merge hitl

api/core/rag/datasource/keyword/jieba/jieba.py

@@ -245,6 +245,7 @@ class Jieba(BaseKeyword):
             segment = pre_segment_data["segment"]
             if pre_segment_data["keywords"]:
                 segment.keywords = pre_segment_data["keywords"]
+                assert segment.index_node_id
                 keyword_table = self._add_text_to_keyword_table(
                     keyword_table or {}, segment.index_node_id, pre_segment_data["keywords"]
                 )
@@ -253,6 +254,7 @@ class Jieba(BaseKeyword):
 
                 keywords = keyword_table_handler.extract_keywords(segment.content, keyword_number)
                 segment.keywords = list(keywords)
+                assert segment.index_node_id
                 keyword_table = self._add_text_to_keyword_table(
                     keyword_table or {}, segment.index_node_id, list(keywords)
                 )

api/core/rag/datasource/retrieval_service.py

@@ -1,5 +1,6 @@
 import concurrent.futures
 import logging
+from collections.abc import Sequence
 from concurrent.futures import ThreadPoolExecutor
 from typing import Any, NotRequired, TypedDict
 
@@ -502,7 +503,7 @@ class RetrievalService:
             index_node_ids = [i for i in index_node_ids if i]
 
             segment_ids: list[str] = []
-            index_node_segments: list[DocumentSegment] = []
+            index_node_segments: Sequence[DocumentSegment] = []
             segments: list[DocumentSegment] = []
             attachment_map: dict[str, list[AttachmentInfoDict]] = {}
             child_chunk_map: dict[str, list[ChildChunk]] = {}
@@ -544,8 +545,9 @@ class RetrievalService:
                         DocumentSegment.status == "completed",
                         DocumentSegment.index_node_id.in_(index_node_ids),
                     )
-                    index_node_segments = session.execute(document_segment_stmt).scalars().all()  # type: ignore
+                    index_node_segments = session.execute(document_segment_stmt).scalars().all()
                     for index_node_segment in index_node_segments:
+                        assert index_node_segment.index_node_id
                         doc_segment_map[index_node_segment.id] = [index_node_segment.index_node_id]
 
                 if segment_ids:

api/core/rag/docstore/dataset_docstore.py

@@ -50,6 +50,7 @@ class DatasetDocumentStore:
 
         output = {}
         for document_segment in document_segments:
+            assert document_segment.index_node_id
             doc_id = document_segment.index_node_id
             output[doc_id] = Document(
                 page_content=document_segment.content,
@@ -103,7 +104,7 @@ class DatasetDocumentStore:
 
             if not segment_document:
                 max_position += 1
-
+                assert self._document_id
                 segment_document = DocumentSegment(
                     tenant_id=self._dataset.tenant_id,
                     dataset_id=self._dataset.id,

api/core/rag/index_processor/index_processor.py

@@ -84,7 +84,7 @@ class IndexProcessor:
                     select(DocumentSegment).where(DocumentSegment.document_id == original_document_id)
                 ).all()
                 if segments:
-                    index_node_ids = [segment.index_node_id for segment in segments]
+                    index_node_ids = [segment.index_node_id for segment in segments if segment.index_node_id]
 
         indexing_start_at = time.perf_counter()
         # delete from vector index

api/models/dataset.py

@@ -8,7 +8,6 @@ import os
 import pickle
 import re
 import time
-from collections.abc import Sequence
 from datetime import datetime
 from json import JSONDecodeError
 from typing import Any, ClassVar, TypedDict, cast
@@ -831,7 +830,7 @@ class Document(Base):
         )
 
 
-class DocumentSegment(Base):
+class DocumentSegment(TypeBase):
     __tablename__ = "document_segments"
     __table_args__ = (
         sa.PrimaryKeyConstraint("id", name="document_segment_pkey"),
@@ -844,35 +843,40 @@ class DocumentSegment(Base):
     )
 
