Merge branch 'main' into feat/mcp

.github/workflows/deploy-rag-dev.yml

@@ -0,0 +1,28 @@
+name: Deploy RAG Dev
+
+permissions:
+  contents: read
+
+on:
+  workflow_run:
+    workflows: ["Build and Push API & Web"]
+    branches:
+      - "deploy/rag-dev"
+    types:
+      - completed
+
+jobs:
+  deploy:
+    runs-on: ubuntu-latest
+    if: |
+      github.event.workflow_run.conclusion == 'success' &&
+      github.event.workflow_run.head_branch == 'deploy/rag-dev'
+    steps:
+      - name: Deploy to server
+        uses: appleboy/ssh-action@v0.1.8
+        with:
+          host: ${{ secrets.RAG_SSH_HOST }}
+          username: ${{ secrets.SSH_USER }}
+          key: ${{ secrets.SSH_PRIVATE_KEY }}
+          script: |
+            ${{ vars.SSH_SCRIPT || secrets.SSH_SCRIPT }}

.github/workflows/expose_service_ports.sh

@@ -10,6 +10,7 @@ yq eval '.services["elasticsearch"].ports += ["9200:9200"]' -i docker/docker-com
 yq eval '.services.couchbase-server.ports += ["8091-8096:8091-8096"]' -i docker/docker-compose.yaml
 yq eval '.services.couchbase-server.ports += ["11210:11210"]' -i docker/docker-compose.yaml
 yq eval '.services.tidb.ports += ["4000:4000"]' -i docker/tidb/docker-compose.yaml
+yq eval '.services.oceanbase.ports += ["2881:2881"]' -i docker/docker-compose.yaml
 yq eval '.services.opengauss.ports += ["6600:6600"]' -i docker/docker-compose.yaml
 
 echo "Ports exposed for sandbox, weaviate, tidb, qdrant, chroma, milvus, pgvector, pgvecto-rs, elasticsearch, couchbase, opengauss"

.github/workflows/vdb-tests.yml

@@ -31,6 +31,13 @@ jobs:
         with:
           persist-credentials: false
 
+      - name: Free Disk Space
+        uses: endersonmenezes/free-disk-space@v2
+        with:
+          remove_dotnet: true
+          remove_haskell: true
+          remove_tool_cache: true
+
       - name: Setup UV and Python
         uses: ./.github/actions/setup-uv
         with:
@@ -59,7 +66,7 @@ jobs:
             tidb
             tiflash
 
-      - name: Set up Vector Stores (Weaviate, Qdrant, PGVector, Milvus, PgVecto-RS, Chroma, MyScale, ElasticSearch, Couchbase)
+      - name: Set up Vector Stores (Weaviate, Qdrant, PGVector, Milvus, PgVecto-RS, Chroma, MyScale, ElasticSearch, Couchbase, OceanBase)
         uses: hoverkraft-tech/compose-action@v2.0.2
         with:
           compose-file: |
@@ -75,8 +82,9 @@ jobs:
             pgvector
             chroma
             elasticsearch
+            oceanbase
 
-      - name: Check TiDB Ready
+      - name: Check VDB Ready (TiDB)
         run: uv run --project api python api/tests/integration_tests/vdb/tidb_vector/check_tiflash_ready.py
 
       - name: Test Vector Stores

.gitignore

@@ -179,6 +179,7 @@ docker/volumes/pgvecto_rs/data/*
 docker/volumes/couchbase/*
 docker/volumes/oceanbase/*
 docker/volumes/plugin_daemon/*
+docker/volumes/matrixone/*
 !docker/volumes/oceanbase/init.d
 
 docker/nginx/conf.d/default.conf
@@ -210,3 +211,6 @@ mise.toml
 
 # Next.js build output
 .next/
+
+# AI Assistant
+.roo/

api/.env.example

@@ -137,7 +137,7 @@ WEB_API_CORS_ALLOW_ORIGINS=http://127.0.0.1:3000,*
 CONSOLE_CORS_ALLOW_ORIGINS=http://127.0.0.1:3000,*
 
 # Vector database configuration
-# support: weaviate, qdrant, milvus, myscale, relyt, pgvecto_rs, pgvector, pgvector, chroma, opensearch, tidb_vector, couchbase, vikingdb, upstash, lindorm, oceanbase, opengauss, tablestore
+# support: weaviate, qdrant, milvus, myscale, relyt, pgvecto_rs, pgvector, pgvector, chroma, opensearch, tidb_vector, couchbase, vikingdb, upstash, lindorm, oceanbase, opengauss, tablestore, matrixone
 VECTOR_STORE=weaviate
 
 # Weaviate configuration
@@ -294,6 +294,13 @@ VIKINGDB_SCHEMA=http
 VIKINGDB_CONNECTION_TIMEOUT=30
 VIKINGDB_SOCKET_TIMEOUT=30
 
+# Matrixone configration
+MATRIXONE_HOST=127.0.0.1
+MATRIXONE_PORT=6001
+MATRIXONE_USER=dump
+MATRIXONE_PASSWORD=111
+MATRIXONE_DATABASE=dify
+
 # Lindorm configuration
 LINDORM_URL=http://ld-*******************-proxy-search-pub.lindorm.aliyuncs.com:30070
 LINDORM_USERNAME=admin
@@ -332,9 +339,11 @@ PROMPT_GENERATION_MAX_TOKENS=512
 CODE_GENERATION_MAX_TOKENS=1024
 PLUGIN_BASED_TOKEN_COUNTING_ENABLED=false
 
-# Mail configuration, support: resend, smtp
+# Mail configuration, support: resend, smtp, sendgrid
 MAIL_TYPE=
+# If using SendGrid, use the 'from' field for authentication if necessary.
 MAIL_DEFAULT_SEND_FROM=no-reply <no-reply@dify.ai>
+# resend configuration
 RESEND_API_KEY=
 RESEND_API_URL=https://api.resend.com
 # smtp configuration
@@ -344,7 +353,8 @@ SMTP_USERNAME=123
 SMTP_PASSWORD=abc
 SMTP_USE_TLS=true
 SMTP_OPPORTUNISTIC_TLS=false
-
+# Sendgid configuration
+SENDGRID_API_KEY=
 # Sentry configuration
 SENTRY_DSN=
 

api/commands.py

@@ -27,7 +27,7 @@ from models.dataset import Dataset, DatasetCollectionBinding, DatasetMetadata, D
 from models.dataset import Document as DatasetDocument
 from models.model import Account, App, AppAnnotationSetting, AppMode, Conversation, MessageAnnotation
 from models.provider import Provider, ProviderModel
-from services.account_service import RegisterService, TenantService
+from services.account_service import AccountService, RegisterService, TenantService
 from services.clear_free_plan_tenant_expired_logs import ClearFreePlanTenantExpiredLogs
 from services.plugin.data_migration import PluginDataMigration
 from services.plugin.plugin_migration import PluginMigration
@@ -68,6 +68,7 @@ def reset_password(email, new_password, password_confirm):
     account.password = base64_password_hashed
     account.password_salt = base64_salt
     db.session.commit()
+    AccountService.reset_login_error_rate_limit(email)
     click.echo(click.style("Password reset successfully.", fg="green"))
 
 
@@ -280,6 +281,7 @@ def migrate_knowledge_vector_database():
         VectorType.ELASTICSEARCH,
         VectorType.OPENGAUSS,
         VectorType.TABLESTORE,
+        VectorType.MATRIXONE,
     }
     lower_collection_vector_types = {
         VectorType.ANALYTICDB,

api/configs/feature/__init__.py

@@ -609,7 +609,7 @@ class MailConfig(BaseSettings):
     """
 
     MAIL_TYPE: Optional[str] = Field(
-        description="Email service provider type ('smtp' or 'resend'), default to None.",
+        description="Email service provider type ('smtp' or 'resend' or 'sendGrid), default to None.",
         default=None,
     )
 
@@ -663,6 +663,11 @@ class MailConfig(BaseSettings):
         default=50,
     )
 
+    SENDGRID_API_KEY: Optional[str] = Field(
+        description="API key for SendGrid service",
+        default=None,
+    )
+
 
 class RagEtlConfig(BaseSettings):
     """

api/configs/middleware/__init__.py

@@ -24,6 +24,7 @@ from .vdb.couchbase_config import CouchbaseConfig
 from .vdb.elasticsearch_config import ElasticsearchConfig
 from .vdb.huawei_cloud_config import HuaweiCloudConfig
 from .vdb.lindorm_config import LindormConfig
+from .vdb.matrixone_config import MatrixoneConfig
 from .vdb.milvus_config import MilvusConfig
 from .vdb.myscale_config import MyScaleConfig
 from .vdb.oceanbase_config import OceanBaseVectorConfig
@@ -323,5 +324,6 @@ class MiddlewareConfig(
     OpenGaussConfig,
     TableStoreConfig,
     DatasetQueueMonitorConfig,
+    MatrixoneConfig,
 ):
     pass

api/configs/middleware/vdb/matrixone_config.py

@@ -0,0 +1,14 @@
+from pydantic import BaseModel, Field
+
+
+class MatrixoneConfig(BaseModel):
+    """Matrixone vector database configuration."""
+
+    MATRIXONE_HOST: str = Field(default="localhost", description="Host address of the Matrixone server")
+    MATRIXONE_PORT: int = Field(default=6001, description="Port number of the Matrixone server")
+    MATRIXONE_USER: str = Field(default="dump", description="Username for authenticating with Matrixone")
+    MATRIXONE_PASSWORD: str = Field(default="111", description="Password for authenticating with Matrixone")
+    MATRIXONE_DATABASE: str = Field(default="dify", description="Name of the Matrixone database to connect to")
+    MATRIXONE_METRIC: str = Field(
+        default="l2", description="Distance metric type for vector similarity search (cosine or l2)"
+    )

api/configs/packaging/__init__.py

@@ -9,7 +9,7 @@ class PackagingInfo(BaseSettings):
 
     CURRENT_VERSION: str = Field(
         description="Dify version",
-        default="1.4.2",
+        default="1.4.3",
     )
 
     COMMIT_SHA: str = Field(

api/controllers/console/app/annotation.py

@@ -208,7 +208,7 @@ class AnnotationBatchImportApi(Resource):
         if len(request.files) > 1:
             raise TooManyFilesError()
         # check file type
-        if not file.filename or not file.filename.endswith(".csv"):
+        if not file.filename or not file.filename.lower().endswith(".csv"):
             raise ValueError("Invalid file type. Only CSV files are allowed")
         return AppAnnotationService.batch_import_app_annotations(app_id, file)
 

api/controllers/console/app/workflow_app_log.py

@@ -34,6 +34,20 @@ class WorkflowAppLogApi(Resource):
         parser.add_argument(
             "created_at__after", type=str, location="args", help="Filter logs created after this timestamp"
         )
+        parser.add_argument(
+            "created_by_end_user_session_id",
+            type=str,
+            location="args",
+            required=False,
+            default=None,
+        )
+        parser.add_argument(
+            "created_by_account",
+            type=str,
+            location="args",
+            required=False,
+            default=None,
+        )
         parser.add_argument("page", type=int_range(1, 99999), default=1, location="args")
         parser.add_argument("limit", type=int_range(1, 100), default=20, location="args")
         args = parser.parse_args()
@@ -57,6 +71,8 @@ class WorkflowAppLogApi(Resource):
                 created_at_after=args.created_at__after,
                 page=args.page,
                 limit=args.limit,
+                created_by_end_user_session_id=args.created_by_end_user_session_id,
+                created_by_account=args.created_by_account,
             )
 
             return workflow_app_log_pagination

api/controllers/console/datasets/datasets.py

@@ -686,6 +686,7 @@ class DatasetRetrievalSettingApi(Resource):
                 | VectorType.TABLESTORE
                 | VectorType.HUAWEI_CLOUD
                 | VectorType.TENCENT
+                | VectorType.MATRIXONE
             ):
                 return {
                     "retrieval_method": [
@@ -733,6 +734,7 @@ class DatasetRetrievalSettingMockApi(Resource):
                 | VectorType.TABLESTORE
                 | VectorType.TENCENT
                 | VectorType.HUAWEI_CLOUD
+                | VectorType.MATRIXONE
             ):
                 return {
                     "retrieval_method": [

api/controllers/console/datasets/datasets_segments.py

@@ -374,7 +374,7 @@ class DatasetDocumentSegmentBatchImportApi(Resource):
         if len(request.files) > 1:
             raise TooManyFilesError()
         # check file type
-        if not file.filename or not file.filename.endswith(".csv"):
+        if not file.filename or not file.filename.lower().endswith(".csv"):
             raise ValueError("Invalid file type. Only CSV files are allowed")
 
         try:

api/controllers/console/workspace/load_balancing_config.py

@@ -15,7 +15,7 @@ class LoadBalancingCredentialsValidateApi(Resource):
     @login_required
     @account_initialization_required
     def post(self, provider: str):
-        if not TenantAccountRole.is_privileged_role(current_user.current_tenant.current_role):
+        if not TenantAccountRole.is_privileged_role(current_user.current_role):
             raise Forbidden()
 
         tenant_id = current_user.current_tenant_id
@@ -64,7 +64,7 @@ class LoadBalancingConfigCredentialsValidateApi(Resource):
     @login_required
     @account_initialization_required
     def post(self, provider: str, config_id: str):
-        if not TenantAccountRole.is_privileged_role(current_user.current_tenant.current_role):
+        if not TenantAccountRole.is_privileged_role(current_user.current_role):
             raise Forbidden()
 
         tenant_id = current_user.current_tenant_id

api/controllers/service_api/app/app.py

@@ -47,7 +47,13 @@ class AppInfoApi(Resource):
     def get(self, app_model: App):
         """Get app information"""
         tags = [tag.name for tag in app_model.tags]
-        return {"name": app_model.name, "description": app_model.description, "tags": tags, "mode": app_model.mode}
+        return {
+            "name": app_model.name,
+            "description": app_model.description,
+            "tags": tags,
+            "mode": app_model.mode,
+            "author_name": app_model.author_name,
+        }
 
 
 api.add_resource(AppParameterApi, "/parameters")

api/controllers/service_api/app/workflow.py

@@ -135,6 +135,20 @@ class WorkflowAppLogApi(Resource):
         parser.add_argument("status", type=str, choices=["succeeded", "failed", "stopped"], location="args")
         parser.add_argument("created_at__before", type=str, location="args")
         parser.add_argument("created_at__after", type=str, location="args")
+        parser.add_argument(
+            "created_by_end_user_session_id",
+            type=str,
+            location="args",
+            required=False,
+            default=None,
+        )
+        parser.add_argument(
+            "created_by_account",
+            type=str,
+            location="args",
+            required=False,
+            default=None,
+        )
         parser.add_argument("page", type=int_range(1, 99999), default=1, location="args")
         parser.add_argument("limit", type=int_range(1, 100), default=20, location="args")
         args = parser.parse_args()
@@ -158,6 +172,8 @@ class WorkflowAppLogApi(Resource):
                 created_at_after=args.created_at__after,
                 page=args.page,
                 limit=args.limit,
+                created_by_end_user_session_id=args.created_by_end_user_session_id,
+                created_by_account=args.created_by_account,
             )
 
             return workflow_app_log_pagination

api/controllers/service_api/dataset/dataset.py

@@ -5,7 +5,11 @@ from werkzeug.exceptions import Forbidden, NotFound
 import services.dataset_service
 from controllers.service_api import api
 from controllers.service_api.dataset.error import DatasetInUseError, DatasetNameDuplicateError
-from controllers.service_api.wraps import DatasetApiResource, validate_dataset_token
+from controllers.service_api.wraps import (
+    DatasetApiResource,
+    cloud_edition_billing_rate_limit_check,
+    validate_dataset_token,
+)
 from core.model_runtime.entities.model_entities import ModelType
 from core.plugin.entities.plugin import ModelProviderID
 from core.provider_manager import ProviderManager
@@ -70,6 +74,7 @@ class DatasetListApi(DatasetApiResource):
         response = {"data": data, "has_more": len(datasets) == limit, "limit": limit, "total": total, "page": page}
         return response, 200
 
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def post(self, tenant_id):
         """Resource for creating datasets."""
         parser = reqparse.RequestParser()
@@ -193,6 +198,7 @@ class DatasetApi(DatasetApiResource):
 
         return data, 200
 
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def patch(self, _, dataset_id):
         dataset_id_str = str(dataset_id)
         dataset = DatasetService.get_dataset(dataset_id_str)
@@ -293,6 +299,7 @@ class DatasetApi(DatasetApiResource):
 
         return result_data, 200
 
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def delete(self, _, dataset_id):
         """
         Deletes a dataset given its ID.

api/controllers/service_api/dataset/document.py

@@ -19,7 +19,11 @@ from controllers.service_api.dataset.error import (
     ArchivedDocumentImmutableError,
     DocumentIndexingError,
 )
-from controllers.service_api.wraps import DatasetApiResource, cloud_edition_billing_resource_check
+from controllers.service_api.wraps import (
+    DatasetApiResource,
+    cloud_edition_billing_rate_limit_check,
+    cloud_edition_billing_resource_check,
+)
 from core.errors.error import ProviderTokenNotInitError
 from extensions.ext_database import db
 from fields.document_fields import document_fields, document_status_fields
@@ -35,6 +39,7 @@ class DocumentAddByTextApi(DatasetApiResource):
 
     @cloud_edition_billing_resource_check("vector_space", "dataset")
     @cloud_edition_billing_resource_check("documents", "dataset")
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def post(self, tenant_id, dataset_id):
         """Create document by text."""
         parser = reqparse.RequestParser()
@@ -99,6 +104,7 @@ class DocumentUpdateByTextApi(DatasetApiResource):
     """Resource for update documents."""
 
     @cloud_edition_billing_resource_check("vector_space", "dataset")
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def post(self, tenant_id, dataset_id, document_id):
         """Update document by text."""
         parser = reqparse.RequestParser()
@@ -158,6 +164,7 @@ class DocumentAddByFileApi(DatasetApiResource):
 
     @cloud_edition_billing_resource_check("vector_space", "dataset")
     @cloud_edition_billing_resource_check("documents", "dataset")
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def post(self, tenant_id, dataset_id):
         """Create document by upload file."""
         args = {}
@@ -232,6 +239,7 @@ class DocumentUpdateByFileApi(DatasetApiResource):
     """Resource for update documents."""
 
