base.jpg)
+![create dataset](https://raw.githubusercontent.com/infiniflow/ragflow-docs/main/images/create_knowledge_base.jpg)
 
-_Each time a knowledge base is created, a folder with the same name is generated in the **root/.knowledgebase** directory._
+_Each time a dataset is created, a folder with the same name is generated in the **root/.knowledgebase** directory._
 
-## Configure knowledge base
+## Configure dataset
 
-The following screenshot shows the configuration page of a knowledge base. A proper configuration of your knowledge base is crucial for future AI chats. For example, choosing the wrong embedding model or chunking method would cause unexpected semantic loss or mismatched answers in chats. 
+The following screenshot shows the configuration page of a dataset. A proper configuration of your dataset is crucial for future AI chats. For example, choosing the wrong embedding model or chunking method would cause unexpected semantic loss or mismatched answers in chats. 
 
-![knowledge base configuration](https://raw.githubusercontent.com/infiniflow/ragflow-docs/main/images/configure_knowledge_base.jpg)
+![dataset configuration](https://raw.githubusercontent.com/infiniflow/ragflow-docs/main/images/configure_knowledge_base.jpg)
 
 This section covers the following topics:
 
@@ -52,7 +52,7 @@ RAGFlow offers multiple chunking template to facilitate chunking files of differ
 | Presentation |                                                                       | PDF, PPTX                                                                                     |
 | Picture      |                                                                       | JPEG, JPG, PNG, TIF, GIF                                                                      |
 | One          | Each document is chunked in its entirety (as one).                    | DOCX, XLSX, XLS (Excel 97-2003), PDF, TXT                                                      |
-| Tag          | The knowledge base functions as a tag set for the others.             | XLSX, CSV/TXT                                                                                 |
+| Tag          | The dataset functions as a tag set for the others.             | XLSX, CSV/TXT                                                                                 |
 
 You can also change a file's chunking method on the **Datasets** page.
 
@@ -60,7 +60,7 @@ You can also change a file's chunking method on the **Datasets** page.
 
 ### Select embedding model
 
-An embedding model converts chunks into embeddings. It cannot be changed once the knowledge base has chunks. To switch to a different embedding model, you must delete all existing chunks in the knowledge base. The obvious reason is that we *must* ensure that files in a specific knowledge base are converted to embeddings using the *same* embedding model (ensure that they are compared in the same embedding space).
+An embedding model converts chunks into embeddings. It cannot be changed once the dataset has chunks. To switch to a different embedding model, you must delete all existing chunks in the dataset. The obvious reason is that we *must* ensure that files in a specific dataset are converted to embeddings using the *same* embedding model (ensure that they are compared in the same embedding space).
 
 The following embedding models can be deployed locally:
 
@@ -73,19 +73,19 @@ These two embedding models are optimized specifically for English and Chinese, s
 
 ### Upload file
 
-- RAGFlow's **File Management** allows you to link a file to multiple knowledge bases, in which case each target knowledge base holds a reference to the file.
-- In **Knowledge Base**, you are also given the option of uploading a single file or a folder of files (bulk upload) from your local machine to a knowledge base, in which case the knowledge base holds file copies. 
+- RAGFlow's **File Management** allows you to link a file to multiple datasets, in which case each target dataset holds a reference to the file.
+- In **Knowledge Base**, you are also given the option of uploading a single file or a folder of files (bulk upload) from your local machine to a dataset, in which case the dataset holds file copies. 
 
-While uploading files directly to a knowledge base seems more convenient, we *highly* recommend uploading files to **File Management** and then linking them to the target knowledge bases. This way, you can avoid permanently deleting files uploaded to the knowledge base. 
+While uploading files directly to a dataset seems more convenient, we *highly* recommend uploading files to **File Management** and then linking them to the target datasets. This way, you can avoid permanently deleting files uploaded to the dataset. 
 
 ### Parse file
 
-File parsing is a crucial topic in knowledge base configuration. The meaning of file parsing in RAGFlow is twofold: chunking files based on file layout and building embedding and full-text (keyword) indexes on these chunks. After having selected the chunking method and embedding model, you can start parsing a file:
+File parsing is a crucial topic in dataset configuration. The meaning of file parsing in RAGFlow is twofold: chunking files based on file layout and building embedding and full-text (keyword) indexes on these chunks. After having selected the chunking method and embedding model, you can start parsing a file:
 
 ![parse file](https://raw.githubusercontent.com/infiniflow/ragflow-docs/main/images/parse_file.jpg)
 
 - As shown above, RAGFlow allows you to use a different chunking method for a particular file, offering flexibility beyond the default method. 
-- As shown above, RAGFlow allows you to enable or disable individual files, offering finer control over knowledge base-based AI chats. 
+- As shown above, RAGFlow allows you to enable or disable individual files, offering finer control over dataset-based AI chats. 
 
 ### Intervene with file parsing results
 
@@ -122,17 +122,17 @@ RAGFlow uses multiple recall of both full-text search and vector search in its c
 
 See [Run retrieval test](./run_retrieval_test.md) for details.
 
-## Search for knowledge base
+## Search for dataset
 
-As of RAGFlow v0.20.5, the search feature is still in a rudimentary form, supporting only knowledge base search by name.
+As of RAGFlow v0.20.5, the search feature is still in a rudimentary form, supporting only dataset search by name.
 
-![search knowledge base](https://raw.githubusercontent.com/infiniflow/ragflow-docs/main/images/search_datasets.jpg)
+![search dataset](https://raw.githubusercontent.com/infiniflow/ragflow-docs/main/images/search_datasets.jpg)
 
-## Delete knowledge base
+## Delete dataset
 
-You are allowed to delete a knowledge base. Hover your mouse over the three dot of the intended knowledge base card and the **Delete** option appears. Once you delete a knowledge base, the associated folder under **root/.knowledge** directory is AUTOMATICALLY REMOVED. The consequence is:
+You are allowed to delete a dataset. Hover your mouse over the three dot of the intended dataset card and the **Delete** option appears. Once you delete a dataset, the associated folder under **root/.knowledge** directory is AUTOMATICALLY REMOVED. The consequence is:
 
-- The files uploaded directly to the knowledge base are gone;  
+- The files uploaded directly to the dataset are gone;  
 - The file references, which you created from within **File Management**, are gone, but the associated files still exist in **File Management**. 
 
-![delete knowledge base](https://raw.githubusercontent.com/infiniflow/ragflow-docs/main/images/delete_datasets.jpg)
+![delete dataset](https://raw.githubusercontent.com/infiniflow/ragflow-docs/main/images/delete_datasets.jpg)