     # initial fields
-    id = mapped_column(StringUUID, nullable=False, default=lambda: str(uuid4()))
-    tenant_id = mapped_column(StringUUID, nullable=False)
-    dataset_id = mapped_column(StringUUID, nullable=False)
-    document_id = mapped_column(StringUUID, nullable=False)
+    id: Mapped[str] = mapped_column(StringUUID, nullable=False, default_factory=lambda: str(uuid4()), init=False)
+    tenant_id: Mapped[str] = mapped_column(StringUUID, nullable=False)
+    dataset_id: Mapped[str] = mapped_column(StringUUID, nullable=False)
+    document_id: Mapped[str] = mapped_column(StringUUID, nullable=False)
     position: Mapped[int]
-    content = mapped_column(LongText, nullable=False)
-    answer = mapped_column(LongText, nullable=True)
+    content: Mapped[str] = mapped_column(LongText, nullable=False)
     word_count: Mapped[int]
     tokens: Mapped[int]
 
-    # indexing fields
-    keywords = mapped_column(sa.JSON, nullable=True)
-    index_node_id = mapped_column(String(255), nullable=True)
-    index_node_hash = mapped_column(String(255), nullable=True)
-
+    created_by: Mapped[str] = mapped_column(StringUUID, nullable=False)
     # basic fields
+    # indexing fields
+    index_node_id: Mapped[str | None] = mapped_column(String(255), nullable=True, default=None)
+    index_node_hash: Mapped[str | None] = mapped_column(String(255), nullable=True, default=None)
+    enabled: Mapped[bool] = mapped_column(sa.Boolean, nullable=False, server_default=sa.text("true"), default=True)
+    answer: Mapped[str | None] = mapped_column(LongText, nullable=True, default=None)
+    keywords: Mapped[Any] = mapped_column(sa.JSON, nullable=True, default=None)
+    disabled_at: Mapped[datetime | None] = mapped_column(DateTime, nullable=True, default=None)
+    disabled_by: Mapped[str | None] = mapped_column(StringUUID, nullable=True, default=None)
+    status: Mapped[SegmentStatus] = mapped_column(
+        EnumText(SegmentStatus, length=255), server_default=sa.text("'waiting'"), default=SegmentStatus.WAITING
+    )
+    created_at: Mapped[datetime] = mapped_column(
+        DateTime, nullable=False, server_default=func.current_timestamp(), init=False
+    )
+    updated_by: Mapped[str | None] = mapped_column(StringUUID, nullable=True, default=None)
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime, nullable=False, server_default=func.current_timestamp(), init=False
+    )
+    indexing_at: Mapped[datetime | None] = mapped_column(DateTime, nullable=True, default=None)
+    completed_at: Mapped[datetime | None] = mapped_column(DateTime, nullable=True, default=None)
+    error: Mapped[str | None] = mapped_column(LongText, nullable=True, default=None)
+    stopped_at: Mapped[datetime | None] = mapped_column(DateTime, nullable=True, default=None)
     hit_count: Mapped[int] = mapped_column(sa.Integer, nullable=False, default=0)
-    enabled: Mapped[bool] = mapped_column(sa.Boolean, nullable=False, server_default=sa.text("true"))
-    disabled_at: Mapped[datetime | None] = mapped_column(DateTime, nullable=True)
-    disabled_by = mapped_column(StringUUID, nullable=True)
-    status: Mapped[str] = mapped_column(EnumText(SegmentStatus, length=255), server_default=sa.text("'waiting'"))
-    created_by = mapped_column(StringUUID, nullable=False)
-    created_at: Mapped[datetime] = mapped_column(DateTime, nullable=False, server_default=func.current_timestamp())
-    updated_by = mapped_column(StringUUID, nullable=True)
-    updated_at: Mapped[datetime] = mapped_column(DateTime, nullable=False, server_default=func.current_timestamp())
-    indexing_at: Mapped[datetime | None] = mapped_column(DateTime, nullable=True)
-    completed_at: Mapped[datetime | None] = mapped_column(DateTime, nullable=True)
-    error = mapped_column(LongText, nullable=True)
-    stopped_at: Mapped[datetime | None] = mapped_column(DateTime, nullable=True)
 