     @cloud_edition_billing_resource_check("vector_space", "dataset")
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def post(self, tenant_id, dataset_id, document_id):
         """Update document by upload file."""
         args = {}
@@ -302,6 +310,7 @@ class DocumentUpdateByFileApi(DatasetApiResource):
 
 
 class DocumentDeleteApi(DatasetApiResource):
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def delete(self, tenant_id, dataset_id, document_id):
         """Delete document."""
         document_id = str(document_id)

api/controllers/service_api/dataset/hit_testing.py

@@ -1,9 +1,10 @@
 from controllers.console.datasets.hit_testing_base import DatasetsHitTestingBase
 from controllers.service_api import api
-from controllers.service_api.wraps import DatasetApiResource
+from controllers.service_api.wraps import DatasetApiResource, cloud_edition_billing_rate_limit_check
 
 
 class HitTestingApi(DatasetApiResource, DatasetsHitTestingBase):
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def post(self, tenant_id, dataset_id):
         dataset_id_str = str(dataset_id)
 

api/controllers/service_api/dataset/metadata.py

@@ -3,7 +3,7 @@ from flask_restful import marshal, reqparse
 from werkzeug.exceptions import NotFound
 
 from controllers.service_api import api
-from controllers.service_api.wraps import DatasetApiResource
+from controllers.service_api.wraps import DatasetApiResource, cloud_edition_billing_rate_limit_check
 from fields.dataset_fields import dataset_metadata_fields
 from services.dataset_service import DatasetService
 from services.entities.knowledge_entities.knowledge_entities import (
@@ -14,6 +14,7 @@ from services.metadata_service import MetadataService
 
 
 class DatasetMetadataCreateServiceApi(DatasetApiResource):
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def post(self, tenant_id, dataset_id):
         parser = reqparse.RequestParser()
         parser.add_argument("type", type=str, required=True, nullable=True, location="json")
@@ -39,6 +40,7 @@ class DatasetMetadataCreateServiceApi(DatasetApiResource):
 
 
 class DatasetMetadataServiceApi(DatasetApiResource):
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def patch(self, tenant_id, dataset_id, metadata_id):
         parser = reqparse.RequestParser()
         parser.add_argument("name", type=str, required=True, nullable=True, location="json")
@@ -54,6 +56,7 @@ class DatasetMetadataServiceApi(DatasetApiResource):
         metadata = MetadataService.update_metadata_name(dataset_id_str, metadata_id_str, args.get("name"))
         return marshal(metadata, dataset_metadata_fields), 200
 
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def delete(self, tenant_id, dataset_id, metadata_id):
         dataset_id_str = str(dataset_id)
         metadata_id_str = str(metadata_id)
@@ -73,6 +76,7 @@ class DatasetMetadataBuiltInFieldServiceApi(DatasetApiResource):
 
 
 class DatasetMetadataBuiltInFieldActionServiceApi(DatasetApiResource):
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def post(self, tenant_id, dataset_id, action):
         dataset_id_str = str(dataset_id)
         dataset = DatasetService.get_dataset(dataset_id_str)
@@ -88,6 +92,7 @@ class DatasetMetadataBuiltInFieldActionServiceApi(DatasetApiResource):
 
 
 class DocumentMetadataEditServiceApi(DatasetApiResource):
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def post(self, tenant_id, dataset_id):
         dataset_id_str = str(dataset_id)
         dataset = DatasetService.get_dataset(dataset_id_str)

api/controllers/service_api/dataset/segment.py

@@ -8,6 +8,7 @@ from controllers.service_api.app.error import ProviderNotInitializeError
 from controllers.service_api.wraps import (
     DatasetApiResource,
     cloud_edition_billing_knowledge_limit_check,
+    cloud_edition_billing_rate_limit_check,
     cloud_edition_billing_resource_check,
 )
 from core.errors.error import LLMBadRequestError, ProviderTokenNotInitError
@@ -35,6 +36,7 @@ class SegmentApi(DatasetApiResource):
 
     @cloud_edition_billing_resource_check("vector_space", "dataset")
     @cloud_edition_billing_knowledge_limit_check("add_segment", "dataset")
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def post(self, tenant_id, dataset_id, document_id):
         """Create single segment."""
         # check dataset
@@ -139,6 +141,7 @@ class SegmentApi(DatasetApiResource):
 
 
 class DatasetSegmentApi(DatasetApiResource):
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def delete(self, tenant_id, dataset_id, document_id, segment_id):
         # check dataset
         dataset_id = str(dataset_id)
@@ -162,6 +165,7 @@ class DatasetSegmentApi(DatasetApiResource):
         return 204
 
     @cloud_edition_billing_resource_check("vector_space", "dataset")
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def post(self, tenant_id, dataset_id, document_id, segment_id):
         # check dataset
         dataset_id = str(dataset_id)
@@ -236,6 +240,7 @@ class ChildChunkApi(DatasetApiResource):
 
     @cloud_edition_billing_resource_check("vector_space", "dataset")
     @cloud_edition_billing_knowledge_limit_check("add_segment", "dataset")
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def post(self, tenant_id, dataset_id, document_id, segment_id):
         """Create child chunk."""
         # check dataset
@@ -332,6 +337,7 @@ class DatasetChildChunkApi(DatasetApiResource):
     """Resource for updating child chunks."""
 
     @cloud_edition_billing_knowledge_limit_check("add_segment", "dataset")
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def delete(self, tenant_id, dataset_id, document_id, segment_id, child_chunk_id):
         """Delete child chunk."""
         # check dataset
@@ -370,6 +376,7 @@ class DatasetChildChunkApi(DatasetApiResource):
 
     @cloud_edition_billing_resource_check("vector_space", "dataset")
     @cloud_edition_billing_knowledge_limit_check("add_segment", "dataset")
+    @cloud_edition_billing_rate_limit_check("knowledge", "dataset")
     def patch(self, tenant_id, dataset_id, document_id, segment_id, child_chunk_id):
         """Update child chunk."""
         # check dataset

api/controllers/web/passport.py

@@ -163,7 +163,7 @@ def exchange_token_for_existing_web_user(app_code: str, enterprise_user_decoded:
         )
         db.session.add(end_user)
         db.session.commit()
-    exp_dt = datetime.now(UTC) + timedelta(hours=dify_config.ACCESS_TOKEN_EXPIRE_MINUTES * 24)
+    exp_dt = datetime.now(UTC) + timedelta(minutes=dify_config.ACCESS_TOKEN_EXPIRE_MINUTES)
     exp = int(exp_dt.timestamp())
     payload = {
         "iss": site.id,

api/core/app/app_config/easy_ui_based_app/dataset/manager.py

@@ -138,15 +138,12 @@ class DatasetConfigManager:
         if not config.get("dataset_configs"):
             config["dataset_configs"] = {"retrieval_model": "single"}
 
+        if not isinstance(config["dataset_configs"], dict):
+            raise ValueError("dataset_configs must be of object type")
+
         if not config["dataset_configs"].get("datasets"):
             config["dataset_configs"]["datasets"] = {"strategy": "router", "datasets": []}
 
-        if not isinstance(config["dataset_configs"], dict):
-            raise ValueError("dataset_configs must be of object type")
-
-        if not isinstance(config["dataset_configs"], dict):
-            raise ValueError("dataset_configs must be of object type")
-
         need_manual_query_datasets = config.get("dataset_configs") and config["dataset_configs"].get(
             "datasets", {}
         ).get("datasets")

api/core/app/apps/advanced_chat/app_generator.py

@@ -5,7 +5,7 @@ import uuid
 from collections.abc import Generator, Mapping
 from typing import Any, Literal, Optional, Union, overload
 
-from flask import Flask, copy_current_request_context, current_app, has_request_context
+from flask import Flask, current_app
 from pydantic import ValidationError
 from sqlalchemy.orm import sessionmaker
 
@@ -31,6 +31,7 @@ from core.workflow.repositories.workflow_execution_repository import WorkflowExe
 from core.workflow.repositories.workflow_node_execution_repository import WorkflowNodeExecutionRepository
 from extensions.ext_database import db
 from factories import file_factory
+from libs.flask_utils import preserve_flask_contexts
 from models import Account, App, Conversation, EndUser, Message, Workflow, WorkflowNodeExecutionTriggeredFrom
 from models.enums import WorkflowRunTriggeredFrom
 from services.conversation_service import ConversationService
@@ -366,6 +367,7 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
         :param user: account or end user
         :param invoke_from: invoke from source
         :param application_generate_entity: application generate entity
+        :param workflow_execution_repository: repository for workflow execution
         :param workflow_node_execution_repository: repository for workflow node execution
         :param conversation: conversation
         :param stream: is stream
@@ -399,20 +401,17 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
         # new thread with request context and contextvars
         context = contextvars.copy_context()
 
-        @copy_current_request_context
-        def worker_with_context():
-            # Run the worker within the copied context
-            return context.run(
-                self._generate_worker,
-                flask_app=current_app._get_current_object(),  # type: ignore
-                application_generate_entity=application_generate_entity,
-                queue_manager=queue_manager,
-                conversation_id=conversation.id,
-                message_id=message.id,
-                context=context,
-            )
-
-        worker_thread = threading.Thread(target=worker_with_context)
+        worker_thread = threading.Thread(
+            target=self._generate_worker,
+            kwargs={
+                "flask_app": current_app._get_current_object(),  # type: ignore
+                "application_generate_entity": application_generate_entity,
+                "queue_manager": queue_manager,
+                "conversation_id": conversation.id,
+                "message_id": message.id,
+                "context": context,
+            },
+        )
 
         worker_thread.start()
 
@@ -449,24 +448,9 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
         :param message_id: message ID
         :return:
         """
-        for var, val in context.items():
-            var.set(val)
 
-        # FIXME(-LAN-): Save current user before entering new app context
-        from flask import g
-
-        saved_user = None
-        if has_request_context() and hasattr(g, "_login_user"):
-            saved_user = g._login_user
-
-        with flask_app.app_context():
+        with preserve_flask_contexts(flask_app, context_vars=context):
             try:
-                # Restore user in new app context
-                if saved_user is not None:
-                    from flask import g
-
-                    g._login_user = saved_user
-
                 # get conversation and message
                 conversation = self._get_conversation(conversation_id)
                 message = self._get_message(message_id)

api/core/app/apps/agent_chat/app_generator.py

@@ -5,7 +5,7 @@ import uuid
 from collections.abc import Generator, Mapping
 from typing import Any, Literal, Union, overload
 
-from flask import Flask, copy_current_request_context, current_app, has_request_context
+from flask import Flask, current_app
 from pydantic import ValidationError
 
 from configs import dify_config
@@ -23,6 +23,7 @@ from core.model_runtime.errors.invoke import InvokeAuthorizationError
 from core.ops.ops_trace_manager import TraceQueueManager
 from extensions.ext_database import db
 from factories import file_factory
+from libs.flask_utils import preserve_flask_contexts
 from models import Account, App, EndUser
 from services.conversation_service import ConversationService
 from services.errors.message import MessageNotExistsError
@@ -182,20 +183,17 @@ class AgentChatAppGenerator(MessageBasedAppGenerator):
         # new thread with request context and contextvars
         context = contextvars.copy_context()
 
-        @copy_current_request_context
-        def worker_with_context():
-            # Run the worker within the copied context
-            return context.run(
-                self._generate_worker,
-                flask_app=current_app._get_current_object(),  # type: ignore
-                context=context,
-                application_generate_entity=application_generate_entity,
-                queue_manager=queue_manager,
-                conversation_id=conversation.id,
-                message_id=message.id,
-            )
-
-        worker_thread = threading.Thread(target=worker_with_context)
+        worker_thread = threading.Thread(
+            target=self._generate_worker,
+            kwargs={
+                "flask_app": current_app._get_current_object(),  # type: ignore
+                "context": context,
+                "application_generate_entity": application_generate_entity,
+                "queue_manager": queue_manager,
+                "conversation_id": conversation.id,
+                "message_id": message.id,
+            },
+        )
 
         worker_thread.start()
 
@@ -229,24 +227,9 @@ class AgentChatAppGenerator(MessageBasedAppGenerator):
         :param message_id: message ID
         :return:
         """
-        for var, val in context.items():
-            var.set(val)
 
-        # FIXME(-LAN-): Save current user before entering new app context
-        from flask import g
-
-        saved_user = None
-        if has_request_context() and hasattr(g, "_login_user"):
-            saved_user = g._login_user
-
-        with flask_app.app_context():
+        with preserve_flask_contexts(flask_app, context_vars=context):
             try:
-                # Restore user in new app context
-                if saved_user is not None:
-                    from flask import g
-
-                    g._login_user = saved_user
-
                 # get conversation and message
                 conversation = self._get_conversation(conversation_id)
                 message = self._get_message(message_id)

api/core/app/apps/base_app_runner.py

@@ -1,3 +1,4 @@
+import logging
 import time
 from collections.abc import Generator, Mapping, Sequence
 from typing import TYPE_CHECKING, Any, Optional, Union
@@ -33,6 +34,8 @@ from models.model import App, AppMode, Message, MessageAnnotation
 if TYPE_CHECKING:
     from core.file.models import File
 
+_logger = logging.getLogger(__name__)
+
 
 class AppRunner:
     def get_pre_calculate_rest_tokens(
@@ -298,7 +301,7 @@ class AppRunner:
         )
 
     def _handle_invoke_result_stream(
-        self, invoke_result: Generator, queue_manager: AppQueueManager, agent: bool
+        self, invoke_result: Generator[LLMResultChunk, None, None], queue_manager: AppQueueManager, agent: bool
     ) -> None:
         """
         Handle invoke result
@@ -317,18 +320,28 @@ class AppRunner:
             else:
                 queue_manager.publish(QueueAgentMessageEvent(chunk=result), PublishFrom.APPLICATION_MANAGER)
 
-            text += result.delta.message.content
+            message = result.delta.message
+            if isinstance(message.content, str):
+                text += message.content
+            elif isinstance(message.content, list):
+                for content in message.content:
+                    if not isinstance(content, str):
+                        # TODO(QuantumGhost): Add multimodal output support for easy ui.
+                        _logger.warning("received multimodal output, type=%s", type(content))
+                        text += content.data
+                    else:
+                        text += content  # failback to str
 
             if not model:
                 model = result.model
 
             if not prompt_messages:
-                prompt_messages = result.prompt_messages
+                prompt_messages = list(result.prompt_messages)
 
             if result.delta.usage:
                 usage = result.delta.usage
 
-        if not usage:
+        if usage is None:
             usage = LLMUsage.empty_usage()
 
         llm_result = LLMResult(

api/core/app/apps/workflow/app_generator.py

@@ -5,7 +5,7 @@ import uuid
 from collections.abc import Generator, Mapping, Sequence
 from typing import Any, Literal, Optional, Union, overload
 
-from flask import Flask, copy_current_request_context, current_app, has_request_context
+from flask import Flask, current_app
 from pydantic import ValidationError
 from sqlalchemy.orm import sessionmaker
 
@@ -29,6 +29,7 @@ from core.workflow.repositories.workflow_execution_repository import WorkflowExe
 from core.workflow.repositories.workflow_node_execution_repository import WorkflowNodeExecutionRepository
 from extensions.ext_database import db
 from factories import file_factory
+from libs.flask_utils import preserve_flask_contexts
 from models import Account, App, EndUser, Workflow, WorkflowNodeExecutionTriggeredFrom
 from models.enums import WorkflowRunTriggeredFrom
 
@@ -194,6 +195,7 @@ class WorkflowAppGenerator(BaseAppGenerator):
         :param user: account or end user
         :param application_generate_entity: application generate entity
         :param invoke_from: invoke from source
+        :param workflow_execution_repository: repository for workflow execution
         :param workflow_node_execution_repository: repository for workflow node execution
         :param streaming: is stream
         :param workflow_thread_pool_id: workflow thread pool id
@@ -209,19 +211,16 @@ class WorkflowAppGenerator(BaseAppGenerator):
         # new thread with request context and contextvars
         context = contextvars.copy_context()
 
-        @copy_current_request_context
-        def worker_with_context():
-            # Run the worker within the copied context
-            return context.run(
-                self._generate_worker,
-                flask_app=current_app._get_current_object(),  # type: ignore
-                application_generate_entity=application_generate_entity,
-                queue_manager=queue_manager,
-                context=context,
-                workflow_thread_pool_id=workflow_thread_pool_id,
-            )
-
-        worker_thread = threading.Thread(target=worker_with_context)
+        worker_thread = threading.Thread(
+            target=self._generate_worker,
+            kwargs={
+                "flask_app": current_app._get_current_object(),  # type: ignore
+                "application_generate_entity": application_generate_entity,
+                "queue_manager": queue_manager,
+                "context": context,
+                "workflow_thread_pool_id": workflow_thread_pool_id,
+            },
+        )
 
         worker_thread.start()
 
@@ -408,24 +407,9 @@ class WorkflowAppGenerator(BaseAppGenerator):
         :param workflow_thread_pool_id: workflow thread pool id
         :return:
         """
-        for var, val in context.items():
-            var.set(val)
 
-        # FIXME(-LAN-): Save current user before entering new app context
-        from flask import g
-
-        saved_user = None
-        if has_request_context() and hasattr(g, "_login_user"):
-            saved_user = g._login_user
-
-        with flask_app.app_context():
+        with preserve_flask_contexts(flask_app, context_vars=context):
             try:
-                # Restore user in new app context
-                if saved_user is not None:
-                    from flask import g
-
-                    g._login_user = saved_user
-
                 # workflow app
                 runner = WorkflowAppRunner(
                     application_generate_entity=application_generate_entity,

api/core/app/task_pipeline/easy_ui_based_generate_task_pipeline.py

@@ -48,6 +48,7 @@ from core.model_manager import ModelInstance
 from core.model_runtime.entities.llm_entities import LLMResult, LLMResultChunk, LLMResultChunkDelta, LLMUsage
 from core.model_runtime.entities.message_entities import (
     AssistantPromptMessage,
+    TextPromptMessageContent,
 )
 from core.model_runtime.model_providers.__base.large_language_model import LargeLanguageModel
 from core.ops.entities.trace_entity import TraceTaskName
@@ -309,6 +310,23 @@ class EasyUIBasedGenerateTaskPipeline(BasedGenerateTaskPipeline):
                 delta_text = chunk.delta.message.content
                 if delta_text is None:
                     continue
+                if isinstance(chunk.delta.message.content, list):
+                    delta_text = ""
+                    for content in chunk.delta.message.content:
+                        logger.debug(
+                            "The content type %s in LLM chunk delta message content.: %r", type(content), content
+                        )
+                        if isinstance(content, TextPromptMessageContent):
+                            delta_text += content.data
+                        elif isinstance(content, str):
+                            delta_text += content  # failback to str
+                        else:
+                            logger.warning(
+                                "Unsupported content type %s in LLM chunk delta message content.: %r",
+                                type(content),
+                                content,
+                            )
+                            continue
 
                 if not self._task_state.llm_result.prompt_messages:
                     self._task_state.llm_result.prompt_messages = chunk.prompt_messages

api/core/model_manager.py

@@ -542,8 +542,6 @@ class LBModelManager:
 
             return config
 
-        return None
-
     def cooldown(self, config: ModelLoadBalancingConfiguration, expire: int = 60) -> None:
         """
         Cooldown model load balancing config

api/core/ops/ops_trace_manager.py

@@ -251,7 +251,7 @@ class OpsTraceManager:
             provider_config_map[tracing_provider]["trace_instance"],
             provider_config_map[tracing_provider]["config_class"],
         )
-        decrypt_trace_config_key = str(decrypt_trace_config)
+        decrypt_trace_config_key = json.dumps(decrypt_trace_config, sort_keys=True)
         tracing_instance = cls.ops_trace_instances_cache.get(decrypt_trace_config_key)
         if tracing_instance is None:
             # create new tracing_instance and update the cache if it absent

api/core/plugin/entities/plugin_daemon.py

@@ -157,9 +157,23 @@ class PluginInstallTaskStartResponse(BaseModel):
     task_id: str = Field(description="The ID of the install task.")
 