     @property
     def dataset(self):
@@ -899,7 +903,7 @@ class DocumentSegment(Base):
         )
 
     @property
-    def child_chunks(self) -> Sequence[Any]:
+    def child_chunks(self):
         if not self.document:
             return []
         process_rule = self.document.dataset_process_rule
@@ -914,7 +918,7 @@ class DocumentSegment(Base):
                     return child_chunks or []
         return []
 
-    def get_child_chunks(self) -> Sequence[Any]:
+    def get_child_chunks(self):
         if not self.document:
             return []
         process_rule = self.document.dataset_process_rule

api/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "dify-api"
-version = "1.14.0"
+version = "1.14.1"
 requires-python = "~=3.12.0"
 
 dependencies = [

api/services/vector_service.py

@@ -111,6 +111,7 @@ class VectorService:
                 "dataset_id": segment.dataset_id,
             },
         )
+        assert segment.index_node_id
         if dataset.indexing_technique == IndexTechniqueType.HIGH_QUALITY:
             # update vector index
             vector = Vector(dataset=dataset)
@@ -138,6 +139,7 @@ class VectorService:
         regenerate: bool = False,
     ):
         index_processor = IndexProcessorFactory(dataset.doc_form).init_index_processor()
+        assert segment.index_node_id
         if regenerate:
             # delete child chunks
             index_processor.clean(dataset, [segment.index_node_id], with_keywords=True, delete_child_chunks=True)

api/tasks/batch_clean_document_task.py

@@ -50,7 +50,7 @@ def batch_clean_document_task(document_ids: list[str], dataset_id: str, doc_form
             ).all()
 
             if segments:
-                index_node_ids = [segment.index_node_id for segment in segments]
+                index_node_ids = [segment.index_node_id for segment in segments if segment.index_node_id]
                 segment_ids = [segment.id for segment in segments]
 
                 # Collect image file IDs from segment content

api/tasks/batch_create_segment_to_index_task.py

@@ -19,6 +19,7 @@ from graphon.model_runtime.entities.model_entities import ModelType
 from libs import helper
 from libs.datetime_utils import naive_utc_now
 from models.dataset import Dataset, Document, DocumentSegment
+from models.enums import SegmentStatus
 from models.model import UploadFile
 from services.vector_service import VectorService
 
@@ -156,7 +157,7 @@ def batch_create_segment_to_index_task(
                 tokens=tokens,
                 created_by=user_id,
                 indexing_at=naive_utc_now(),
-                status="completed",
+                status=SegmentStatus.COMPLETED,
                 completed_at=naive_utc_now(),
             )
             if document_config["doc_form"] == IndexStructureType.QA_INDEX:

api/tasks/clean_document_task.py

@@ -53,7 +53,7 @@ def clean_document_task(document_id: str, dataset_id: str, doc_form: str, file_i
             binding_ids = [binding.id for binding, _ in attachments_with_bindings]
             total_attachment_files.extend([attachment_file.key for _, attachment_file in attachments_with_bindings])
 
-            index_node_ids = [segment.index_node_id for segment in segments]
+            index_node_ids = [segment.index_node_id for segment in segments if segment.index_node_id]
             segment_contents = [segment.content for segment in segments]
         except Exception:
             logger.exception("Cleaned document when document deleted failed")

api/tasks/clean_notion_document_task.py

@@ -38,7 +38,7 @@ def clean_notion_document_task(document_ids: list[str], dataset_id: str):
 
         for document_id in document_ids:
             segments = session.scalars(select(DocumentSegment).where(DocumentSegment.document_id == document_id)).all()
-            total_index_node_ids.extend([segment.index_node_id for segment in segments])
+            total_index_node_ids.extend([segment.index_node_id for segment in segments if segment.index_node_id])
 
     # Wrap vector / keyword index cleanup in try/except so that a transient
     # failure here (e.g. billing API hiccup propagated via FeatureService when

api/tasks/disable_segment_from_index_task.py

@@ -9,6 +9,7 @@ from core.db.session_factory import session_factory
 from core.rag.index_processor.index_processor_factory import IndexProcessorFactory
 from extensions.ext_redis import redis_client
 from models.dataset import DocumentSegment
+from models.enums import SegmentStatus
 
 logger = logging.getLogger(__name__)
 
@@ -30,7 +31,7 @@ def disable_segment_from_index_task(segment_id: str):
             logger.info(click.style(f"Segment not found: {segment_id}", fg="red"))
             return
 
-        if segment.status != "completed":
+        if segment.status != SegmentStatus.COMPLETED:
             logger.info(click.style(f"Segment is not completed, disable is not allowed: {segment_id}", fg="red"))
             return
 
@@ -59,6 +60,7 @@ def disable_segment_from_index_task(segment_id: str):
 
             index_type = dataset_document.doc_form
             index_processor = IndexProcessorFactory(index_type).init_index_processor()
+            assert segment.index_node_id
             index_processor.clean(dataset, [segment.index_node_id])
 
             # Disable summary index for this segment

api/tasks/disable_segments_from_index_task.py

@@ -55,7 +55,7 @@ def disable_segments_from_index_task(segment_ids: list, dataset_id: str, documen
             return
 
         try:
-            index_node_ids = [segment.index_node_id for segment in segments]
+            index_node_ids = [segment.index_node_id for segment in segments if segment.index_node_id]
             if dataset.is_multimodal:
                 segment_ids = [segment.id for segment in segments]
                 segment_attachment_bindings = session.scalars(

api/tasks/document_indexing_sync_task.py

@@ -69,7 +69,7 @@ def document_indexing_sync_task(dataset_id: str, document_id: str):
         index_type = document.doc_form
 
         segments = session.scalars(select(DocumentSegment).where(DocumentSegment.document_id == document_id)).all()
-        index_node_ids = [segment.index_node_id for segment in segments]
+        index_node_ids = [segment.index_node_id for segment in segments if segment.index_node_id]
 
     # Get credentials from datasource provider
     datasource_provider_service = DatasourceProviderService()

api/tasks/document_indexing_update_task.py

@@ -45,7 +45,7 @@ def document_indexing_update_task(dataset_id: str, document_id: str):
 
         index_type = document.doc_form
         segments = session.scalars(select(DocumentSegment).where(DocumentSegment.document_id == document_id)).all()
-        index_node_ids = [segment.index_node_id for segment in segments]
+        index_node_ids = [segment.index_node_id for segment in segments if segment.index_node_id]
 
     clean_success = False
     try:

api/tasks/duplicate_document_indexing_task.py

@@ -137,7 +137,7 @@ def _duplicate_document_indexing_task(dataset_id: str, document_ids: Sequence[st
                     select(DocumentSegment).where(DocumentSegment.document_id == document.id)
                 ).all()
                 if segments:
-                    index_node_ids = [segment.index_node_id for segment in segments]
+                    index_node_ids = [segment.index_node_id for segment in segments if segment.index_node_id]
 
                     # delete from vector index
                     index_processor.clean(dataset, index_node_ids, with_keywords=True, delete_child_chunks=True)

api/tasks/remove_document_from_index_task.py

@@ -61,7 +61,7 @@ def remove_document_from_index_task(document_id: str):
                 except Exception as e:
                     logger.warning("Failed to disable summaries for document %s: %s", document.id, str(e))
 