 
-class PluginUploadResponse(BaseModel):
+class PluginVerification(BaseModel):
+    """
+    Verification of the plugin.
+    """
+
+    class AuthorizedCategory(StrEnum):
+        Langgenius = "langgenius"
+        Partner = "partner"
+        Community = "community"
+
+    authorized_category: AuthorizedCategory = Field(description="The authorized category of the plugin.")
+
+
+class PluginDecodeResponse(BaseModel):
     unique_identifier: str = Field(description="The unique identifier of the plugin.")
     manifest: PluginDeclaration
+    verification: Optional[PluginVerification] = Field(default=None, description="Basic verification information")
 
 
 class PluginOAuthAuthorizationUrlResponse(BaseModel):

api/core/plugin/impl/plugin.py

@@ -10,10 +10,10 @@ from core.plugin.entities.plugin import (
     PluginInstallationSource,
 )
 from core.plugin.entities.plugin_daemon import (
+    PluginDecodeResponse,
     PluginInstallTask,
     PluginInstallTaskStartResponse,
     PluginListResponse,
-    PluginUploadResponse,
 )
 from core.plugin.impl.base import BasePluginClient
 
@@ -53,7 +53,7 @@ class PluginInstaller(BasePluginClient):
         tenant_id: str,
         pkg: bytes,
         verify_signature: bool = False,
-    ) -> PluginUploadResponse:
+    ) -> PluginDecodeResponse:
         """
         Upload a plugin package and return the plugin unique identifier.
         """
@@ -68,7 +68,7 @@ class PluginInstaller(BasePluginClient):
         return self._request_with_plugin_daemon_response(
             "POST",
             f"plugin/{tenant_id}/management/install/upload/package",
-            PluginUploadResponse,
+            PluginDecodeResponse,
             files=body,
             data=data,
         )
@@ -176,6 +176,18 @@ class PluginInstaller(BasePluginClient):
             params={"plugin_unique_identifier": plugin_unique_identifier},
         )
 
+    def decode_plugin_from_identifier(self, tenant_id: str, plugin_unique_identifier: str) -> PluginDecodeResponse:
+        """
+        Decode a plugin from an identifier.
+        """
+        return self._request_with_plugin_daemon_response(
+            "GET",
+            f"plugin/{tenant_id}/management/decode/from_identifier",
+            PluginDecodeResponse,
+            data={"plugin_unique_identifier": plugin_unique_identifier},
+            headers={"Content-Type": "application/json"},
+        )
+
     def fetch_plugin_installation_by_ids(
         self, tenant_id: str, plugin_ids: Sequence[str]
     ) -> Sequence[PluginInstallation]:

api/core/rag/datasource/vdb/matrixone/matrixone_vector.py

@@ -0,0 +1,233 @@
+import json
+import logging
+import uuid
+from functools import wraps
+from typing import Any, Optional
+
+from mo_vector.client import MoVectorClient  # type: ignore
+from pydantic import BaseModel, model_validator
+
+from configs import dify_config
+from core.rag.datasource.vdb.vector_base import BaseVector
+from core.rag.datasource.vdb.vector_factory import AbstractVectorFactory
+from core.rag.datasource.vdb.vector_type import VectorType
+from core.rag.embedding.embedding_base import Embeddings
+from core.rag.models.document import Document
+from extensions.ext_redis import redis_client
+from models.dataset import Dataset
+
+logger = logging.getLogger(__name__)
+
+
+class MatrixoneConfig(BaseModel):
+    host: str = "localhost"
+    port: int = 6001
+    user: str = "dump"
+    password: str = "111"
+    database: str = "dify"
+    metric: str = "l2"
+
+    @model_validator(mode="before")
+    @classmethod
+    def validate_config(cls, values: dict) -> dict:
+        if not values["host"]:
+            raise ValueError("config host is required")
+        if not values["port"]:
+            raise ValueError("config port is required")
+        if not values["user"]:
+            raise ValueError("config user is required")
+        if not values["password"]:
+            raise ValueError("config password is required")
+        if not values["database"]:
+            raise ValueError("config database is required")
+        return values
+
+
+def ensure_client(func):
+    @wraps(func)
+    def wrapper(self, *args, **kwargs):
+        if self.client is None:
+            self.client = self._get_client(None, False)
+        return func(self, *args, **kwargs)
+
+    return wrapper
+
+
+class MatrixoneVector(BaseVector):
+    """
+    Matrixone vector storage implementation.
+    """
+
+    def __init__(self, collection_name: str, config: MatrixoneConfig):
+        super().__init__(collection_name)
+        self.config = config
+        self.collection_name = collection_name.lower()
+        self.client = None
+
+    @property
+    def collection_name(self):
+        return self._collection_name
+
+    @collection_name.setter
+    def collection_name(self, value):
+        self._collection_name = value
+
+    def get_type(self) -> str:
+        return VectorType.MATRIXONE
+
+    def create(self, texts: list[Document], embeddings: list[list[float]], **kwargs):
+        if self.client is None:
+            self.client = self._get_client(len(embeddings[0]), True)
+        return self.add_texts(texts, embeddings)
+
+    def _get_client(self, dimension: Optional[int] = None, create_table: bool = False) -> MoVectorClient:
+        """
+        Create a new client for the collection.
+
+        The collection will be created if it doesn't exist.
+        """
+        lock_name = f"vector_indexing_lock_{self._collection_name}"
+        with redis_client.lock(lock_name, timeout=20):
+            client = MoVectorClient(
+                connection_string=f"mysql+pymysql://{self.config.user}:{self.config.password}@{self.config.host}:{self.config.port}/{self.config.database}",
+                table_name=self.collection_name,
+                vector_dimension=dimension,
+                create_table=create_table,
+            )
+            collection_exist_cache_key = f"vector_indexing_{self._collection_name}"
+            if redis_client.get(collection_exist_cache_key):
+                return client
+            try:
+                client.create_full_text_index()
+            except Exception as e:
+                logger.exception("Failed to create full text index")
+            redis_client.set(collection_exist_cache_key, 1, ex=3600)
+            return client
+
+    def add_texts(self, documents: list[Document], embeddings: list[list[float]], **kwargs):
+        if self.client is None:
+            self.client = self._get_client(len(embeddings[0]), True)
+        assert self.client is not None
+        ids = []
+        for _, doc in enumerate(documents):
+            if doc.metadata is not None:
+                doc_id = doc.metadata.get("doc_id", str(uuid.uuid4()))
+                ids.append(doc_id)
+        self.client.insert(
+            texts=[doc.page_content for doc in documents],
+            embeddings=embeddings,
+            metadatas=[doc.metadata for doc in documents],
+            ids=ids,
+        )
+        return ids
+
+    @ensure_client
+    def text_exists(self, id: str) -> bool:
+        assert self.client is not None
+        result = self.client.get(ids=[id])
+        return len(result) > 0
+
+    @ensure_client
+    def delete_by_ids(self, ids: list[str]) -> None:
+        assert self.client is not None
+        if not ids:
+            return
+        self.client.delete(ids=ids)
+
+    @ensure_client
+    def get_ids_by_metadata_field(self, key: str, value: str):
+        assert self.client is not None
+        results = self.client.query_by_metadata(filter={key: value})
+        return [result.id for result in results]
+
+    @ensure_client
+    def delete_by_metadata_field(self, key: str, value: str) -> None:
+        assert self.client is not None
+        self.client.delete(filter={key: value})
+
+    @ensure_client
+    def search_by_vector(self, query_vector: list[float], **kwargs: Any) -> list[Document]:
+        assert self.client is not None
+        top_k = kwargs.get("top_k", 5)
+        document_ids_filter = kwargs.get("document_ids_filter")
+        filter = None
+        if document_ids_filter:
+            filter = {"document_id": {"$in": document_ids_filter}}
+
+        results = self.client.query(
+            query_vector=query_vector,
+            k=top_k,
+            filter=filter,
+        )
+
+        docs = []
+        # TODO: add the score threshold to the query
+        for result in results:
+            metadata = result.metadata
+            docs.append(
+                Document(
+                    page_content=result.document,
+                    metadata=metadata,
+                )
+            )
+        return docs
+
+    @ensure_client
+    def search_by_full_text(self, query: str, **kwargs: Any) -> list[Document]:
+        assert self.client is not None
+        top_k = kwargs.get("top_k", 5)
+        document_ids_filter = kwargs.get("document_ids_filter")
+        filter = None
+        if document_ids_filter:
+            filter = {"document_id": {"$in": document_ids_filter}}
+        score_threshold = float(kwargs.get("score_threshold", 0.0))
+
+        results = self.client.full_text_query(
+            keywords=[query],
+            k=top_k,
+            filter=filter,
+        )
+
+        docs = []
+        for result in results:
+            metadata = result.metadata
+            if isinstance(metadata, str):
+                import json
+
+                metadata = json.loads(metadata)
+            score = 1 - result.distance
+            if score >= score_threshold:
+                metadata["score"] = score
+                docs.append(
+                    Document(
+                        page_content=result.document,
+                        metadata=metadata,
+                    )
+                )
+        return docs
+
+    @ensure_client
+    def delete(self) -> None:
+        assert self.client is not None
+        self.client.delete()
+
+
+class MatrixoneVectorFactory(AbstractVectorFactory):
+    def init_vector(self, dataset: Dataset, attributes: list, embeddings: Embeddings) -> MatrixoneVector:
+        if dataset.index_struct_dict:
+            class_prefix: str = dataset.index_struct_dict["vector_store"]["class_prefix"]
+            collection_name = class_prefix
+        else:
+            dataset_id = dataset.id
+            collection_name = Dataset.gen_collection_name_by_id(dataset_id)
+            dataset.index_struct = json.dumps(self.gen_index_struct_dict(VectorType.MATRIXONE, collection_name))
+
+        config = MatrixoneConfig(
+            host=dify_config.MATRIXONE_HOST or "localhost",
+            port=dify_config.MATRIXONE_PORT or 6001,
+            user=dify_config.MATRIXONE_USER or "dump",
+            password=dify_config.MATRIXONE_PASSWORD or "111",
+            database=dify_config.MATRIXONE_DATABASE or "dify",
+            metric=dify_config.MATRIXONE_METRIC or "l2",
+        )
+        return MatrixoneVector(collection_name=collection_name, config=config)

api/core/rag/datasource/vdb/oceanbase/oceanbase_vector.py

@@ -80,6 +80,23 @@ class OceanBaseVector(BaseVector):
 
             self.delete()
 
+            vals = []
+            params = self._client.perform_raw_text_sql("SHOW PARAMETERS LIKE '%ob_vector_memory_limit_percentage%'")
+            for row in params:
+                val = int(row[6])
+                vals.append(val)
+            if len(vals) == 0:
+                raise ValueError("ob_vector_memory_limit_percentage not found in parameters.")
+            if any(val == 0 for val in vals):
+                try:
+                    self._client.perform_raw_text_sql("ALTER SYSTEM SET ob_vector_memory_limit_percentage = 30")
+                except Exception as e:
+                    raise Exception(
+                        "Failed to set ob_vector_memory_limit_percentage. "
+                        + "Maybe the database user has insufficient privilege.",
+                        e,
+                    )
+
             cols = [
                 Column("id", String(36), primary_key=True, autoincrement=False),
                 Column("vector", VECTOR(self._vec_dim)),
@@ -110,22 +127,6 @@ class OceanBaseVector(BaseVector):
                     + "to support fulltext index and vector index in the same table",
                     e,
                 )
-            vals = []
-            params = self._client.perform_raw_text_sql("SHOW PARAMETERS LIKE '%ob_vector_memory_limit_percentage%'")
-            for row in params:
-                val = int(row[6])
-                vals.append(val)
-            if len(vals) == 0:
-                raise ValueError("ob_vector_memory_limit_percentage not found in parameters.")
-            if any(val == 0 for val in vals):
-                try:
-                    self._client.perform_raw_text_sql("ALTER SYSTEM SET ob_vector_memory_limit_percentage = 30")
-                except Exception as e:
-                    raise Exception(
-                        "Failed to set ob_vector_memory_limit_percentage. "
-                        + "Maybe the database user has insufficient privilege.",
-                        e,
-                    )
             redis_client.set(collection_exist_cache_key, 1, ex=3600)
 
     def _check_hybrid_search_support(self) -> bool:

api/core/rag/datasource/vdb/vector_factory.py

@@ -164,6 +164,10 @@ class Vector:
                 from core.rag.datasource.vdb.huawei.huawei_cloud_vector import HuaweiCloudVectorFactory
 
                 return HuaweiCloudVectorFactory
+            case VectorType.MATRIXONE:
+                from core.rag.datasource.vdb.matrixone.matrixone_vector import MatrixoneVectorFactory
+
+                return MatrixoneVectorFactory
             case _:
                 raise ValueError(f"Vector store {vector_type} is not supported.")
 

api/core/rag/datasource/vdb/vector_type.py

@@ -29,3 +29,4 @@ class VectorType(StrEnum):
     OPENGAUSS = "opengauss"
     TABLESTORE = "tablestore"
     HUAWEI_CLOUD = "huawei_cloud"
+    MATRIXONE = "matrixone"

api/core/rag/datasource/vdb/weaviate/weaviate_vector.py

@@ -41,6 +41,13 @@ class WeaviateVector(BaseVector):
 
         weaviate.connect.connection.has_grpc = False
 
+        # Fix to minimize the performance impact of the deprecation check in weaviate-client 3.24.0,
+        # by changing the connection timeout to pypi.org from 1 second to 0.001 seconds.
+        # TODO: This can be removed once weaviate-client is updated to 3.26.7 or higher,
+        #       which does not contain the deprecation check.
+        if hasattr(weaviate.connect.connection, "PYPI_TIMEOUT"):
+            weaviate.connect.connection.PYPI_TIMEOUT = 0.001
+
         try:
             client = weaviate.Client(
                 url=config.endpoint, auth_client_secret=auth_config, timeout_config=(5, 60), startup_period=None

api/core/rag/extractor/firecrawl/firecrawl_app.py

@@ -22,6 +22,7 @@ class FirecrawlApp:
             "formats": ["markdown"],
             "onlyMainContent": True,
             "timeout": 30000,
+            "integration": "dify",
         }
         if params:
             json_data.update(params)
@@ -39,7 +40,7 @@ class FirecrawlApp:
     def crawl_url(self, url, params=None) -> str:
         # Documentation: https://docs.firecrawl.dev/api-reference/endpoint/crawl-post
         headers = self._prepare_headers()
-        json_data = {"url": url}
+        json_data = {"url": url, "integration": "dify"}
         if params:
             json_data.update(params)
         response = self._post_request(f"{self.base_url}/v1/crawl", json_data, headers)
@@ -49,7 +50,6 @@ class FirecrawlApp:
             return cast(str, job_id)
         else:
             self._handle_error(response, "start crawl job")
-            # FIXME: unreachable code for mypy
             return ""  # unreachable
 
     def check_crawl_status(self, job_id) -> dict[str, Any]:
@@ -82,7 +82,6 @@ class FirecrawlApp:
                 )
         else:
             self._handle_error(response, "check crawl status")
-            # FIXME: unreachable code for mypy
             return {}  # unreachable
 
     def _format_crawl_status_response(
@@ -126,4 +125,31 @@ class FirecrawlApp:
 
     def _handle_error(self, response, action) -> None:
         error_message = response.json().get("error", "Unknown error occurred")
-        raise Exception(f"Failed to {action}. Status code: {response.status_code}. Error: {error_message}")
+        raise Exception(f"Failed to {action}. Status code: {response.status_code}. Error: {error_message}")  # type: ignore[return]
+
+    def search(self, query: str, params: dict[str, Any] | None = None) -> dict[str, Any]:
+        # Documentation: https://docs.firecrawl.dev/api-reference/endpoint/search
+        headers = self._prepare_headers()
+        json_data = {
+            "query": query,
+            "limit": 5,
+            "lang": "en",
+            "country": "us",
+            "timeout": 60000,
+            "ignoreInvalidURLs": False,
+            "scrapeOptions": {},
+            "integration": "dify",
+        }
+        if params:
+            json_data.update(params)
+        response = self._post_request(f"{self.base_url}/v1/search", json_data, headers)
+        if response.status_code == 200:
+            response_data = response.json()
+            if not response_data.get("success"):
+                raise Exception(f"Search failed. Error: {response_data.get('warning', 'Unknown error')}")
+            return cast(dict[str, Any], response_data)
+        elif response.status_code in {402, 409, 500, 429, 408}:
+            self._handle_error(response, "perform search")
+            return {}  # Avoid additional exception after handling error
+        else:
+            raise Exception(f"Failed to perform search. Status code: {response.status_code}")

api/core/rag/extractor/notion_extractor.py

@@ -79,55 +79,71 @@ class NotionExtractor(BaseExtractor):
     def _get_notion_database_data(self, database_id: str, query_dict: dict[str, Any] = {}) -> list[Document]:
         """Get all the pages from a Notion database."""
         assert self._notion_access_token is not None, "Notion access token is required"
-        res = requests.post(
-            DATABASE_URL_TMPL.format(database_id=database_id),
-            headers={
-                "Authorization": "Bearer " + self._notion_access_token,
-                "Content-Type": "application/json",
-                "Notion-Version": "2022-06-28",
-            },
-            json=query_dict,
-        )
-
-        data = res.json()
 
         database_content = []
-        if "results" not in data or data["results"] is None:
+        next_cursor = None
+        has_more = True
+
+        while has_more:
+            current_query = query_dict.copy()
+            if next_cursor:
+                current_query["start_cursor"] = next_cursor
+
+            res = requests.post(
+                DATABASE_URL_TMPL.format(database_id=database_id),
+                headers={
+                    "Authorization": "Bearer " + self._notion_access_token,
+                    "Content-Type": "application/json",
+                    "Notion-Version": "2022-06-28",
+                },
+                json=current_query,
+            )
+
+            response_data = res.json()
+
+            if "results" not in response_data or response_data["results"] is None:
+                break
+
+            for result in response_data["results"]:
+                properties = result["properties"]
+                data = {}
+                value: Any
+                for property_name, property_value in properties.items():
+                    type = property_value["type"]
+                    if type == "multi_select":
+                        value = []
+                        multi_select_list = property_value[type]
+                        for multi_select in multi_select_list:
+                            value.append(multi_select["name"])
+                    elif type in {"rich_text", "title"}:
+                        if len(property_value[type]) > 0:
+                            value = property_value[type][0]["plain_text"]
+                        else:
+                            value = ""
+                    elif type in {"select", "status"}:
+                        if property_value[type]:
+                            value = property_value[type]["name"]
+                        else:
+                            value = ""
+                    else:
+                        value = property_value[type]
+                    data[property_name] = value
+                row_dict = {k: v for k, v in data.items() if v}
+                row_content = ""
+                for key, value in row_dict.items():
+                    if isinstance(value, dict):
+                        value_dict = {k: v for k, v in value.items() if v}
+                        value_content = "".join(f"{k}:{v} " for k, v in value_dict.items())
+                        row_content = row_content + f"{key}:{value_content}\n"
+                    else:
+                        row_content = row_content + f"{key}:{value}\n"
+                database_content.append(row_content)
+
+            has_more = response_data.get("has_more", False)
+            next_cursor = response_data.get("next_cursor")
+
+        if not database_content:
             return []
-        for result in data["results"]:
-            properties = result["properties"]
-            data = {}
-            value: Any
-            for property_name, property_value in properties.items():
-                type = property_value["type"]
-                if type == "multi_select":
-                    value = []
-                    multi_select_list = property_value[type]
-                    for multi_select in multi_select_list:
-                        value.append(multi_select["name"])
-                elif type in {"rich_text", "title"}:
-                    if len(property_value[type]) > 0:
-                        value = property_value[type][0]["plain_text"]
-                    else:
-                        value = ""
-                elif type in {"select", "status"}:
-                    if property_value[type]:
-                        value = property_value[type]["name"]
-                    else:
-                        value = ""
-                else:
-                    value = property_value[type]
-                data[property_name] = value
-            row_dict = {k: v for k, v in data.items() if v}
-            row_content = ""
-            for key, value in row_dict.items():
-                if isinstance(value, dict):
-                    value_dict = {k: v for k, v in value.items() if v}
-                    value_content = "".join(f"{k}:{v} " for k, v in value_dict.items())
-                    row_content = row_content + f"{key}:{value_content}\n"
-                else:
-                    row_content = row_content + f"{key}:{value}\n"
-            database_content.append(row_content)
 
         return [Document(page_content="\n".join(database_content))]
 

api/core/rag/index_processor/processor/qa_index_processor.py

@@ -104,7 +104,7 @@ class QAIndexProcessor(BaseIndexProcessor):
 
     def format_by_template(self, file: FileStorage, **kwargs) -> list[Document]:
         # check file type
-        if not file.filename or not file.filename.endswith(".csv"):
+        if not file.filename or not file.filename.lower().endswith(".csv"):
             raise ValueError("Invalid file type. Only CSV files are allowed")
 
         try:

api/core/rag/retrieval/dataset_retrieval.py

@@ -496,6 +496,8 @@ class DatasetRetrieval:
                     all_documents = self.calculate_keyword_score(query, all_documents, top_k)
                 elif index_type == "high_quality":
                     all_documents = self.calculate_vector_score(all_documents, top_k, score_threshold)
+                else:
+                    all_documents = all_documents[:top_k] if top_k else all_documents
 
         self._on_query(query, dataset_ids, app_id, user_from, user_id)
 

api/core/repositories/sqlalchemy_workflow_execution_repository.py

@@ -6,7 +6,7 @@ import json
 import logging
 from typing import Optional, Union
 
-from sqlalchemy import select
+from sqlalchemy import func, select
 from sqlalchemy.engine import Engine
 from sqlalchemy.orm import sessionmaker
 
@@ -151,11 +151,11 @@ class SQLAlchemyWorkflowExecutionRepository(WorkflowExecutionRepository):
             existing = session.scalar(select(WorkflowRun).where(WorkflowRun.id == domain_model.id_))
             if not existing:
                 # For new records, get the next sequence number
-                stmt = select(WorkflowRun.sequence_number).where(
+                stmt = select(func.max(WorkflowRun.sequence_number)).where(
                     WorkflowRun.app_id == self._app_id,
                     WorkflowRun.tenant_id == self._tenant_id,
                 )
-                max_sequence = session.scalar(stmt.order_by(WorkflowRun.sequence_number.desc()))
+                max_sequence = session.scalar(stmt)
                 db_model.sequence_number = (max_sequence or 0) + 1
             else:
                 # For updates, keep the existing sequence number

api/core/workflow/graph_engine/graph_engine.py

@@ -9,7 +9,7 @@ from copy import copy, deepcopy
 from datetime import UTC, datetime
 from typing import Any, Optional, cast
 
-from flask import Flask, current_app, has_request_context
+from flask import Flask, current_app
 
 from configs import dify_config
 from core.app.apps.base_app_queue_manager import GenerateTaskStoppedError
@@ -53,6 +53,7 @@ from core.workflow.nodes.end.end_stream_processor import EndStreamProcessor
 from core.workflow.nodes.enums import ErrorStrategy, FailBranchSourceHandle
 from core.workflow.nodes.event import RunCompletedEvent, RunRetrieverResourceEvent, RunStreamChunkEvent
 from core.workflow.nodes.node_mapping import NODE_TYPE_CLASSES_MAPPING
+from libs.flask_utils import preserve_flask_contexts
 from models.enums import UserFrom
 from models.workflow import WorkflowType
 
@@ -537,24 +538,9 @@ class GraphEngine:
         """
         Run parallel nodes
         """
-        for var, val in context.items():
-            var.set(val)
 
-        # FIXME(-LAN-): Save current user before entering new app context
-        from flask import g
-
-        saved_user = None
-        if has_request_context() and hasattr(g, "_login_user"):
-            saved_user = g._login_user
-
-        with flask_app.app_context():
+        with preserve_flask_contexts(flask_app, context_vars=context):
             try:
-                # Restore user in new app context
-                if saved_user is not None:
-                    from flask import g
-
-                    g._login_user = saved_user
-
                 q.put(
                     ParallelBranchRunStartedEvent(
                         parallel_id=parallel_id,
@@ -653,26 +639,19 @@ class GraphEngine:
                 retry_start_at = datetime.now(UTC).replace(tzinfo=None)
                 # yield control to other threads
                 time.sleep(0.001)
-                generator = node_instance.run()
-                for item in generator:
-                    if isinstance(item, GraphEngineEvent):
-                        if isinstance(item, BaseIterationEvent):
-                            # add parallel info to iteration event
-                            item.parallel_id = parallel_id
-                            item.parallel_start_node_id = parallel_start_node_id
-                            item.parent_parallel_id = parent_parallel_id
-                            item.parent_parallel_start_node_id = parent_parallel_start_node_id
-                        elif isinstance(item, BaseLoopEvent):
-                            # add parallel info to loop event
-                            item.parallel_id = parallel_id
-                            item.parallel_start_node_id = parallel_start_node_id
-                            item.parent_parallel_id = parent_parallel_id
-                            item.parent_parallel_start_node_id = parent_parallel_start_node_id
-
-                        yield item
+                event_stream = node_instance.run()
+                for event in event_stream:
+                    if isinstance(event, GraphEngineEvent):
+                        # add parallel info to iteration event
+                        if isinstance(event, BaseIterationEvent | BaseLoopEvent):
+                            event.parallel_id = parallel_id
+                            event.parallel_start_node_id = parallel_start_node_id
+                            event.parent_parallel_id = parent_parallel_id
+                            event.parent_parallel_start_node_id = parent_parallel_start_node_id
+                        yield event
                     else:
-                        if isinstance(item, RunCompletedEvent):
-                            run_result = item.run_result
+                        if isinstance(event, RunCompletedEvent):
+                            run_result = event.run_result
                             if run_result.status == WorkflowNodeExecutionStatus.FAILED:
                                 if (
                                     retries == max_retries
@@ -708,7 +687,7 @@ class GraphEngine:
                                     # if run failed, handle error
                                     run_result = self._handle_continue_on_error(
                                         node_instance,
-                                        item.run_result,
+                                        event.run_result,
                                         self.graph_runtime_state.variable_pool,
                                         handle_exceptions=handle_exceptions,
                                     )
@@ -811,28 +790,28 @@ class GraphEngine:
                                 should_continue_retry = False
 
                             break
-                        elif isinstance(item, RunStreamChunkEvent):
+                        elif isinstance(event, RunStreamChunkEvent):
                             yield NodeRunStreamChunkEvent(
                                 id=node_instance.id,
                                 node_id=node_instance.node_id,
                                 node_type=node_instance.node_type,
                                 node_data=node_instance.node_data,
-                                chunk_content=item.chunk_content,
-                                from_variable_selector=item.from_variable_selector,
+                                chunk_content=event.chunk_content,
+                                from_variable_selector=event.from_variable_selector,
                                 route_node_state=route_node_state,
                                 parallel_id=parallel_id,
                                 parallel_start_node_id=parallel_start_node_id,
                                 parent_parallel_id=parent_parallel_id,
                                 parent_parallel_start_node_id=parent_parallel_start_node_id,
                             )
-                        elif isinstance(item, RunRetrieverResourceEvent):
+                        elif isinstance(event, RunRetrieverResourceEvent):
                             yield NodeRunRetrieverResourceEvent(
                                 id=node_instance.id,
                                 node_id=node_instance.node_id,
                                 node_type=node_instance.node_type,
                                 node_data=node_instance.node_data,
-                                retriever_resources=item.retriever_resources,
-                                context=item.context,
+                                retriever_resources=event.retriever_resources,
+                                context=event.context,
                                 route_node_state=route_node_state,
                                 parallel_id=parallel_id,
                                 parallel_start_node_id=parallel_start_node_id,

api/core/workflow/nodes/agent/agent_node.py

@@ -219,7 +219,7 @@ class AgentNode(ToolNode):
                         )
                         if tool_runtime.entity.description:
                             tool_runtime.entity.description.llm = (
-                                extra.get("descrption", "") or tool_runtime.entity.description.llm
+                                extra.get("description", "") or tool_runtime.entity.description.llm
                             )
                         for tool_runtime_params in tool_runtime.entity.parameters:
                             tool_runtime_params.form = (

api/core/workflow/nodes/answer/base_stream_processor.py

@@ -57,7 +57,6 @@ class StreamProcessor(ABC):
 
                     # The branch_identify parameter is added to ensure that
                     # only nodes in the correct logical branch are included.
-                    reachable_node_ids.append(edge.target_node_id)
                     ids = self._fetch_node_ids_in_reachable_branch(edge.target_node_id, run_result.edge_source_handle)
                     reachable_node_ids.extend(ids)
                 else:
@@ -74,6 +73,8 @@ class StreamProcessor(ABC):
                 self._remove_node_ids_in_unreachable_branch(node_id, reachable_node_ids)
 
     def _fetch_node_ids_in_reachable_branch(self, node_id: str, branch_identify: Optional[str] = None) -> list[str]:
+        if node_id not in self.rest_node_ids:
+            self.rest_node_ids.append(node_id)
         node_ids = []
         for edge in self.graph.edge_mapping.get(node_id, []):
             if edge.target_node_id == self.graph.root_node_id:

api/core/workflow/nodes/event/event.py

@@ -6,7 +6,6 @@ from pydantic import BaseModel, Field
 from core.model_runtime.entities.llm_entities import LLMUsage
 from core.rag.entities.citation_metadata import RetrievalSourceMetadata
 from core.workflow.entities.node_entities import NodeRunResult
-from core.workflow.entities.workflow_node_execution import WorkflowNodeExecutionStatus
 
 
 class RunCompletedEvent(BaseModel):
@@ -39,11 +38,3 @@ class RunRetryEvent(BaseModel):
     error: str = Field(..., description="error")
     retry_index: int = Field(..., description="Retry attempt number")
     start_at: datetime = Field(..., description="Retry start time")
-
-
-class SingleStepRetryEvent(NodeRunResult):
-    """Single step retry event"""
-
-    status: WorkflowNodeExecutionStatus = WorkflowNodeExecutionStatus.RETRY
-
-    elapsed_time: float = Field(..., description="elapsed time")

api/core/workflow/nodes/iteration/iteration_node.py

@@ -7,7 +7,7 @@ from datetime import UTC, datetime
 from queue import Empty, Queue
 from typing import TYPE_CHECKING, Any, Optional, cast
 
-from flask import Flask, current_app, has_request_context
+from flask import Flask, current_app
 
 from configs import dify_config
 from core.variables import ArrayVariable, IntegerVariable, NoneVariable
@@ -37,6 +37,7 @@ from core.workflow.nodes.base import BaseNode
 from core.workflow.nodes.enums import NodeType
 from core.workflow.nodes.event import NodeEvent, RunCompletedEvent
 from core.workflow.nodes.iteration.entities import ErrorHandleMode, IterationNodeData
+from libs.flask_utils import preserve_flask_contexts
 
 from .exc import (
     InvalidIteratorValueError,
@@ -583,23 +584,8 @@ class IterationNode(BaseNode[IterationNodeData]):
         """
         run single iteration in parallel mode
         """
-        for var, val in context.items():
-            var.set(val)
-
-        # FIXME(-LAN-): Save current user before entering new app context
-        from flask import g
-
-        saved_user = None
-        if has_request_context() and hasattr(g, "_login_user"):
-            saved_user = g._login_user
-
-        with flask_app.app_context():
-            # Restore user in new app context
-            if saved_user is not None:
-                from flask import g
-
-                g._login_user = saved_user
 
+        with preserve_flask_contexts(flask_app, context_vars=context):
             parallel_mode_run_id = uuid.uuid4().hex
             graph_engine_copy = graph_engine.create_copy()
             variable_pool_copy = graph_engine_copy.graph_runtime_state.variable_pool

api/core/workflow/nodes/llm/node.py

@@ -525,6 +525,8 @@ class LLMNode(BaseNode[LLMNodeData]):
                 # Set appropriate response format based on model capabilities
                 self._set_response_format(completion_params, model_schema.parameter_rules)
         model_config_with_cred.parameters = completion_params
+        # NOTE(-LAN-): This line modify the `self.node_data.model`, which is used in `_invoke_llm()`.
+        node_data_model.completion_params = completion_params
         return model, model_config_with_cred
 
     def _fetch_prompt_messages(

api/core/workflow/nodes/variable_assigner/v2/constants.py

@@ -8,4 +8,5 @@ EMPTY_VALUE_MAPPING = {
     SegmentType.ARRAY_STRING: [],
     SegmentType.ARRAY_NUMBER: [],
     SegmentType.ARRAY_OBJECT: [],
+    SegmentType.ARRAY_FILE: [],
 }

api/core/workflow/nodes/variable_assigner/v2/helpers.py

@@ -1,5 +1,6 @@
 from typing import Any
 
+from core.file import File
 from core.variables import SegmentType
 
 from .enums import Operation
@@ -85,6 +86,8 @@ def is_input_value_valid(*, variable_type: SegmentType, operation: Operation, va
             return isinstance(value, int | float)
         case SegmentType.ARRAY_OBJECT if operation == Operation.APPEND:
             return isinstance(value, dict)
+        case SegmentType.ARRAY_FILE if operation == Operation.APPEND:
+            return isinstance(value, File)
 
         # Array & Extend / Overwrite
         case SegmentType.ARRAY_ANY if operation in {Operation.EXTEND, Operation.OVER_WRITE}:
@@ -95,6 +98,8 @@ def is_input_value_valid(*, variable_type: SegmentType, operation: Operation, va
             return isinstance(value, list) and all(isinstance(item, int | float) for item in value)
         case SegmentType.ARRAY_OBJECT if operation in {Operation.EXTEND, Operation.OVER_WRITE}:
             return isinstance(value, list) and all(isinstance(item, dict) for item in value)
+        case SegmentType.ARRAY_FILE if operation in {Operation.EXTEND, Operation.OVER_WRITE}:
+            return isinstance(value, list) and all(isinstance(item, File) for item in value)
 
         case _:
             return False

api/extensions/ext_mail.py

@@ -54,6 +54,15 @@ class Mail:
                     use_tls=dify_config.SMTP_USE_TLS,
                     opportunistic_tls=dify_config.SMTP_OPPORTUNISTIC_TLS,
                 )
+            case "sendgrid":
+                from libs.sendgrid import SendGridClient
+
+                if not dify_config.SENDGRID_API_KEY:
+                    raise ValueError("SENDGRID_API_KEY is required for SendGrid mail type")
+
+                self._client = SendGridClient(
+                    sendgrid_api_key=dify_config.SENDGRID_API_KEY, _from=dify_config.MAIL_DEFAULT_SEND_FROM or ""
+                )
             case _:
                 raise ValueError("Unsupported mail type {}".format(mail_type))
 

api/factories/variable_factory.py

@@ -39,10 +39,6 @@ from core.variables.variables import (
 from core.workflow.constants import CONVERSATION_VARIABLE_NODE_ID, ENVIRONMENT_VARIABLE_NODE_ID
 
 
-class InvalidSelectorError(ValueError):
-    pass
-
-
 class UnsupportedSegmentTypeError(Exception):
     pass
 
@@ -105,6 +101,8 @@ def _build_variable_from_mapping(*, mapping: Mapping[str, Any], selector: Sequen
             result = ArrayNumberVariable.model_validate(mapping)
         case SegmentType.ARRAY_OBJECT if isinstance(value, list):
             result = ArrayObjectVariable.model_validate(mapping)
+        case SegmentType.ARRAY_FILE if isinstance(value, list):
+            result = ArrayFileVariable.model_validate(mapping)
         case _:
             raise VariableError(f"not supported value type {value_type}")
     if result.size > dify_config.MAX_VARIABLE_SIZE:

api/libs/flask_utils.py

@@ -0,0 +1,65 @@
+import contextvars
+from collections.abc import Iterator
+from contextlib import contextmanager
+from typing import TypeVar
+
+from flask import Flask, g, has_request_context
+
+T = TypeVar("T")
+
+
+@contextmanager
+def preserve_flask_contexts(
+    flask_app: Flask,
+    context_vars: contextvars.Context,
+) -> Iterator[None]:
+    """
+    A context manager that handles:
+    1. flask-login's UserProxy copy
+    2. ContextVars copy
+    3. flask_app.app_context()
+
+    This context manager ensures that the Flask application context is properly set up,
+    the current user is preserved across context boundaries, and any provided context variables
+    are set within the new context.
+
+    Note:
+        This manager aims to allow use current_user cross thread and app context,
+        but it's not the recommend use, it's better to pass user directly in parameters.
+
+    Args:
+        flask_app: The Flask application instance
+        context_vars: contextvars.Context object containing context variables to be set in the new context
+
+    Yields:
+        None
+
+    Example:
+        ```python
+        with preserve_flask_contexts(flask_app, context_vars=context_vars):
+            # Code that needs Flask app context and context variables
+            # Current user will be preserved if available
+        ```
+    """
+    # Set context variables if provided
+    if context_vars:
+        for var, val in context_vars.items():
+            var.set(val)
+
+    # Save current user before entering new app context
+    saved_user = None
+    if has_request_context() and hasattr(g, "_login_user"):
+        saved_user = g._login_user
+
+    # Enter Flask app context
+    with flask_app.app_context():
+        try:
+            # Restore user in new app context if it was saved
+            if saved_user is not None:
+                g._login_user = saved_user
+
+            # Yield control back to the caller
+            yield
+        finally:
+            # Any cleanup can be added here if needed
+            pass

api/libs/sendgrid.py

@@ -0,0 +1,42 @@
+import logging
+
+import sendgrid  # type: ignore
+from python_http_client.exceptions import ForbiddenError, UnauthorizedError
+from sendgrid.helpers.mail import Content, Email, Mail, To  # type: ignore
+
+
+class SendGridClient:
+    def __init__(self, sendgrid_api_key: str, _from: str):
+        self.sendgrid_api_key = sendgrid_api_key
+        self._from = _from
+
+    def send(self, mail: dict):
+        logging.debug("Sending email with SendGrid")
+
+        try:
+            _to = mail["to"]
+
+            if not _to:
+                raise ValueError("SendGridClient: Cannot send email: recipient address is missing.")
+
+            sg = sendgrid.SendGridAPIClient(api_key=self.sendgrid_api_key)
+            from_email = Email(self._from)
+            to_email = To(_to)
+            subject = mail["subject"]
+            content = Content("text/html", mail["html"])
+            mail = Mail(from_email, to_email, subject, content)
+            mail_json = mail.get()  # type: ignore
+            response = sg.client.mail.send.post(request_body=mail_json)
+            logging.debug(response.status_code)
+            logging.debug(response.body)
+            logging.debug(response.headers)
+
+        except TimeoutError as e:
+            logging.exception("SendGridClient Timeout occurred while sending email")
+            raise
+        except (UnauthorizedError, ForbiddenError) as e:
+            logging.exception("SendGridClient Authentication failed. Verify that your credentials and the 'from")
+            raise
+        except Exception as e:
+            logging.exception(f"SendGridClient Unexpected error occurred while sending email to {_to}")
+            raise

api/models/model.py

@@ -10,7 +10,6 @@ from core.plugin.entities.plugin import GenericProviderID
 from core.tools.entities.tool_entities import ToolProviderType
 from core.tools.signature import sign_tool_file
 from core.workflow.entities.workflow_execution import WorkflowExecutionStatus
-from services.plugin.plugin_service import PluginService
 
 if TYPE_CHECKING:
     from models.workflow import Workflow
@@ -169,6 +168,7 @@ class App(Base):
     @property
     def deleted_tools(self) -> list:
         from core.tools.tool_manager import ToolManager
+        from services.plugin.plugin_service import PluginService
 
         # get agent mode tools
         app_model_config = self.app_model_config

api/mypy.ini

@@ -18,4 +18,3 @@ ignore_missing_imports=True
 
 [mypy-flask_restful.inputs]
 ignore_missing_imports=True
-

api/pyproject.toml

@@ -83,6 +83,7 @@ dependencies = [
     "webvtt-py~=0.5.1",
     "sseclient-py>=1.8.0",
     "httpx-sse>=0.4.0",
+    "sendgrid~=6.12.3",
 ]
 # Before adding new dependency, consider place it in
 # alphabet order (a-z) and suitable group.
@@ -204,4 +205,5 @@ vdb = [
     "volcengine-compat~=1.0.0",
     "weaviate-client~=3.24.0",
     "xinference-client~=1.2.2",
+    "mo-vector~=0.1.13",
 ]

api/services/app_dsl_service.py

@@ -421,7 +421,7 @@ class AppDslService:
 