-            index_node_ids = [segment.index_node_id for segment in segments]
+            index_node_ids = [segment.index_node_id for segment in segments if segment.index_node_id]
             if index_node_ids:
                 try:
                     index_processor.clean(dataset, index_node_ids, with_keywords=True, delete_child_chunks=False)

api/tasks/retry_document_indexing_task.py

@@ -85,7 +85,7 @@ def retry_document_indexing_task(dataset_id: str, document_ids: list[str], user_
                         select(DocumentSegment).where(DocumentSegment.document_id == document_id)
                     ).all()
                     if segments:
-                        index_node_ids = [segment.index_node_id for segment in segments]
+                        index_node_ids = [segment.index_node_id for segment in segments if segment.index_node_id]
                         # delete from vector index
                         index_processor.clean(dataset, index_node_ids, with_keywords=True, delete_child_chunks=True)
 

api/tasks/sync_website_document_indexing_task.py

@@ -70,7 +70,7 @@ def sync_website_document_indexing_task(dataset_id: str, document_id: str):
 
             segments = session.scalars(select(DocumentSegment).where(DocumentSegment.document_id == document_id)).all()
             if segments:
-                index_node_ids = [segment.index_node_id for segment in segments]
+                index_node_ids = [segment.index_node_id for segment in segments if segment.index_node_id]
                 # delete from vector index
                 index_processor.clean(dataset, index_node_ids, with_keywords=True, delete_child_chunks=True)
 

api/tests/test_containers_integration_tests/services/test_dataset_service_get_segments.py

@@ -13,9 +13,9 @@ from uuid import uuid4
 from sqlalchemy.orm import Session
 
 from core.rag.index_processor.constant.index_type import IndexTechniqueType
-from models import Account, Tenant, TenantAccountJoin, TenantAccountRole
+from models import Account, AccountStatus, Tenant, TenantAccountJoin, TenantAccountRole, TenantStatus
 from models.dataset import Dataset, DatasetPermissionEnum, Document, DocumentSegment
-from models.enums import DataSourceType, DocumentCreatedFrom
+from models.enums import DataSourceType, DocumentCreatedFrom, SegmentStatus
 from services.dataset_service import SegmentService
 
 
@@ -35,13 +35,13 @@ class SegmentServiceTestDataFactory:
             email=f"{uuid4()}@example.com",
             name=f"user-{uuid4()}",
             interface_language="en-US",
-            status="active",
+            status=AccountStatus.ACTIVE,
         )
         db_session_with_containers.add(account)
         db_session_with_containers.commit()
 
         if tenant is None:
-            tenant = Tenant(name=f"tenant-{uuid4()}", status="normal")
+            tenant = Tenant(name=f"tenant-{uuid4()}", status=TenantStatus.NORMAL)
             db_session_with_containers.add(tenant)
             db_session_with_containers.commit()
 
@@ -103,7 +103,7 @@ class SegmentServiceTestDataFactory:
         created_by: str,
         position: int = 1,
         content: str = "Test content",
-        status: str = "completed",
+        status: SegmentStatus = SegmentStatus.COMPLETED,
         word_count: int = 10,
         tokens: int = 15,
     ) -> DocumentSegment:
@@ -203,7 +203,7 @@ class TestSegmentServiceGetSegments:
             document_id=document.id,
             created_by=owner.id,
             position=1,
-            status="completed",
+            status=SegmentStatus.COMPLETED,
         )
         SegmentServiceTestDataFactory.create_segment(
             db_session_with_containers,
@@ -212,7 +212,7 @@ class TestSegmentServiceGetSegments:
             document_id=document.id,
             created_by=owner.id,
             position=2,
-            status="indexing",
+            status=SegmentStatus.INDEXING,
         )
         SegmentServiceTestDataFactory.create_segment(
             db_session_with_containers,
@@ -221,7 +221,7 @@ class TestSegmentServiceGetSegments:
             document_id=document.id,
             created_by=owner.id,
             position=3,
-            status="waiting",
+            status=SegmentStatus.WAITING,
         )
 
         # Act
@@ -257,7 +257,7 @@ class TestSegmentServiceGetSegments:
             document_id=document.id,
             created_by=owner.id,
             position=1,
-            status="completed",
+            status=SegmentStatus.COMPLETED,
         )
         SegmentServiceTestDataFactory.create_segment(
             db_session_with_containers,
@@ -266,7 +266,7 @@ class TestSegmentServiceGetSegments:
             document_id=document.id,
             created_by=owner.id,
             position=2,
-            status="indexing",
+            status=SegmentStatus.INDEXING,
         )
 
         # Act
@@ -415,7 +415,7 @@ class TestSegmentServiceGetSegments:
             document_id=document.id,
             created_by=owner.id,
             position=1,
-            status="completed",
+            status=SegmentStatus.COMPLETED,
             content="This is important information",
         )
         SegmentServiceTestDataFactory.create_segment(
@@ -425,7 +425,7 @@ class TestSegmentServiceGetSegments:
             document_id=document.id,
             created_by=owner.id,
             position=2,
-            status="indexing",
+            status=SegmentStatus.INDEXING,
             content="This is also important",
         )
         SegmentServiceTestDataFactory.create_segment(
@@ -435,7 +435,7 @@ class TestSegmentServiceGetSegments:
             document_id=document.id,
             created_by=owner.id,
             position=3,
-            status="completed",
+            status=SegmentStatus.COMPLETED,
             content="This is irrelevant",
         )
 
@@ -477,7 +477,7 @@ class TestSegmentServiceGetSegments:
             document_id=document.id,
             created_by=owner.id,
             position=1,
-            status="completed",
+            status=SegmentStatus.COMPLETED,
         )
         SegmentServiceTestDataFactory.create_segment(
             db_session_with_containers,
@@ -486,7 +486,7 @@ class TestSegmentServiceGetSegments:
             document_id=document.id,
             created_by=owner.id,
             position=2,
-            status="waiting",
+            status=SegmentStatus.WAITING,
         )
 
         # Act

api/tests/test_containers_integration_tests/tasks/test_add_document_to_index_task.py

@@ -128,7 +128,6 @@ class TestAddDocumentToIndexTask:
 
         for i in range(3):
             segment = DocumentSegment(
-                id=fake.uuid4(),
                 tenant_id=document.tenant_id,
                 dataset_id=dataset.id,
                 document_id=document.id,
@@ -451,7 +450,6 @@ class TestAddDocumentToIndexTask:
         segments = []
         for i in range(3):
             segment = DocumentSegment(
-                id=fake.uuid4(),
                 tenant_id=document.tenant_id,
                 dataset_id=dataset.id,
                 document_id=document.id,
@@ -630,7 +628,6 @@ class TestAddDocumentToIndexTask:
 
         # Segment 1: Should be processed (enabled=False, status=SegmentStatus.COMPLETED)
         segment1 = DocumentSegment(
-            id=fake.uuid4(),
             tenant_id=document.tenant_id,
             dataset_id=dataset.id,
             document_id=document.id,
@@ -650,7 +647,6 @@ class TestAddDocumentToIndexTask:
         # Segment 2: Should be processed (enabled=True, status=SegmentStatus.COMPLETED)
         # Note: Implementation doesn't filter by enabled status, only by status=SegmentStatus.COMPLETED
         segment2 = DocumentSegment(
-            id=fake.uuid4(),
             tenant_id=document.tenant_id,
             dataset_id=dataset.id,
             document_id=document.id,
@@ -669,7 +665,6 @@ class TestAddDocumentToIndexTask:
 
         # Segment 3: Should NOT be processed (enabled=False, status="processing")
         segment3 = DocumentSegment(
-            id=fake.uuid4(),
             tenant_id=document.tenant_id,
             dataset_id=dataset.id,
             document_id=document.id,
@@ -688,7 +683,6 @@ class TestAddDocumentToIndexTask:
 
         # Segment 4: Should be processed (enabled=False, status=SegmentStatus.COMPLETED)
         segment4 = DocumentSegment(
-            id=fake.uuid4(),
             tenant_id=document.tenant_id,
             dataset_id=dataset.id,
             document_id=document.id,

api/tests/test_containers_integration_tests/tasks/test_batch_clean_document_task.py