         # Set icon type
         icon_type_value = icon_type or app_data.get("icon_type")
-        if icon_type_value in ["emoji", "link"]:
+        if icon_type_value in ["emoji", "link", "image"]:
             icon_type = icon_type_value
         else:
             icon_type = "emoji"

api/services/dataset_service.py

@@ -1402,16 +1402,16 @@ class DocumentService:
                 knowledge_config.embedding_model,  # type: ignore
             )
             dataset_collection_binding_id = dataset_collection_binding.id
-            if knowledge_config.retrieval_model:
-                retrieval_model = knowledge_config.retrieval_model
-            else:
-                retrieval_model = RetrievalModel(
-                    search_method=RetrievalMethod.SEMANTIC_SEARCH.value,
-                    reranking_enable=False,
-                    reranking_model=RerankingModel(reranking_provider_name="", reranking_model_name=""),
-                    top_k=2,
-                    score_threshold_enabled=False,
-                )
+        if knowledge_config.retrieval_model:
+            retrieval_model = knowledge_config.retrieval_model
+        else:
+            retrieval_model = RetrievalModel(
+                search_method=RetrievalMethod.SEMANTIC_SEARCH.value,
+                reranking_enable=False,
+                reranking_model=RerankingModel(reranking_provider_name="", reranking_model_name=""),
+                top_k=2,
+                score_threshold_enabled=False,
+            )
         # save dataset
         dataset = Dataset(
             tenant_id=tenant_id,

api/services/entities/knowledge_entities/knowledge_entities.py

@@ -101,7 +101,7 @@ class WeightModel(BaseModel):
 
 
 class RetrievalModel(BaseModel):
-    search_method: Literal["hybrid_search", "semantic_search", "full_text_search"]
+    search_method: Literal["hybrid_search", "semantic_search", "full_text_search", "keyword_search"]
     reranking_enable: bool
     reranking_model: Optional[RerankingModel] = None
     reranking_mode: Optional[str] = None

api/services/errors/__init__.py

@@ -4,7 +4,6 @@ from . import (
     app_model_config,
     audio,
     base,
-    completion,
     conversation,
     dataset,
     document,
@@ -19,7 +18,6 @@ __all__ = [
     "app_model_config",
     "audio",
     "base",
-    "completion",
     "conversation",
     "dataset",
     "document",

api/services/errors/account.py

@@ -55,7 +55,3 @@ class MemberNotInTenantError(BaseServiceError):
 
 class RoleAlreadyAssignedError(BaseServiceError):
     pass
-
-
-class RateLimitExceededError(BaseServiceError):
-    pass

api/services/errors/plugin.py

@@ -1,5 +1,5 @@
 from services.errors.base import BaseServiceError
 
 
-class CompletionStoppedError(BaseServiceError):
+class PluginInstallationForbiddenError(BaseServiceError):
     pass

api/services/feature_service.py

@@ -88,6 +88,26 @@ class WebAppAuthModel(BaseModel):
     allow_email_password_login: bool = False
 
 
+class PluginInstallationScope(StrEnum):
+    NONE = "none"
+    OFFICIAL_ONLY = "official_only"
+    OFFICIAL_AND_SPECIFIC_PARTNERS = "official_and_specific_partners"
+    ALL = "all"
+
+
+class PluginInstallationPermissionModel(BaseModel):
+    # Plugin installation scope – possible values:
+    #   none: prohibit all plugin installations
+    #   official_only: allow only Dify official plugins
+    #   official_and_specific_partners: allow official and specific partner plugins
+    #   all: allow installation of all plugins
+    plugin_installation_scope: PluginInstallationScope = PluginInstallationScope.ALL
+
+    # If True, restrict plugin installation to the marketplace only
+    # Equivalent to ForceEnablePluginVerification
+    restrict_to_marketplace_only: bool = False
+
+
 class FeatureModel(BaseModel):
     billing: BillingModel = BillingModel()
     education: EducationModel = EducationModel()
@@ -128,6 +148,7 @@ class SystemFeatureModel(BaseModel):
     license: LicenseModel = LicenseModel()
     branding: BrandingModel = BrandingModel()
     webapp_auth: WebAppAuthModel = WebAppAuthModel()
+    plugin_installation_permission: PluginInstallationPermissionModel = PluginInstallationPermissionModel()
 
 
 class FeatureService:
@@ -291,3 +312,12 @@ class FeatureService:
                 features.license.workspaces.enabled = license_info["workspaces"]["enabled"]
                 features.license.workspaces.limit = license_info["workspaces"]["limit"]
                 features.license.workspaces.size = license_info["workspaces"]["used"]
+
+        if "PluginInstallationPermission" in enterprise_info:
+            plugin_installation_info = enterprise_info["PluginInstallationPermission"]
+            features.plugin_installation_permission.plugin_installation_scope = plugin_installation_info[
+                "pluginInstallationScope"
+            ]
+            features.plugin_installation_permission.restrict_to_marketplace_only = plugin_installation_info[
+                "restrictToMarketplaceOnly"
+            ]

api/services/plugin/data_migration.py

@@ -3,7 +3,7 @@ import logging
 
 import click
 
-from core.entities import DEFAULT_PLUGIN_ID
+from core.plugin.entities.plugin import GenericProviderID, ModelProviderID, ToolProviderID
 from models.engine import db
 
 logger = logging.getLogger(__name__)
@@ -12,17 +12,17 @@ logger = logging.getLogger(__name__)
 class PluginDataMigration:
     @classmethod
     def migrate(cls) -> None:
-        cls.migrate_db_records("providers", "provider_name")  # large table
-        cls.migrate_db_records("provider_models", "provider_name")
-        cls.migrate_db_records("provider_orders", "provider_name")
-        cls.migrate_db_records("tenant_default_models", "provider_name")
-        cls.migrate_db_records("tenant_preferred_model_providers", "provider_name")
-        cls.migrate_db_records("provider_model_settings", "provider_name")
-        cls.migrate_db_records("load_balancing_model_configs", "provider_name")
+        cls.migrate_db_records("providers", "provider_name", ModelProviderID)  # large table
+        cls.migrate_db_records("provider_models", "provider_name", ModelProviderID)
+        cls.migrate_db_records("provider_orders", "provider_name", ModelProviderID)
+        cls.migrate_db_records("tenant_default_models", "provider_name", ModelProviderID)
+        cls.migrate_db_records("tenant_preferred_model_providers", "provider_name", ModelProviderID)
+        cls.migrate_db_records("provider_model_settings", "provider_name", ModelProviderID)
+        cls.migrate_db_records("load_balancing_model_configs", "provider_name", ModelProviderID)
         cls.migrate_datasets()
-        cls.migrate_db_records("embeddings", "provider_name")  # large table
-        cls.migrate_db_records("dataset_collection_bindings", "provider_name")
-        cls.migrate_db_records("tool_builtin_providers", "provider")
+        cls.migrate_db_records("embeddings", "provider_name", ModelProviderID)  # large table
+        cls.migrate_db_records("dataset_collection_bindings", "provider_name", ModelProviderID)
+        cls.migrate_db_records("tool_builtin_providers", "provider_name", ToolProviderID)
 
     @classmethod
     def migrate_datasets(cls) -> None:
@@ -66,9 +66,10 @@ limit 1000"""
                                     fg="white",
                                 )
                             )
-                            retrieval_model["reranking_model"]["reranking_provider_name"] = (
-                                f"{DEFAULT_PLUGIN_ID}/{retrieval_model['reranking_model']['reranking_provider_name']}/{retrieval_model['reranking_model']['reranking_provider_name']}"
-                            )
+                            # update google to langgenius/gemini/google etc.
+                            retrieval_model["reranking_model"]["reranking_provider_name"] = ModelProviderID(
+                                retrieval_model["reranking_model"]["reranking_provider_name"]
+                            ).to_string()
                             retrieval_model_changed = True
 
                     click.echo(
@@ -86,9 +87,11 @@ limit 1000"""
                             update_retrieval_model_sql = ", retrieval_model = :retrieval_model"
                             params["retrieval_model"] = json.dumps(retrieval_model)
 
+                        params["provider_name"] = ModelProviderID(provider_name).to_string()
+
                         sql = f"""update {table_name}
                         set {provider_column_name} =
-                        concat('{DEFAULT_PLUGIN_ID}/', {provider_column_name}, '/', {provider_column_name})
+                        :provider_name
                         {update_retrieval_model_sql}
                         where id = :record_id"""
                         conn.execute(db.text(sql), params)
@@ -122,7 +125,9 @@ limit 1000"""
         )
 
     @classmethod
-    def migrate_db_records(cls, table_name: str, provider_column_name: str) -> None:
+    def migrate_db_records(
+        cls, table_name: str, provider_column_name: str, provider_cls: type[GenericProviderID]
+    ) -> None:
         click.echo(click.style(f"Migrating [{table_name}] data for plugin", fg="white"))
 
         processed_count = 0
@@ -166,7 +171,8 @@ limit 1000"""
                     )
 
                     try:
-                        updated_value = f"{DEFAULT_PLUGIN_ID}/{provider_name}/{provider_name}"
+                        # update jina to langgenius/jina_tool/jina etc.
+                        updated_value = provider_cls(provider_name).to_string()
                         batch_updates.append((updated_value, record_id))
                     except Exception as e:
                         failed_ids.append(record_id)

api/services/plugin/plugin_service.py

@@ -17,11 +17,18 @@ from core.plugin.entities.plugin import (
     PluginInstallation,
     PluginInstallationSource,
 )
-from core.plugin.entities.plugin_daemon import PluginInstallTask, PluginListResponse, PluginUploadResponse
+from core.plugin.entities.plugin_daemon import (
+    PluginDecodeResponse,
+    PluginInstallTask,
+    PluginListResponse,
+    PluginVerification,
+)
 from core.plugin.impl.asset import PluginAssetManager
 from core.plugin.impl.debugging import PluginDebuggingClient
 from core.plugin.impl.plugin import PluginInstaller
 from extensions.ext_redis import redis_client
+from services.errors.plugin import PluginInstallationForbiddenError
+from services.feature_service import FeatureService, PluginInstallationScope
 
 logger = logging.getLogger(__name__)
 
@@ -86,6 +93,42 @@ class PluginService:
             logger.exception("failed to fetch latest plugin version")
             return result
 
+    @staticmethod
+    def _check_marketplace_only_permission():
+        """
+        Check if the marketplace only permission is enabled
+        """
+        features = FeatureService.get_system_features()
+        if features.plugin_installation_permission.restrict_to_marketplace_only:
+            raise PluginInstallationForbiddenError("Plugin installation is restricted to marketplace only")
+
+    @staticmethod
+    def _check_plugin_installation_scope(plugin_verification: Optional[PluginVerification]):
+        """
+        Check the plugin installation scope
+        """
+        features = FeatureService.get_system_features()
+
+        match features.plugin_installation_permission.plugin_installation_scope:
+            case PluginInstallationScope.OFFICIAL_ONLY:
+                if (
+                    plugin_verification is None
+                    or plugin_verification.authorized_category != PluginVerification.AuthorizedCategory.Langgenius
+                ):
+                    raise PluginInstallationForbiddenError("Plugin installation is restricted to official only")
+            case PluginInstallationScope.OFFICIAL_AND_SPECIFIC_PARTNERS:
+                if plugin_verification is None or plugin_verification.authorized_category not in [
+                    PluginVerification.AuthorizedCategory.Langgenius,
+                    PluginVerification.AuthorizedCategory.Partner,
+                ]:
+                    raise PluginInstallationForbiddenError(
+                        "Plugin installation is restricted to official and specific partners"
+                    )
+            case PluginInstallationScope.NONE:
+                raise PluginInstallationForbiddenError("Installing plugins is not allowed")
+            case PluginInstallationScope.ALL:
+                pass
+
     @staticmethod
     def get_debugging_key(tenant_id: str) -> str:
         """
@@ -208,6 +251,8 @@ class PluginService:
         # check if plugin pkg is already downloaded
         manager = PluginInstaller()
 
+        features = FeatureService.get_system_features()
+
         try:
             manager.fetch_plugin_manifest(tenant_id, new_plugin_unique_identifier)
             # already downloaded, skip, and record install event
@@ -215,7 +260,14 @@ class PluginService:
         except Exception:
             # plugin not installed, download and upload pkg
             pkg = download_plugin_pkg(new_plugin_unique_identifier)
-            manager.upload_pkg(tenant_id, pkg, verify_signature=False)
+            response = manager.upload_pkg(
+                tenant_id,
+                pkg,
+                verify_signature=features.plugin_installation_permission.restrict_to_marketplace_only,
+            )
+
+            # check if the plugin is available to install
+            PluginService._check_plugin_installation_scope(response.verification)
 
         return manager.upgrade_plugin(
             tenant_id,
@@ -239,6 +291,7 @@ class PluginService:
         """
         Upgrade plugin with github
         """
+        PluginService._check_marketplace_only_permission()
         manager = PluginInstaller()
         return manager.upgrade_plugin(
             tenant_id,
@@ -253,33 +306,43 @@ class PluginService:
         )
 
     @staticmethod
-    def upload_pkg(tenant_id: str, pkg: bytes, verify_signature: bool = False) -> PluginUploadResponse:
+    def upload_pkg(tenant_id: str, pkg: bytes, verify_signature: bool = False) -> PluginDecodeResponse:
         """
         Upload plugin package files
 
         returns: plugin_unique_identifier
         """
+        PluginService._check_marketplace_only_permission()
         manager = PluginInstaller()
-        return manager.upload_pkg(tenant_id, pkg, verify_signature)
+        features = FeatureService.get_system_features()
+        response = manager.upload_pkg(
+            tenant_id,
+            pkg,
+            verify_signature=features.plugin_installation_permission.restrict_to_marketplace_only,
+        )
+        return response
 
     @staticmethod
     def upload_pkg_from_github(
         tenant_id: str, repo: str, version: str, package: str, verify_signature: bool = False
-    ) -> PluginUploadResponse:
+    ) -> PluginDecodeResponse:
         """
         Install plugin from github release package files,
         returns plugin_unique_identifier
         """
+        PluginService._check_marketplace_only_permission()
         pkg = download_with_size_limit(
             f"https://github.com/{repo}/releases/download/{version}/{package}", dify_config.PLUGIN_MAX_PACKAGE_SIZE
         )
+        features = FeatureService.get_system_features()
 
         manager = PluginInstaller()
-        return manager.upload_pkg(
+        response = manager.upload_pkg(
             tenant_id,
             pkg,
-            verify_signature,
+            verify_signature=features.plugin_installation_permission.restrict_to_marketplace_only,
         )
+        return response
 
     @staticmethod
     def upload_bundle(
@@ -289,11 +352,15 @@ class PluginService:
         Upload a plugin bundle and return the dependencies.
         """
         manager = PluginInstaller()
+        PluginService._check_marketplace_only_permission()
         return manager.upload_bundle(tenant_id, bundle, verify_signature)
 
     @staticmethod
     def install_from_local_pkg(tenant_id: str, plugin_unique_identifiers: Sequence[str]):
+        PluginService._check_marketplace_only_permission()
+
         manager = PluginInstaller()
+
         return manager.install_from_identifiers(
             tenant_id,
             plugin_unique_identifiers,
@@ -307,6 +374,8 @@ class PluginService:
         Install plugin from github release package files,
         returns plugin_unique_identifier
         """
+        PluginService._check_marketplace_only_permission()
+
         manager = PluginInstaller()
         return manager.install_from_identifiers(
             tenant_id,
@@ -322,28 +391,33 @@ class PluginService:
         )
 
     @staticmethod
-    def fetch_marketplace_pkg(
-        tenant_id: str, plugin_unique_identifier: str, verify_signature: bool = False
-    ) -> PluginDeclaration:
+    def fetch_marketplace_pkg(tenant_id: str, plugin_unique_identifier: str) -> PluginDeclaration:
         """
         Fetch marketplace package
         """
         if not dify_config.MARKETPLACE_ENABLED:
             raise ValueError("marketplace is not enabled")
 
+        features = FeatureService.get_system_features()
+
         manager = PluginInstaller()
         try:
             declaration = manager.fetch_plugin_manifest(tenant_id, plugin_unique_identifier)
         except Exception:
             pkg = download_plugin_pkg(plugin_unique_identifier)
-            declaration = manager.upload_pkg(tenant_id, pkg, verify_signature).manifest
+            response = manager.upload_pkg(
+                tenant_id,
+                pkg,
+                verify_signature=features.plugin_installation_permission.restrict_to_marketplace_only,
+            )
+            # check if the plugin is available to install
+            PluginService._check_plugin_installation_scope(response.verification)
+            declaration = response.manifest
 
         return declaration
 
     @staticmethod
-    def install_from_marketplace_pkg(
-        tenant_id: str, plugin_unique_identifiers: Sequence[str], verify_signature: bool = False
-    ):
+    def install_from_marketplace_pkg(tenant_id: str, plugin_unique_identifiers: Sequence[str]):
         """
         Install plugin from marketplace package files,
         returns installation task id
@@ -353,15 +427,26 @@ class PluginService:
 
         manager = PluginInstaller()
 
+        features = FeatureService.get_system_features()
+
         # check if already downloaded
         for plugin_unique_identifier in plugin_unique_identifiers:
             try:
                 manager.fetch_plugin_manifest(tenant_id, plugin_unique_identifier)
+                plugin_decode_response = manager.decode_plugin_from_identifier(tenant_id, plugin_unique_identifier)
+                # check if the plugin is available to install
+                PluginService._check_plugin_installation_scope(plugin_decode_response.verification)
                 # already downloaded, skip
             except Exception:
                 # plugin not installed, download and upload pkg
                 pkg = download_plugin_pkg(plugin_unique_identifier)
-                manager.upload_pkg(tenant_id, pkg, verify_signature)
+                response = manager.upload_pkg(
+                    tenant_id,
+                    pkg,
+                    verify_signature=features.plugin_installation_permission.restrict_to_marketplace_only,
+                )
+                # check if the plugin is available to install
+                PluginService._check_plugin_installation_scope(response.verification)
 
         return manager.install_from_identifiers(
             tenant_id,

api/services/tools/builtin_tools_manage_service.py

@@ -7,7 +7,7 @@ from sqlalchemy.orm import Session
 from configs import dify_config
 from core.helper.position_helper import is_filtered
 from core.model_runtime.utils.encoders import jsonable_encoder
-from core.plugin.entities.plugin import GenericProviderID, ToolProviderID
+from core.plugin.entities.plugin import ToolProviderID
 from core.plugin.impl.exc import PluginDaemonClientSideError
 from core.tools.builtin_tool.providers._positions import BuiltinToolProviderSort
 from core.tools.entities.api_entities import ToolApiEntity, ToolProviderApiEntity
@@ -290,7 +290,7 @@ class BuiltinToolManageService:
     def _fetch_builtin_provider(provider_name: str, tenant_id: str) -> BuiltinToolProvider | None:
         try:
             full_provider_name = provider_name
-            provider_id_entity = GenericProviderID(provider_name)
+            provider_id_entity = ToolProviderID(provider_name)
             provider_name = provider_id_entity.provider_name
             if provider_id_entity.organization != "langgenius":
                 provider_obj = (
@@ -315,7 +315,7 @@ class BuiltinToolManageService:
             if provider_obj is None:
                 return None
 
-            provider_obj.provider = GenericProviderID(provider_obj.provider).to_string()
+            provider_obj.provider = ToolProviderID(provider_obj.provider).to_string()
             return provider_obj
         except Exception:
             # it's an old provider without organization

api/services/workflow_app_service.py

@@ -5,7 +5,7 @@ from sqlalchemy import and_, func, or_, select
 from sqlalchemy.orm import Session
 
 from core.workflow.entities.workflow_execution import WorkflowExecutionStatus
-from models import App, EndUser, WorkflowAppLog, WorkflowRun
+from models import Account, App, EndUser, WorkflowAppLog, WorkflowRun
 from models.enums import CreatorUserRole
 
 
@@ -21,6 +21,8 @@ class WorkflowAppService:
         created_at_after: datetime | None = None,
         page: int = 1,
         limit: int = 20,
+        created_by_end_user_session_id: str | None = None,
+        created_by_account: str | None = None,
     ) -> dict:
         """
         Get paginate workflow app logs using SQLAlchemy 2.0 style
@@ -32,6 +34,8 @@ class WorkflowAppService:
         :param created_at_after: filter logs created after this timestamp
         :param page: page number
         :param limit: items per page
+        :param created_by_end_user_session_id: filter by end user session id
+        :param created_by_account: filter by account email
         :return: Pagination object
         """
         # Build base statement using SQLAlchemy 2.0 style
@@ -71,6 +75,26 @@ class WorkflowAppService:
         if created_at_after:
             stmt = stmt.where(WorkflowAppLog.created_at >= created_at_after)
 