@@ -177,7 +177,6 @@ class TestBatchCleanDocumentTask:
         fake = Faker()
 
         segment = DocumentSegment(
-            id=str(uuid.uuid4()),
             tenant_id=account.current_tenant.id,
             dataset_id=document.dataset_id,
             document_id=document.id,
@@ -290,10 +289,9 @@ class TestBatchCleanDocumentTask:
         account = self._create_test_account(db_session_with_containers)
         dataset = self._create_test_dataset(db_session_with_containers, account)
         document = self._create_test_document(db_session_with_containers, dataset, account)
-
+        assert account.current_tenant
         # Create segment with simple content (no image references)
         segment = DocumentSegment(
-            id=str(uuid.uuid4()),
             tenant_id=account.current_tenant.id,
             dataset_id=document.dataset_id,
             document_id=document.id,
@@ -692,9 +690,9 @@ class TestBatchCleanDocumentTask:
 
         # Create multiple segments for the document
         segments = []
+        assert account.current_tenant
         for i in range(3):
             segment = DocumentSegment(
-                id=str(uuid.uuid4()),
                 tenant_id=account.current_tenant.id,
                 dataset_id=document.dataset_id,
                 document_id=document.id,

api/tests/test_containers_integration_tests/tasks/test_clean_dataset_task.py

@@ -220,7 +220,6 @@ class TestCleanDatasetTask:
             DocumentSegment: Created document segment instance
         """
         segment = DocumentSegment(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             dataset_id=dataset.id,
             document_id=document.id,
@@ -232,8 +231,6 @@ class TestCleanDatasetTask:
             status=SegmentStatus.COMPLETED,
             index_node_id=str(uuid.uuid4()),
             index_node_hash="test_hash",
-            created_at=datetime.now(),
-            updated_at=datetime.now(),
         )
 
         db_session_with_containers.add(segment)
@@ -614,7 +611,6 @@ class TestCleanDatasetTask:
         """
 
         segment = DocumentSegment(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             dataset_id=dataset.id,
             document_id=document.id,
@@ -626,8 +622,6 @@ class TestCleanDatasetTask:
             status=SegmentStatus.COMPLETED,
             index_node_id=str(uuid.uuid4()),
             index_node_hash="test_hash",
-            created_at=datetime.now(),
-            updated_at=datetime.now(),
         )
 
         db_session_with_containers.add(segment)
@@ -729,8 +723,6 @@ class TestCleanDatasetTask:
                 type=DatasetMetadataType.STRING,
                 created_by=account.id,
             )
-            metadata.id = str(uuid.uuid4())
-            metadata.created_at = datetime.now()
             metadata_items.append(metadata)
 
             # Create binding for each metadata item
@@ -741,8 +733,6 @@ class TestCleanDatasetTask:
                 document_id=documents[i % len(documents)].id,
                 created_by=account.id,
             )
-            binding.id = str(uuid.uuid4())
-            binding.created_at = datetime.now()
             bindings.append(binding)
 
         db_session_with_containers.add_all(metadata_items)
@@ -946,7 +936,6 @@ class TestCleanDatasetTask:
         long_content = "Very long content " * 100  # Long content within reasonable limits
         segment_content = f"Segment with special chars: {special_content}\n{long_content}"
         segment = DocumentSegment(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             dataset_id=dataset.id,
             document_id=document.id,
@@ -958,8 +947,6 @@ class TestCleanDatasetTask:
             status=SegmentStatus.COMPLETED,
             index_node_id=str(uuid.uuid4()),
             index_node_hash="test_hash_" + "x" * 50,  # Long hash within limits
-            created_at=datetime.now(),
-            updated_at=datetime.now(),
         )
         db_session_with_containers.add(segment)
         db_session_with_containers.commit()

api/tests/test_containers_integration_tests/tasks/test_clean_notion_document_task.py