+        # Filter by end user session id or account email
+        if created_by_end_user_session_id:
+            stmt = stmt.join(
+                EndUser,
+                and_(
+                    WorkflowAppLog.created_by == EndUser.id,
+                    WorkflowAppLog.created_by_role == CreatorUserRole.END_USER,
+                    EndUser.session_id == created_by_end_user_session_id,
+                ),
+            )
+        if created_by_account:
+            stmt = stmt.join(
+                Account,
+                and_(
+                    WorkflowAppLog.created_by == Account.id,
+                    WorkflowAppLog.created_by_role == CreatorUserRole.ACCOUNT,
+                    Account.email == created_by_account,
+                ),
+            )
+
         stmt = stmt.order_by(WorkflowAppLog.created_at.desc())
 
         # Get total count using the same filters

api/tasks/retry_document_indexing_task.py

@@ -30,11 +30,11 @@ def retry_document_indexing_task(dataset_id: str, document_ids: list[str]):
         logging.info(click.style("Dataset not found: {}".format(dataset_id), fg="red"))
         db.session.close()
         return
-
+    tenant_id = dataset.tenant_id
     for document_id in document_ids:
         retry_indexing_cache_key = "document_{}_is_retried".format(document_id)
         # check document limit
-        features = FeatureService.get_features(dataset.tenant_id)
+        features = FeatureService.get_features(tenant_id)
         try:
             if features.billing.enabled:
                 vector_space = features.vector_space

api/tests/integration_tests/vdb/matrixone/test_matrixone.py

@@ -0,0 +1,25 @@
+from core.rag.datasource.vdb.matrixone.matrixone_vector import MatrixoneConfig, MatrixoneVector
+from tests.integration_tests.vdb.test_vector_store import (
+    AbstractVectorTest,
+    get_example_text,
+    setup_mock_redis,
+)
+
+
+class MatrixoneVectorTest(AbstractVectorTest):
+    def __init__(self):
+        super().__init__()
+        self.vector = MatrixoneVector(
+            collection_name=self.collection_name,
+            config=MatrixoneConfig(
+                host="localhost", port=6001, user="dump", password="111", database="dify", metric="l2"
+            ),
+        )
+
+    def get_ids_by_metadata_field(self):
+        ids = self.vector.get_ids_by_metadata_field(key="document_id", value=self.example_doc_id)
+        assert len(ids) == 1
+
+
+def test_matrixone_vector(setup_mock_redis):
+    MatrixoneVectorTest().run_all_tests()

api/tests/integration_tests/vdb/oceanbase/test_oceanbase.py

@@ -1,15 +1,11 @@
-from unittest.mock import MagicMock, patch
-
 import pytest
 
 from core.rag.datasource.vdb.oceanbase.oceanbase_vector import (
     OceanBaseVector,
     OceanBaseVectorConfig,
 )
-from tests.integration_tests.vdb.__mock.tcvectordb import setup_tcvectordb_mock
 from tests.integration_tests.vdb.test_vector_store import (
     AbstractVectorTest,
-    get_example_text,
     setup_mock_redis,
 )
 
@@ -20,10 +16,11 @@ def oceanbase_vector():
         "dify_test_collection",
         config=OceanBaseVectorConfig(
             host="127.0.0.1",
-            port="2881",
-            user="root@test",
+            port=2881,
+            user="root",
             database="test",
-            password="test",
+            password="difyai123456",
+            enable_hybrid_search=True,
         ),
     )
 
@@ -33,39 +30,13 @@ class OceanBaseVectorTest(AbstractVectorTest):
         super().__init__()
         self.vector = vector
 
-    def search_by_vector(self):
-        hits_by_vector = self.vector.search_by_vector(query_vector=self.example_embedding)
-        assert len(hits_by_vector) == 0
-
-    def search_by_full_text(self):
-        hits_by_full_text = self.vector.search_by_full_text(query=get_example_text())
-        assert len(hits_by_full_text) == 0
-
-    def text_exists(self):
-        exist = self.vector.text_exists(self.example_doc_id)
-        assert exist == True
-
     def get_ids_by_metadata_field(self):
         ids = self.vector.get_ids_by_metadata_field(key="document_id", value=self.example_doc_id)
-        assert len(ids) == 0
-
-
-@pytest.fixture
-def setup_mock_oceanbase_client():
-    with patch("core.rag.datasource.vdb.oceanbase.oceanbase_vector.ObVecClient", new_callable=MagicMock) as mock_client:
-        yield mock_client
-
-
-@pytest.fixture
-def setup_mock_oceanbase_vector(oceanbase_vector):
-    with patch.object(oceanbase_vector, "_client"):
-        yield oceanbase_vector
+        assert len(ids) == 1
 
 
 def test_oceanbase_vector(
     setup_mock_redis,
-    setup_mock_oceanbase_client,
-    setup_mock_oceanbase_vector,
     oceanbase_vector,
 ):
     OceanBaseVectorTest(oceanbase_vector).run_all_tests()

api/tests/unit_tests/libs/test_flask_utils.py

@@ -0,0 +1,124 @@
+import contextvars
+import threading
+from typing import Optional
+
+import pytest
+from flask import Flask
+from flask_login import LoginManager, UserMixin, current_user, login_user
+
+from libs.flask_utils import preserve_flask_contexts
+
+
+class User(UserMixin):
+    """Simple User class for testing."""
+
+    def __init__(self, id: str):
+        self.id = id
+
+    def get_id(self) -> str:
+        return self.id
+
+
+@pytest.fixture
+def login_app(app: Flask) -> Flask:
+    """Set up a Flask app with flask-login."""
+    # Set a secret key for the app
+    app.config["SECRET_KEY"] = "test-secret-key"
+
+    login_manager = LoginManager()
+    login_manager.init_app(app)
+
+    @login_manager.user_loader
+    def load_user(user_id: str) -> Optional[User]:
+        if user_id == "test_user":
+            return User("test_user")
+        return None
+
+    return app
+
+
+@pytest.fixture
+def test_user() -> User:
+    """Create a test user."""
+    return User("test_user")
+
+
+def test_current_user_not_accessible_across_threads(login_app: Flask, test_user: User):
+    """
+    Test that current_user is not accessible in a different thread without preserve_flask_contexts.
+
+    This test demonstrates that without the preserve_flask_contexts, we cannot access
+    current_user in a different thread, even with app_context.
+    """
+    # Log in the user in the main thread
+    with login_app.test_request_context():
+        login_user(test_user)
+        assert current_user.is_authenticated
+        assert current_user.id == "test_user"
+
+        # Store the result of the thread execution
+        result = {"user_accessible": True, "error": None}
+
+        # Define a function to run in a separate thread
+        def check_user_in_thread():
+            try:
+                # Try to access current_user in a different thread with app_context
+                with login_app.app_context():
+                    # This should fail because current_user is not accessible across threads
+                    # without preserve_flask_contexts
+                    result["user_accessible"] = current_user.is_authenticated
+            except Exception as e:
+                result["error"] = str(e)  # type: ignore
+
+        # Run the function in a separate thread
+        thread = threading.Thread(target=check_user_in_thread)
+        thread.start()
+        thread.join()
+
+        # Verify that we got an error or current_user is not authenticated
+        assert result["error"] is not None or (result["user_accessible"] is not None and not result["user_accessible"])
+
+
+def test_current_user_accessible_with_preserve_flask_contexts(login_app: Flask, test_user: User):
+    """
+    Test that current_user is accessible in a different thread with preserve_flask_contexts.
+
+    This test demonstrates that with the preserve_flask_contexts, we can access
+    current_user in a different thread.
+    """
+    # Log in the user in the main thread
+    with login_app.test_request_context():
+        login_user(test_user)
+        assert current_user.is_authenticated
+        assert current_user.id == "test_user"
+
+        # Save the context variables
+        context_vars = contextvars.copy_context()
+
+        # Store the result of the thread execution
+        result = {"user_accessible": False, "user_id": None, "error": None}
+
+        # Define a function to run in a separate thread
+        def check_user_in_thread_with_manager():
+            try:
+                # Use preserve_flask_contexts to access current_user in a different thread
+                with preserve_flask_contexts(login_app, context_vars):
+                    from flask_login import current_user
+
+                    if current_user:
+                        result["user_accessible"] = True
+                        result["user_id"] = current_user.id
+                    else:
+                        result["user_accessible"] = False
+            except Exception as e:
+                result["error"] = str(e)  # type: ignore
+
+        # Run the function in a separate thread
+        thread = threading.Thread(target=check_user_in_thread_with_manager)
+        thread.start()
+        thread.join()
+
+        # Verify that current_user is accessible and has the correct ID
+        assert result["error"] is None
+        assert result["user_accessible"] is True
+        assert result["user_id"] == "test_user"

docker/.env.example

@@ -399,7 +399,7 @@ SUPABASE_URL=your-server-url
 # ------------------------------
 
 # The type of vector store to use.
-# Supported values are `weaviate`, `qdrant`, `milvus`, `myscale`, `relyt`, `pgvector`, `pgvecto-rs`, `chroma`, `opensearch`, `oracle`, `tencent`, `elasticsearch`, `elasticsearch-ja`, `analyticdb`, `couchbase`, `vikingdb`, `oceanbase`, `opengauss`, `tablestore`,`vastbase`,`tidb`,`tidb_on_qdrant`,`baidu`,`lindorm`,`huawei_cloud`,`upstash`.
+# Supported values are `weaviate`, `qdrant`, `milvus`, `myscale`, `relyt`, `pgvector`, `pgvecto-rs`, `chroma`, `opensearch`, `oracle`, `tencent`, `elasticsearch`, `elasticsearch-ja`, `analyticdb`, `couchbase`, `vikingdb`, `oceanbase`, `opengauss`, `tablestore`,`vastbase`,`tidb`,`tidb_on_qdrant`,`baidu`,`lindorm`,`huawei_cloud`,`upstash`, `matrixone`.
 VECTOR_STORE=weaviate
 
 # The Weaviate endpoint URL. Only available when VECTOR_STORE is `weaviate`.
@@ -490,6 +490,13 @@ TIDB_VECTOR_USER=
 TIDB_VECTOR_PASSWORD=
 TIDB_VECTOR_DATABASE=dify
 
+# Matrixone vector configurations.
+MATRIXONE_HOST=matrixone
+MATRIXONE_PORT=6001
+MATRIXONE_USER=dump
+MATRIXONE_PASSWORD=111
+MATRIXONE_DATABASE=dify
+
 # Tidb on qdrant configuration, only available when VECTOR_STORE is `tidb_on_qdrant`
 TIDB_ON_QDRANT_URL=http://127.0.0.1
 TIDB_ON_QDRANT_API_KEY=dify
@@ -719,10 +726,11 @@ NOTION_INTERNAL_SECRET=
 # Mail related configuration
 # ------------------------------
 
-# Mail type, support: resend, smtp
+# Mail type, support: resend, smtp, sendgrid
 MAIL_TYPE=resend
 
 # Default send from email address, if not specified
+# If using SendGrid, use the 'from' field for authentication if necessary.
 MAIL_DEFAULT_SEND_FROM=
 
 # API-Key for the Resend email provider, used when MAIL_TYPE is `resend`.
@@ -738,6 +746,9 @@ SMTP_PASSWORD=
 SMTP_USE_TLS=true
 SMTP_OPPORTUNISTIC_TLS=false
 
+# Sendgid configuration
+SENDGRID_API_KEY=
+
 # ------------------------------
 # Others Configuration
 # ------------------------------
@@ -815,7 +826,8 @@ TEXT_GENERATION_TIMEOUT_MS=60000
 # Environment Variables for db Service
 # ------------------------------
 
-PGUSER=${DB_USERNAME}
+# The name of the default postgres user.
+POSTGRES_USER=${DB_USERNAME}
 # The password for the default postgres user.
 POSTGRES_PASSWORD=${DB_PASSWORD}
 # The name of the default postgres database.
@@ -1067,6 +1079,7 @@ PLUGIN_MEDIA_CACHE_PATH=assets
 # Plugin oss bucket
 PLUGIN_STORAGE_OSS_BUCKET=
 # Plugin oss s3 credentials
+PLUGIN_S3_USE_AWS=false
 PLUGIN_S3_USE_AWS_MANAGED_IAM=false
 PLUGIN_S3_ENDPOINT=
 PLUGIN_S3_USE_PATH_STYLE=false

docker/docker-compose-template.yaml

@@ -2,7 +2,7 @@ x-shared-env: &shared-api-worker-env
 services:
   # API service
   api:
-    image: langgenius/dify-api:1.4.2
+    image: langgenius/dify-api:1.4.3
     restart: always
     environment:
       # Use the shared environment variables.
@@ -31,7 +31,7 @@ services:
   # worker service
   # The Celery worker for processing the queue.
   worker:
-    image: langgenius/dify-api:1.4.2
+    image: langgenius/dify-api:1.4.3
     restart: always
     environment:
       # Use the shared environment variables.
@@ -57,7 +57,7 @@ services:
 
   # Frontend web application.
   web:
-    image: langgenius/dify-web:1.4.2
+    image: langgenius/dify-web:1.4.3
     restart: always
     environment:
       CONSOLE_API_URL: ${CONSOLE_API_URL:-}
@@ -84,7 +84,7 @@ services:
     image: postgres:15-alpine
     restart: always
     environment:
-      PGUSER: ${PGUSER:-postgres}
+      POSTGRES_USER: ${POSTGRES_USER:-postgres}
       POSTGRES_PASSWORD: ${POSTGRES_PASSWORD:-difyai123456}
       POSTGRES_DB: ${POSTGRES_DB:-dify}
       PGDATA: ${PGDATA:-/var/lib/postgresql/data/pgdata}
@@ -168,6 +168,7 @@ services:
       PLUGIN_MEDIA_CACHE_PATH: ${PLUGIN_MEDIA_CACHE_PATH:-assets}
       PLUGIN_STORAGE_OSS_BUCKET: ${PLUGIN_STORAGE_OSS_BUCKET:-}
       S3_USE_AWS_MANAGED_IAM: ${PLUGIN_S3_USE_AWS_MANAGED_IAM:-false}
+      S3_USE_AWS: ${PLUGIN_S3_USE_AWS:-}
       S3_ENDPOINT: ${PLUGIN_S3_ENDPOINT:-}
       S3_USE_PATH_STYLE: ${PLUGIN_S3_USE_PATH_STYLE:-false}
       AWS_ACCESS_KEY: ${PLUGIN_AWS_ACCESS_KEY:-}
@@ -434,7 +435,7 @@ services:
 
   # OceanBase vector database
   oceanbase:
-    image: oceanbase/oceanbase-ce:4.3.5.1-101000042025031818
+    image: oceanbase/oceanbase-ce:4.3.5-lts
     container_name: oceanbase
     profiles:
       - oceanbase
@@ -449,9 +450,15 @@ services:
       OB_TENANT_PASSWORD: ${OCEANBASE_VECTOR_PASSWORD:-difyai123456}
       OB_CLUSTER_NAME: ${OCEANBASE_CLUSTER_NAME:-difyai}
       OB_SERVER_IP: 127.0.0.1
-      MODE: MINI
+      MODE: mini
     ports:
       - "${OCEANBASE_VECTOR_PORT:-2881}:2881"
+    healthcheck:
+      test: [ 'CMD-SHELL', 'obclient -h127.0.0.1 -P2881 -uroot@test -p$${OB_TENANT_PASSWORD} -e "SELECT 1;"' ]
+      interval: 10s
+      retries: 30
+      start_period: 30s
+      timeout: 10s
 
   # Oracle vector database
   oracle:
@@ -610,6 +617,18 @@ services:
     ports:
       - ${MYSCALE_PORT:-8123}:${MYSCALE_PORT:-8123}
 
+  # Matrixone vector store.
+  matrixone:
+    hostname: matrixone
+    image: matrixorigin/matrixone:2.1.1
+    profiles:
+      - matrixone
+    restart: always
+    volumes:
+      - ./volumes/matrixone/data:/mo-data
+    ports:
+      - ${MATRIXONE_PORT:-6001}:${MATRIXONE_PORT:-6001}
+
   # https://www.elastic.co/guide/en/elasticsearch/reference/current/settings.html
   # https://www.elastic.co/guide/en/elasticsearch/reference/current/docker.html#docker-prod-prerequisites
   elasticsearch:

docker/docker-compose.middleware.yaml

@@ -104,6 +104,7 @@ services:
       PLUGIN_PACKAGE_CACHE_PATH: ${PLUGIN_PACKAGE_CACHE_PATH:-plugin_packages}
       PLUGIN_MEDIA_CACHE_PATH: ${PLUGIN_MEDIA_CACHE_PATH:-assets}
       PLUGIN_STORAGE_OSS_BUCKET: ${PLUGIN_STORAGE_OSS_BUCKET:-}
+      S3_USE_AWS: ${PLUGIN_S3_USE_AWS:-false}
       S3_USE_AWS_MANAGED_IAM: ${PLUGIN_S3_USE_AWS_MANAGED_IAM:-false}
       S3_ENDPOINT: ${PLUGIN_S3_ENDPOINT:-}
       S3_USE_PATH_STYLE: ${PLUGIN_S3_USE_PATH_STYLE:-false}

docker/docker-compose.yaml

@@ -195,6 +195,11 @@ x-shared-env: &shared-api-worker-env
   TIDB_VECTOR_USER: ${TIDB_VECTOR_USER:-}
   TIDB_VECTOR_PASSWORD: ${TIDB_VECTOR_PASSWORD:-}
   TIDB_VECTOR_DATABASE: ${TIDB_VECTOR_DATABASE:-dify}
+  MATRIXONE_HOST: ${MATRIXONE_HOST:-matrixone}
+  MATRIXONE_PORT: ${MATRIXONE_PORT:-6001}
+  MATRIXONE_USER: ${MATRIXONE_USER:-dump}
+  MATRIXONE_PASSWORD: ${MATRIXONE_PASSWORD:-111}
+  MATRIXONE_DATABASE: ${MATRIXONE_DATABASE:-dify}
   TIDB_ON_QDRANT_URL: ${TIDB_ON_QDRANT_URL:-http://127.0.0.1}
   TIDB_ON_QDRANT_API_KEY: ${TIDB_ON_QDRANT_API_KEY:-dify}
   TIDB_ON_QDRANT_CLIENT_TIMEOUT: ${TIDB_ON_QDRANT_CLIENT_TIMEOUT:-20}
@@ -322,6 +327,7 @@ x-shared-env: &shared-api-worker-env
   SMTP_PASSWORD: ${SMTP_PASSWORD:-}
   SMTP_USE_TLS: ${SMTP_USE_TLS:-true}
   SMTP_OPPORTUNISTIC_TLS: ${SMTP_OPPORTUNISTIC_TLS:-false}
+  SENDGRID_API_KEY: ${SENDGRID_API_KEY:-}
   INDEXING_MAX_SEGMENTATION_TOKENS_LENGTH: ${INDEXING_MAX_SEGMENTATION_TOKENS_LENGTH:-4000}
   INVITE_EXPIRY_HOURS: ${INVITE_EXPIRY_HOURS:-72}
   RESET_PASSWORD_TOKEN_EXPIRY_MINUTES: ${RESET_PASSWORD_TOKEN_EXPIRY_MINUTES:-5}
@@ -356,7 +362,7 @@ x-shared-env: &shared-api-worker-env
   MAX_PARALLEL_LIMIT: ${MAX_PARALLEL_LIMIT:-10}
   MAX_ITERATIONS_NUM: ${MAX_ITERATIONS_NUM:-99}
   TEXT_GENERATION_TIMEOUT_MS: ${TEXT_GENERATION_TIMEOUT_MS:-60000}
-  PGUSER: ${PGUSER:-${DB_USERNAME}}
+  POSTGRES_USER: ${POSTGRES_USER:-${DB_USERNAME}}
   POSTGRES_PASSWORD: ${POSTGRES_PASSWORD:-${DB_PASSWORD}}
   POSTGRES_DB: ${POSTGRES_DB:-${DB_DATABASE}}
   PGDATA: ${PGDATA:-/var/lib/postgresql/data/pgdata}
@@ -467,6 +473,7 @@ x-shared-env: &shared-api-worker-env
   PLUGIN_PACKAGE_CACHE_PATH: ${PLUGIN_PACKAGE_CACHE_PATH:-plugin_packages}
   PLUGIN_MEDIA_CACHE_PATH: ${PLUGIN_MEDIA_CACHE_PATH:-assets}
   PLUGIN_STORAGE_OSS_BUCKET: ${PLUGIN_STORAGE_OSS_BUCKET:-}
+  PLUGIN_S3_USE_AWS: ${PLUGIN_S3_USE_AWS:-false}
   PLUGIN_S3_USE_AWS_MANAGED_IAM: ${PLUGIN_S3_USE_AWS_MANAGED_IAM:-false}
   PLUGIN_S3_ENDPOINT: ${PLUGIN_S3_ENDPOINT:-}
   PLUGIN_S3_USE_PATH_STYLE: ${PLUGIN_S3_USE_PATH_STYLE:-false}
@@ -508,7 +515,7 @@ x-shared-env: &shared-api-worker-env
 services:
   # API service
   api:
-    image: langgenius/dify-api:1.4.2
+    image: langgenius/dify-api:1.4.3
     restart: always
     environment:
       # Use the shared environment variables.
@@ -537,7 +544,7 @@ services:
   # worker service
   # The Celery worker for processing the queue.
   worker:
-    image: langgenius/dify-api:1.4.2
+    image: langgenius/dify-api:1.4.3
     restart: always
     environment:
       # Use the shared environment variables.
@@ -563,7 +570,7 @@ services:
 
   # Frontend web application.
   web:
-    image: langgenius/dify-web:1.4.2
+    image: langgenius/dify-web:1.4.3
     restart: always
     environment:
       CONSOLE_API_URL: ${CONSOLE_API_URL:-}
@@ -590,7 +597,7 @@ services:
     image: postgres:15-alpine
     restart: always
     environment:
-      PGUSER: ${PGUSER:-postgres}
+      POSTGRES_USER: ${POSTGRES_USER:-postgres}
       POSTGRES_PASSWORD: ${POSTGRES_PASSWORD:-difyai123456}
       POSTGRES_DB: ${POSTGRES_DB:-dify}
       PGDATA: ${PGDATA:-/var/lib/postgresql/data/pgdata}
@@ -674,6 +681,7 @@ services:
       PLUGIN_MEDIA_CACHE_PATH: ${PLUGIN_MEDIA_CACHE_PATH:-assets}
       PLUGIN_STORAGE_OSS_BUCKET: ${PLUGIN_STORAGE_OSS_BUCKET:-}
       S3_USE_AWS_MANAGED_IAM: ${PLUGIN_S3_USE_AWS_MANAGED_IAM:-false}
+      S3_USE_AWS: ${PLUGIN_S3_USE_AWS:-}
       S3_ENDPOINT: ${PLUGIN_S3_ENDPOINT:-}
       S3_USE_PATH_STYLE: ${PLUGIN_S3_USE_PATH_STYLE:-false}
       AWS_ACCESS_KEY: ${PLUGIN_AWS_ACCESS_KEY:-}
@@ -940,7 +948,7 @@ services:
 
   # OceanBase vector database
   oceanbase:
-    image: oceanbase/oceanbase-ce:4.3.5.1-101000042025031818
+    image: oceanbase/oceanbase-ce:4.3.5-lts
     container_name: oceanbase
     profiles:
       - oceanbase
@@ -955,9 +963,15 @@ services:
       OB_TENANT_PASSWORD: ${OCEANBASE_VECTOR_PASSWORD:-difyai123456}
       OB_CLUSTER_NAME: ${OCEANBASE_CLUSTER_NAME:-difyai}
       OB_SERVER_IP: 127.0.0.1
-      MODE: MINI
+      MODE: mini
     ports:
       - "${OCEANBASE_VECTOR_PORT:-2881}:2881"
+    healthcheck:
+      test: [ 'CMD-SHELL', 'obclient -h127.0.0.1 -P2881 -uroot@test -p$${OB_TENANT_PASSWORD} -e "SELECT 1;"' ]
+      interval: 10s
+      retries: 30
+      start_period: 30s
+      timeout: 10s
 
   # Oracle vector database
   oracle:
@@ -1116,6 +1130,18 @@ services:
     ports:
       - ${MYSCALE_PORT:-8123}:${MYSCALE_PORT:-8123}
 
+  # Matrixone vector store.
+  matrixone:
+    hostname: matrixone
+    image: matrixorigin/matrixone:2.1.1
+    profiles:
+      - matrixone
+    restart: always
+    volumes:
+      - ./volumes/matrixone/data:/mo-data
+    ports:
+      - ${MATRIXONE_PORT:-6001}:${MATRIXONE_PORT:-6001}
+
   # https://www.elastic.co/guide/en/elasticsearch/reference/current/settings.html
   # https://www.elastic.co/guide/en/elasticsearch/reference/current/docker.html#docker-prod-prerequisites
   elasticsearch:

docker/middleware.env.example

@@ -1,7 +1,7 @@
 # ------------------------------
 # Environment Variables for db Service
 # ------------------------------
-PGUSER=postgres
+POSTGRES_USER=postgres
 # The password for the default postgres user.
 POSTGRES_PASSWORD=difyai123456
 # The name of the default postgres database.
@@ -133,6 +133,7 @@ PLUGIN_MEDIA_CACHE_PATH=assets
 PLUGIN_STORAGE_OSS_BUCKET=
 # Plugin oss s3 credentials
 PLUGIN_S3_USE_AWS_MANAGED_IAM=false
+PLUGIN_S3_USE_AWS=false
 PLUGIN_S3_ENDPOINT=
 PLUGIN_S3_USE_PATH_STYLE=false
 PLUGIN_AWS_ACCESS_KEY=

web/app/(commonLayout)/app/(appDetailLayout)/[appId]/overview/page.tsx

@@ -15,7 +15,7 @@ const Overview = async (props: IDevelopProps) => {
   } = params
 
   return (
-    <div className="h-full overflow-scroll bg-chatbot-bg px-4 py-6 sm:px-12">
+    <div className="h-full overflow-y-auto bg-chatbot-bg px-4 py-6 sm:px-12">
       <ApikeyInfoPanel />
       <ChartView
         appId={appId}

web/app/(commonLayout)/apps/Apps.tsx

@@ -9,6 +9,7 @@ import { useTranslation } from 'react-i18next'
 import { useDebounceFn } from 'ahooks'
 import {
   RiApps2Line,
+  RiDragDropLine,
   RiExchange2Line,
   RiFile4Line,
   RiMessage3Line,
@@ -16,7 +17,8 @@ import {
 } from '@remixicon/react'
 import AppCard from './AppCard'
 import NewAppCard from './NewAppCard'
-import useAppsQueryState from './hooks/useAppsQueryState'
+import useAppsQueryState from './hooks/use-apps-query-state'
+import { useDSLDragDrop } from './hooks/use-dsl-drag-drop'
 import type { AppListResponse } from '@/models/app'
 import { fetchAppList } from '@/service/apps'
 import { useAppContext } from '@/context/app-context'
@@ -29,6 +31,7 @@ import { useStore as useTagStore } from '@/app/components/base/tag-management/st
 import TagManagementModal from '@/app/components/base/tag-management'
 import TagFilter from '@/app/components/base/tag-management/filter'
 import CheckboxWithLabel from '@/app/components/datasets/create/website/base/checkbox-with-label'
+import CreateFromDSLModal from '@/app/components/app/create-from-dsl-modal'
 
 const getKey = (
   pageIndex: number,
@@ -67,6 +70,9 @@ const Apps = () => {
   const [tagFilterValue, setTagFilterValue] = useState<string[]>(tagIDs)
   const [searchKeywords, setSearchKeywords] = useState(keywords)
   const newAppCardRef = useRef<HTMLDivElement>(null)
+  const containerRef = useRef<HTMLDivElement>(null)
+  const [showCreateFromDSLModal, setShowCreateFromDSLModal] = useState(false)
+  const [droppedDSLFile, setDroppedDSLFile] = useState<File | undefined>()
   const setKeywords = useCallback((keywords: string) => {
     setQuery(prev => ({ ...prev, keywords }))
   }, [setQuery])
@@ -74,6 +80,17 @@ const Apps = () => {
     setQuery(prev => ({ ...prev, tagIDs }))
   }, [setQuery])
 
+  const handleDSLFileDropped = useCallback((file: File) => {
+    setDroppedDSLFile(file)
+    setShowCreateFromDSLModal(true)
+  }, [])
+
+  const { dragging } = useDSLDragDrop({
+    onDSLFileDropped: handleDSLFileDropped,
+    containerRef,
+    enabled: isCurrentWorkspaceEditor,
+  })
+
   const { data, isLoading, error, setSize, mutate } = useSWRInfinite(
     (pageIndex: number, previousPageData: AppListResponse) => getKey(pageIndex, previousPageData, activeTab, isCreatedByMe, tagIDs, searchKeywords),
     fetchAppList,
@@ -151,47 +168,81 @@ const Apps = () => {
 
   return (
     <>
-      <div className='sticky top-0 z-10 flex flex-wrap items-center justify-between gap-y-2 bg-background-body px-12 pb-2 pt-4 leading-[56px]'>
-        <TabSliderNew
-          value={activeTab}
-          onChange={setActiveTab}
-          options={options}
-        />
-        <div className='flex items-center gap-2'>
-          <CheckboxWithLabel
-            className='mr-2'
-            label={t('app.showMyCreatedAppsOnly')}
-            isChecked={isCreatedByMe}
-            onChange={handleCreatedByMeChange}
-          />
-          <TagFilter type='app' value={tagFilterValue} onChange={handleTagsChange} />
-          <Input
-            showLeftIcon
-            showClearIcon
-            wrapperClassName='w-[200px]'
-            value={keywords}
-            onChange={e => handleKeywordsChange(e.target.value)}
-            onClear={() => handleKeywordsChange('')}
+      <div ref={containerRef} className='relative flex h-0 shrink-0 grow flex-col overflow-y-auto bg-background-body'>
+        {dragging && (
+          <div className="absolute inset-0 z-50 m-0.5 rounded-2xl border-2 border-dashed border-components-dropzone-border-accent bg-[rgba(21,90,239,0.14)] p-2">
+          </div>
+        )}
+
+        <div className='sticky top-0 z-10 flex flex-wrap items-center justify-between gap-y-2 bg-background-body px-12 pb-2 pt-4 leading-[56px]'>
+          <TabSliderNew
+            value={activeTab}
+            onChange={setActiveTab}
+            options={options}
           />
+          <div className='flex items-center gap-2'>
+            <CheckboxWithLabel
+              className='mr-2'
+              label={t('app.showMyCreatedAppsOnly')}
+              isChecked={isCreatedByMe}
+              onChange={handleCreatedByMeChange}
+            />
+            <TagFilter type='app' value={tagFilterValue} onChange={handleTagsChange} />
+            <Input
+              showLeftIcon
+              showClearIcon
+              wrapperClassName='w-[200px]'
+              value={keywords}
+              onChange={e => handleKeywordsChange(e.target.value)}
+              onClear={() => handleKeywordsChange('')}
+            />
+          </div>
         </div>
+        {(data && data[0].total > 0)
+          ? <div className='relative grid grow grid-cols-1 content-start gap-4 px-12 pt-2 sm:grid-cols-1 md:grid-cols-2 xl:grid-cols-4 2xl:grid-cols-5 2k:grid-cols-6'>
+            {isCurrentWorkspaceEditor
+              && <NewAppCard ref={newAppCardRef} onSuccess={mutate} />}
+            {data.map(({ data: apps }) => apps.map(app => (
+              <AppCard key={app.id} app={app} onRefresh={mutate} />
+            )))}
+          </div>
+          : <div className='relative grid grow grid-cols-1 content-start gap-4 overflow-hidden px-12 pt-2 sm:grid-cols-1 md:grid-cols-2 xl:grid-cols-4 2xl:grid-cols-5 2k:grid-cols-6'>
+            {isCurrentWorkspaceEditor
+              && <NewAppCard ref={newAppCardRef} className='z-10' onSuccess={mutate} />}
+            <NoAppsFound />
+          </div>}
+
+        {isCurrentWorkspaceEditor && (
+          <div
+            className={`flex items-center justify-center gap-2 py-4 ${dragging ? 'text-text-accent' : 'text-text-quaternary'}`}
+            role="region"
+            aria-label={t('app.newApp.dropDSLToCreateApp')}
+          >
+            <RiDragDropLine className="h-4 w-4" />
+            <span className="system-xs-regular">{t('app.newApp.dropDSLToCreateApp')}</span>
+          </div>
+        )}
+        <CheckModal />
+        <div ref={anchorRef} className='h-0'> </div>
+        {showTagManagementModal && (
+          <TagManagementModal type='app' show={showTagManagementModal} />
+        )}
       </div>
-      {(data && data[0].total > 0)
-        ? <div className='relative grid grow grid-cols-1 content-start gap-4 px-12 pt-2 sm:grid-cols-1 md:grid-cols-2 xl:grid-cols-4 2xl:grid-cols-5 2k:grid-cols-6'>
-          {isCurrentWorkspaceEditor
-            && <NewAppCard ref={newAppCardRef} onSuccess={mutate} />}
-          {data.map(({ data: apps }) => apps.map(app => (
-            <AppCard key={app.id} app={app} onRefresh={mutate} />
-          )))}
-        </div>
-        : <div className='relative grid grow grid-cols-1 content-start gap-4 overflow-hidden px-12 pt-2 sm:grid-cols-1 md:grid-cols-2 xl:grid-cols-4 2xl:grid-cols-5 2k:grid-cols-6'>
-          {isCurrentWorkspaceEditor
-            && <NewAppCard ref={newAppCardRef} className='z-10' onSuccess={mutate} />}
-          <NoAppsFound />
-        </div>}
-      <CheckModal />
-      <div ref={anchorRef} className='h-0'> </div>
-      {showTagManagementModal && (
-        <TagManagementModal type='app' show={showTagManagementModal} />
+
+      {showCreateFromDSLModal && (
+        <CreateFromDSLModal
+          show={showCreateFromDSLModal}
+          onClose={() => {
+            setShowCreateFromDSLModal(false)
+            setDroppedDSLFile(undefined)
+          }}
+          onSuccess={() => {
+            setShowCreateFromDSLModal(false)
+            setDroppedDSLFile(undefined)
+            mutate()
+          }}
+          droppedFile={droppedDSLFile}
+        />
       )}
     </>
   )

web/app/(commonLayout)/apps/hooks/use-dsl-drag-drop.ts

@@ -0,0 +1,72 @@
+import { useEffect, useState } from 'react'
+
+type DSLDragDropHookProps = {
+  onDSLFileDropped: (file: File) => void
+  containerRef: React.RefObject<HTMLDivElement>
+  enabled?: boolean
+}
+
+export const useDSLDragDrop = ({ onDSLFileDropped, containerRef, enabled = true }: DSLDragDropHookProps) => {
+  const [dragging, setDragging] = useState(false)
+
+  const handleDragEnter = (e: DragEvent) => {
+    e.preventDefault()
+    e.stopPropagation()
+    if (e.dataTransfer?.types.includes('Files'))
+      setDragging(true)
+  }
+
+  const handleDragOver = (e: DragEvent) => {
+    e.preventDefault()
+    e.stopPropagation()
+  }
+
+  const handleDragLeave = (e: DragEvent) => {
+    e.preventDefault()
+    e.stopPropagation()
+    if (e.relatedTarget === null || !containerRef.current?.contains(e.relatedTarget as Node))
+      setDragging(false)
+  }
+
+  const handleDrop = (e: DragEvent) => {
+    e.preventDefault()
+    e.stopPropagation()
+    setDragging(false)
+
+    if (!e.dataTransfer)
+      return
+
+    const files = [...e.dataTransfer.files]
+    if (files.length === 0)
+      return
+
+    const file = files[0]
+    if (file.name.toLowerCase().endsWith('.yaml') || file.name.toLowerCase().endsWith('.yml'))
+      onDSLFileDropped(file)
+  }
+
+  useEffect(() => {
+    if (!enabled)
+      return
+
+    const current = containerRef.current
+    if (current) {
+      current.addEventListener('dragenter', handleDragEnter)
+      current.addEventListener('dragover', handleDragOver)
+      current.addEventListener('dragleave', handleDragLeave)
+      current.addEventListener('drop', handleDrop)
+    }
+    return () => {
+      if (current) {
+        current.removeEventListener('dragenter', handleDragEnter)
+        current.removeEventListener('dragover', handleDragOver)
+        current.removeEventListener('dragleave', handleDragLeave)
+        current.removeEventListener('drop', handleDrop)
+      }
+    }
+  }, [containerRef, enabled])
+
+  return {
+    dragging: enabled ? dragging : false,
+  }
+}

web/app/(commonLayout)/datasets/(datasetDetailLayout)/[datasetId]/layout-main.tsx

@@ -25,9 +25,8 @@ import Loading from '@/app/components/base/loading'
 import DatasetDetailContext from '@/context/dataset-detail'
 import { DataSourceType } from '@/models/datasets'
 import useBreakpoints, { MediaType } from '@/hooks/use-breakpoints'
-import { LanguagesSupported } from '@/i18n/language'
 import { useStore } from '@/app/components/app/store'
-import { getLocaleOnClient } from '@/i18n'
+import { useDocLink } from '@/context/i18n'
 import { useAppContext } from '@/context/app-context'
 import Tooltip from '@/app/components/base/tooltip'
 import LinkedAppsPanel from '@/app/components/base/linked-apps-panel'
@@ -45,9 +44,9 @@ type IExtraInfoProps = {
 }
 
 const ExtraInfo = ({ isMobile, relatedApps, expand }: IExtraInfoProps) => {
-  const locale = getLocaleOnClient()
   const [isShowTips, { toggle: toggleTips, set: setShowTips }] = useBoolean(!isMobile)
   const { t } = useTranslation()
+  const docLink = useDocLink()
 
   const hasRelatedApps = relatedApps?.data && relatedApps?.data?.length > 0
   const relatedAppsTotal = relatedApps?.data?.length || 0
@@ -97,11 +96,7 @@ const ExtraInfo = ({ isMobile, relatedApps, expand }: IExtraInfoProps) => {
             <div className='my-2 text-xs text-text-tertiary'>{t('common.datasetMenus.emptyTip')}</div>
             <a
               className='mt-2 inline-flex cursor-pointer items-center text-xs text-text-accent'
-              href={
-                locale === LanguagesSupported[1]
-                  ? 'https://docs.dify.ai/zh-hans/guides/knowledge-base/integrate-knowledge-within-application'
-                  : 'https://docs.dify.ai/guides/knowledge-base/integrate-knowledge-within-application'
-              }
+              href={docLink('/guides/knowledge-base/integrate-knowledge-within-application')}
               target='_blank' rel='noopener noreferrer'
             >
               <RiBookOpenLine className='mr-1 text-text-accent' />

web/app/(commonLayout)/datasets/Datasets.tsx

@@ -81,7 +81,7 @@ const Datasets = ({
       currentContainer?.removeEventListener('scroll', onScroll)
       onScroll.cancel()
     }
-  }, [onScroll])
+  }, [containerRef, onScroll])
 
   return (
     <nav className='grid shrink-0 grow grid-cols-1 content-start gap-4 px-12 pt-2 sm:grid-cols-2 md:grid-cols-3 lg:grid-cols-4'>

web/app/(commonLayout)/datasets/NewDatasetCard.tsx

@@ -5,34 +5,34 @@ import {
   RiAddLine,
   RiArrowRightLine,
 } from '@remixicon/react'
+import Link from 'next/link'
 
-const CreateAppCard = (
-  {
-    ref,
-    ..._
-  },
-) => {
+type CreateAppCardProps = {
+  ref?: React.Ref<HTMLAnchorElement>
+}
+
+const CreateAppCard = ({ ref }: CreateAppCardProps) => {
   const { t } = useTranslation()
 
   return (
     <div className='bg-background-default-dimm flex min-h-[160px] flex-col rounded-xl border-[0.5px]
       border-components-panel-border transition-all duration-200 ease-in-out'
     >
-      <a ref={ref} className='group flex grow cursor-pointer items-start p-4' href={`${basePath}/datasets/create`}>
+      <Link ref={ref} className='group flex grow cursor-pointer items-start p-4' href={`${basePath}/datasets/create`}>
         <div className='flex items-center gap-3'>
           <div className='flex h-10 w-10 items-center justify-center rounded-lg border border-dashed border-divider-regular bg-background-default-lighter
             p-2 group-hover:border-solid group-hover:border-effects-highlight group-hover:bg-background-default-dodge'
           >
-            <RiAddLine className='h-4 w-4 text-text-tertiary group-hover:text-text-accent'/>
+            <RiAddLine className='h-4 w-4 text-text-tertiary group-hover:text-text-accent' />
           </div>
           <div className='system-md-semibold text-text-secondary group-hover:text-text-accent'>{t('dataset.createDataset')}</div>
         </div>
-      </a>
+      </Link>
       <div className='system-xs-regular p-4 pt-0 text-text-tertiary'>{t('dataset.createDatasetIntro')}</div>
-      <a className='group flex cursor-pointer items-center gap-1 rounded-b-xl border-t-[0.5px] border-divider-subtle p-4' href={`${basePath}/datasets/connect`}>
+      <Link className='group flex cursor-pointer items-center gap-1 rounded-b-xl border-t-[0.5px] border-divider-subtle p-4' href={`${basePath}/datasets/connect`}>
         <div className='system-xs-medium text-text-tertiary group-hover:text-text-accent'>{t('dataset.connectDataset')}</div>
         <RiArrowRightLine className='h-3.5 w-3.5 text-text-tertiary group-hover:text-text-accent' />
-      </a>
+      </Link>
     </div>
   )
 }

web/app/(commonLayout)/datasets/layout.tsx

@@ -1,9 +1,25 @@
 'use client'
 
+import Loading from '@/app/components/base/loading'
+import { useAppContext } from '@/context/app-context'
 import { ExternalApiPanelProvider } from '@/context/external-api-panel-context'
 import { ExternalKnowledgeApiProvider } from '@/context/external-knowledge-api-context'
+import { useRouter } from 'next/navigation'
+import { useEffect } from 'react'
 
 export default function DatasetsLayout({ children }: { children: React.ReactNode }) {
+  const { isCurrentWorkspaceEditor, isCurrentWorkspaceDatasetOperator, currentWorkspace, isLoadingCurrentWorkspace } = useAppContext()
+  const router = useRouter()
+
+  useEffect(() => {
+    if (isLoadingCurrentWorkspace || !currentWorkspace.id)
+      return
+    if (!(isCurrentWorkspaceEditor || isCurrentWorkspaceDatasetOperator))
+      router.replace('/apps')
+  }, [isCurrentWorkspaceEditor, isCurrentWorkspaceDatasetOperator, isLoadingCurrentWorkspace, currentWorkspace, router])
+
+  if (isLoadingCurrentWorkspace || !(isCurrentWorkspaceEditor || isCurrentWorkspaceDatasetOperator))
+    return <Loading type='app' />
   return (
     <ExternalKnowledgeApiProvider>
       <ExternalApiPanelProvider>

web/app/(commonLayout)/datasets/template/template.en.mdx

@@ -54,7 +54,7 @@ import { Row, Col, Properties, Property, Heading, SubProperty, PropertyInstructi
       </Property>
       <Property name='indexing_technique' type='string' key='indexing_technique'>
         Index mode
-          - <code>high_quality</code> High quality: embedding using embedding model, built as vector database index
+          - <code>high_quality</code> High quality: Embedding using embedding model, built as vector database index
           - <code>economy</code> Economy: Build using inverted index of keyword table index
       </Property>
       <Property name='doc_form' type='string' key='doc_form'>

web/app/(commonLayout)/datasets/template/template.zh.mdx

@@ -55,7 +55,7 @@ import { Row, Col, Properties, Property, Heading, SubProperty, PropertyInstructi
       <Property name='indexing_technique' type='string' key='indexing_technique'>
         索引方式
           - <code>high_quality</code> 高质量：使用  
-        ding 模型进行嵌入，构建为向量数据库索引
+        Embedding 模型进行嵌入，构建为向量数据库索引
           - <code>economy</code> 经济：使用 keyword table index 的倒排索引进行构建
       </Property>
       <Property name='doc_form' type='string' key='doc_form'>

web/app/(shareLayout)/layout.tsx

@@ -19,7 +19,7 @@ const Layout: FC<{
   const [isLoading, setIsLoading] = useState(true)
   useEffect(() => {
     (async () => {
-      if (!systemFeatures.webapp_auth.enabled) {
+      if (!isGlobalPending && !systemFeatures.webapp_auth.enabled) {
         setIsLoading(false)
         return
       }
@@ -37,7 +37,7 @@ const Layout: FC<{
       setWebAppAccessMode(ret?.accessMode || AccessMode.PUBLIC)
       setIsLoading(false)
     })()
-  }, [pathname, redirectUrl, setWebAppAccessMode])
+  }, [pathname, redirectUrl, setWebAppAccessMode, isGlobalPending, systemFeatures.webapp_auth.enabled])
   if (isLoading || isGlobalPending) {
     return <div className='flex h-full w-full items-center justify-center'>
       <Loading />

web/app/components/app/app-publisher/index.tsx

@@ -314,10 +314,10 @@ const AppPublisher = ({
                   {!isAppAccessSet && <p className='system-xs-regular mt-1 text-text-warning'>{t('app.publishApp.notSetDesc')}</p>}
                 </div>}
                 <div className='flex flex-col gap-y-1 border-t-[0.5px] border-t-divider-regular p-4 pt-3'>
-                  <Tooltip triggerClassName='flex' disabled={!systemFeatures.webapp_auth.enabled || userCanAccessApp?.result} popupContent={t('app.noAccessPermission')} asChild={false}>
+                  <Tooltip triggerClassName='flex' disabled={!systemFeatures.webapp_auth.enabled || appDetail?.access_mode === AccessMode.EXTERNAL_MEMBERS || userCanAccessApp?.result} popupContent={t('app.noAccessPermission')} asChild={false}>
                     <SuggestedAction
                       className='flex-1'
-                      disabled={!publishedAt || (systemFeatures.webapp_auth.enabled && !userCanAccessApp?.result)}
+                      disabled={!publishedAt || (systemFeatures.webapp_auth.enabled && appDetail?.access_mode !== AccessMode.EXTERNAL_MEMBERS && !userCanAccessApp?.result)}
                       link={appURL}
                       icon={<RiPlayCircleLine className='h-4 w-4' />}
                     >
@@ -326,10 +326,10 @@ const AppPublisher = ({
                   </Tooltip>
                   {appDetail?.mode === 'workflow' || appDetail?.mode === 'completion'
                     ? (
-                      <Tooltip triggerClassName='flex' disabled={!systemFeatures.webapp_auth.enabled || userCanAccessApp?.result} popupContent={t('app.noAccessPermission')} asChild={false}>
+                      <Tooltip triggerClassName='flex' disabled={!systemFeatures.webapp_auth.enabled || appDetail.access_mode === AccessMode.EXTERNAL_MEMBERS || userCanAccessApp?.result} popupContent={t('app.noAccessPermission')} asChild={false}>
                         <SuggestedAction
                           className='flex-1'
-                          disabled={!publishedAt || (systemFeatures.webapp_auth.enabled && !userCanAccessApp?.result)}
+                          disabled={!publishedAt || (systemFeatures.webapp_auth.enabled && appDetail.access_mode !== AccessMode.EXTERNAL_MEMBERS && !userCanAccessApp?.result)}
                           link={`${appURL}${appURL.includes('?') ? '&' : '?'}mode=batch`}
                           icon={<RiPlayList2Line className='h-4 w-4' />}
                         >

web/app/components/app/configuration/config-prompt/conversation-history/history-panel.tsx

@@ -1,13 +1,11 @@
 'use client'
 import type { FC } from 'react'
 import React from 'react'
-import { useContext } from 'use-context-selector'
 import { useTranslation } from 'react-i18next'
 import OperationBtn from '@/app/components/app/configuration/base/operation-btn'
 import Panel from '@/app/components/app/configuration/base/feature-panel'
 import { MessageClockCircle } from '@/app/components/base/icons/src/vender/solid/general'
-import I18n from '@/context/i18n'
-import { LanguagesSupported } from '@/i18n/language'
+import { useDocLink } from '@/context/i18n'
 
 type Props = {
   showWarning: boolean
@@ -19,7 +17,7 @@ const HistoryPanel: FC<Props> = ({
   onShowEditModal,
 }) => {
   const { t } = useTranslation()
-  const { locale } = useContext(I18n)
+  const docLink = useDocLink()
 
   return (
     <Panel
@@ -45,9 +43,8 @@ const HistoryPanel: FC<Props> = ({
       {showWarning && (
         <div className='flex justify-between rounded-b-xl bg-background-section-burn px-3 py-2 text-xs text-text-secondary'>
           <div>{t('appDebug.feature.conversationHistory.tip')}
-            <a href={`${locale === LanguagesSupported[1]
-              ? 'https://docs.dify.ai/zh-hans/learn-more/extended-reading/prompt-engineering/README'
-              : 'https://docs.dify.ai/en/features/prompt-engineering'}`}
+            <a href={docLink('/learn-more/extended-reading/what-is-llmops',
+              { 'zh-Hans': '/learn-more/extended-reading/prompt-engineering/README' })}
             target='_blank' rel='noopener noreferrer'
             className='text-[#155EEF]'>{t('appDebug.feature.conversationHistory.learnMore')}
             </a>

web/app/components/app/configuration/dataset-config/settings-modal/index.tsx

@@ -31,6 +31,7 @@ import {
 import { ModelTypeEnum } from '@/app/components/header/account-setting/model-provider-page/declarations'
 import { fetchMembers } from '@/service/common'
 import type { Member } from '@/models/common'
+import { useDocLink } from '@/context/i18n'
 
 type SettingsModalProps = {
   currentDataset: DataSet
@@ -58,6 +59,7 @@ const SettingsModal: FC<SettingsModalProps> = ({
     currentModel: isRerankDefaultModelValid,
   } = useModelListAndDefaultModelAndCurrentProviderAndModel(ModelTypeEnum.rerank)
   const { t } = useTranslation()
+  const docLink = useDocLink()
   const { notify } = useToastContext()
   const ref = useRef(null)
   const isExternal = currentDataset.provider === 'external'
@@ -328,7 +330,7 @@ const SettingsModal: FC<SettingsModalProps> = ({
               <div>
                 <div className='system-sm-semibold text-text-secondary'>{t('datasetSettings.form.retrievalSetting.title')}</div>
                 <div className='text-xs font-normal leading-[18px] text-text-tertiary'>
-                  <a target='_blank' rel='noopener noreferrer' href='https://docs.dify.ai/guides/knowledge-base/create-knowledge-and-upload-documents#id-4-retrieval-settings' className='text-text-accent'>{t('datasetSettings.form.retrievalSetting.learnMore')}</a>
+                  <a target='_blank' rel='noopener noreferrer' href={docLink('/guides/knowledge-base/create-knowledge-and-upload-documents/setting-indexing-methods#setting-the-retrieval-setting')} className='text-text-accent'>{t('datasetSettings.form.retrievalSetting.learnMore')}</a>
                   {t('datasetSettings.form.retrievalSetting.description')}
                 </div>
               </div>

web/app/components/app/configuration/debug/index.tsx

@@ -156,12 +156,11 @@ const Debug: FC<IDebug> = ({
     }
     let hasEmptyInput = ''
     const requiredVars = modelConfig.configs.prompt_variables.filter(({ key, name, required, type }) => {
-      if (type !== 'string' && type !== 'paragraph' && type !== 'select')
+      if (type !== 'string' && type !== 'paragraph' && type !== 'select' && type !== 'number')
         return false
       const res = (!key || !key.trim()) || (!name || !name.trim()) || (required || required === undefined || required === null)
       return res
     }) // compatible with old version
-    // debugger
     requiredVars.forEach(({ key, name }) => {
       if (hasEmptyInput)
         return

web/app/components/app/configuration/prompt-mode/advanced-mode-waring.tsx

@@ -2,9 +2,7 @@
 import type { FC } from 'react'
 import React from 'react'
 import { useTranslation } from 'react-i18next'
-import { useContext } from 'use-context-selector'
-import I18n from '@/context/i18n'
-import { LanguagesSupported } from '@/i18n/language'
+import { useDocLink } from '@/context/i18n'
 type Props = {
   onReturnToSimpleMode: () => void
 }
@@ -13,7 +11,7 @@ const AdvancedModeWarning: FC<Props> = ({
   onReturnToSimpleMode,
 }) => {
   const { t } = useTranslation()
-  const { locale } = useContext(I18n)
+  const docLink = useDocLink()
   const [show, setShow] = React.useState(true)
   if (!show)
     return null
@@ -25,7 +23,7 @@ const AdvancedModeWarning: FC<Props> = ({
           <span className='text-gray-700'>{t('appDebug.promptMode.advancedWarning.description')}</span>
           <a
             className='font-medium text-[#155EEF]'
-            href={`https://docs.dify.ai/${locale === LanguagesSupported[1] ? '/guides/features/prompt-engineering' : 'features/prompt-engineering'}`}
+            href={docLink('/guides/features/prompt-engineering')}
             target='_blank' rel='noopener noreferrer'
           >
             {t('appDebug.promptMode.advancedWarning.learnMore')}

web/app/components/app/configuration/tools/external-data-tool-modal.tsx

@@ -20,6 +20,7 @@ import type {
 import { useToastContext } from '@/app/components/base/toast'
 import AppIcon from '@/app/components/base/app-icon'
 import { noop } from 'lodash-es'
+import { useDocLink } from '@/context/i18n'
 
 const systemTypes = ['api']
 type ExternalDataToolModalProps = {
@@ -40,6 +41,7 @@ const ExternalDataToolModal: FC<ExternalDataToolModalProps> = ({
   onValidateBeforeSave,
 }) => {
   const { t } = useTranslation()
+  const docLink = useDocLink()
   const { notify } = useToastContext()
   const { locale } = useContext(I18n)
   const [localeData, setLocaleData] = useState(data.type ? data : { ...data, type: 'api' })
@@ -243,7 +245,7 @@ const ExternalDataToolModal: FC<ExternalDataToolModalProps> = ({
             <div className='flex h-9 items-center justify-between text-sm font-medium text-gray-900'>
               {t('common.apiBasedExtension.selector.title')}
               <a
-                href={t('common.apiBasedExtension.linkUrl') || '/'}
+                href={docLink('/guides/extension/api-based-extension/README')}
                 target='_blank' rel='noopener noreferrer'
                 className='group flex items-center text-xs font-normal text-gray-500 hover:text-primary-600'
               >

web/app/components/app/create-app-modal/index.tsx

@@ -29,6 +29,7 @@ import { NEED_REFRESH_APP_LIST_KEY } from '@/config'
 import { getRedirection } from '@/utils/app-redirection'
 import FullScreenModal from '@/app/components/base/fullscreen-modal'
 import useTheme from '@/hooks/use-theme'
+import { useDocLink } from '@/context/i18n'
 
 type CreateAppProps = {
   onSuccess: () => void
@@ -303,31 +304,41 @@ function AppTypeCard({ icon, title, description, active, onClick }: AppTypeCardP
 
 function AppPreview({ mode }: { mode: AppMode }) {
   const { t } = useTranslation()
+  const docLink = useDocLink()
   const modeToPreviewInfoMap = {
     'chat': {
       title: t('app.types.chatbot'),
       description: t('app.newApp.chatbotUserDescription'),
-      link: 'https://docs.dify.ai/guides/application-orchestrate/readme',
+      link: docLink('/guides/application-orchestrate/chatbot-application'),
     },
     'advanced-chat': {
       title: t('app.types.advanced'),
       description: t('app.newApp.advancedUserDescription'),
-      link: 'https://docs.dify.ai/en/guides/workflow/README',
+      link: docLink('/guides/workflow/README', {
+        'zh-Hans': '/guides/workflow/readme',
+        'ja-JP': '/guides/workflow/concepts',
+      }),
     },
     'agent-chat': {
       title: t('app.types.agent'),
       description: t('app.newApp.agentUserDescription'),
-      link: 'https://docs.dify.ai/en/guides/application-orchestrate/agent',
+      link: docLink('/guides/application-orchestrate/agent'),
     },
     'completion': {
       title: t('app.newApp.completeApp'),
       description: t('app.newApp.completionUserDescription'),
-      link: null,
+      link: docLink('/guides/application-orchestrate/text-generator', {
+        'zh-Hans': '/guides/application-orchestrate/readme',
+        'ja-JP': '/guides/application-orchestrate/README',
+      }),
     },
     'workflow': {
       title: t('app.types.workflow'),
       description: t('app.newApp.workflowUserDescription'),
-      link: 'https://docs.dify.ai/en/guides/workflow/README',
+      link: docLink('/guides/workflow/README', {
+        'zh-Hans': '/guides/workflow/readme',
+        'ja-JP': '/guides/workflow/concepts',
+      }),
     },
   }
   const previewInfo = modeToPreviewInfoMap[mode]

web/app/components/app/create-from-dsl-modal/index.tsx

@@ -1,7 +1,7 @@
 'use client'
 
 import type { MouseEventHandler } from 'react'
-import { useMemo, useRef, useState } from 'react'
+import { useEffect, useMemo, useRef, useState } from 'react'
 import { useRouter } from 'next/navigation'
 import { useContext } from 'use-context-selector'
 import { useTranslation } from 'react-i18next'
@@ -35,6 +35,7 @@ type CreateFromDSLModalProps = {
   onClose: () => void
   activeTab?: string
   dslUrl?: string
+  droppedFile?: File
 }
 
 export enum CreateFromDSLModalTab {
@@ -42,11 +43,11 @@ export enum CreateFromDSLModalTab {
   FROM_URL = 'from-url',
 }
 
-const CreateFromDSLModal = ({ show, onSuccess, onClose, activeTab = CreateFromDSLModalTab.FROM_FILE, dslUrl = '' }: CreateFromDSLModalProps) => {
+const CreateFromDSLModal = ({ show, onSuccess, onClose, activeTab = CreateFromDSLModalTab.FROM_FILE, dslUrl = '', droppedFile }: CreateFromDSLModalProps) => {
   const { push } = useRouter()
   const { t } = useTranslation()
   const { notify } = useContext(ToastContext)
-  const [currentFile, setDSLFile] = useState<File>()
+  const [currentFile, setDSLFile] = useState<File | undefined>(droppedFile)
   const [fileContent, setFileContent] = useState<string>()
   const [currentTab, setCurrentTab] = useState(activeTab)
   const [dslUrlValue, setDslUrlValue] = useState(dslUrl)
@@ -78,6 +79,11 @@ const CreateFromDSLModal = ({ show, onSuccess, onClose, activeTab = CreateFromDS
 
   const isCreatingRef = useRef(false)
 
+  useEffect(() => {
+    if (droppedFile)
+      handleFile(droppedFile)
+  }, [droppedFile])
+
   const onCreate: MouseEventHandler = async () => {
     if (currentTab === CreateFromDSLModalTab.FROM_FILE && !currentFile)
       return