@@ -123,11 +123,10 @@ class TestCleanNotionDocumentTask:
             db_session_with_containers.add(document)
             db_session_with_containers.flush()
             document_ids.append(document.id)
-
+            assert tenant
             # Create segments for each document
             for j in range(2):
                 segment = DocumentSegment(
-                    id=str(uuid.uuid4()),
                     tenant_id=tenant.id,
                     dataset_id=dataset.id,
                     document_id=document.id,
@@ -303,10 +302,9 @@ class TestCleanNotionDocumentTask:
             )
             db_session_with_containers.add(document)
             db_session_with_containers.flush()
-
+            assert tenant
             # Create test segment
             segment = DocumentSegment(
-                id=str(uuid.uuid4()),
                 tenant_id=tenant.id,
                 dataset_id=dataset.id,
                 document_id=document.id,
@@ -390,12 +388,11 @@ class TestCleanNotionDocumentTask:
         )
         db_session_with_containers.add(document)
         db_session_with_containers.flush()
-
+        assert tenant
         # Create segments without index_node_ids
         segments = []
         for i in range(3):
             segment = DocumentSegment(
-                id=str(uuid.uuid4()),
                 tenant_id=tenant.id,
                 dataset_id=dataset.id,
                 document_id=document.id,
@@ -484,11 +481,10 @@ class TestCleanNotionDocumentTask:
             db_session_with_containers.add(document)
             db_session_with_containers.flush()
             documents.append(document)
-
+            assert tenant
             # Create segments for each document
             for j in range(2):
                 segment = DocumentSegment(
-                    id=str(uuid.uuid4()),
                     tenant_id=tenant.id,
                     dataset_id=dataset.id,
                     document_id=document.id,
@@ -612,10 +608,9 @@ class TestCleanNotionDocumentTask:
         segment_statuses = [SegmentStatus.WAITING, SegmentStatus.INDEXING, SegmentStatus.COMPLETED, SegmentStatus.ERROR]
         segments = []
         index_node_ids = []
-
+        assert tenant
         for i, status in enumerate(segment_statuses):
             segment = DocumentSegment(
-                id=str(uuid.uuid4()),
                 tenant_id=tenant.id,
                 dataset_id=dataset.id,
                 document_id=document.id,
@@ -718,10 +713,9 @@ class TestCleanNotionDocumentTask:
         )
         db_session_with_containers.add(document)
         db_session_with_containers.flush()
-
+        assert tenant
         # Create segment
         segment = DocumentSegment(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             dataset_id=dataset.id,
             document_id=document.id,
@@ -818,12 +812,11 @@ class TestCleanNotionDocumentTask:
             db_session_with_containers.add(document)
             db_session_with_containers.flush()
             documents.append(document)
-
+            assert tenant
             # Create multiple segments for each document
             num_segments_per_doc = 5
             for j in range(num_segments_per_doc):
                 segment = DocumentSegment(
-                    id=str(uuid.uuid4()),
                     tenant_id=tenant.id,
                     dataset_id=dataset.id,
                     document_id=document.id,
@@ -940,7 +933,6 @@ class TestCleanNotionDocumentTask:
             # Create segments for each document
             for j in range(3):
                 segment = DocumentSegment(
-                    id=str(uuid.uuid4()),
                     tenant_id=account.current_tenant.id,
                     dataset_id=dataset.id,
                     document_id=document.id,
@@ -1066,11 +1058,10 @@ class TestCleanNotionDocumentTask:
             db_session_with_containers.add(document)
             db_session_with_containers.flush()
             documents.append(document)
-
+            assert tenant
             # Create segments for each document
             for j in range(2):
                 segment = DocumentSegment(
-                    id=str(uuid.uuid4()),
                     tenant_id=tenant.id,
                     dataset_id=dataset.id,
                     document_id=document.id,
@@ -1184,10 +1175,9 @@ class TestCleanNotionDocumentTask:
         # Create segments with metadata
         segments = []
         index_node_ids = []
-
+        assert tenant
         for i in range(3):
             segment = DocumentSegment(
-                id=str(uuid.uuid4()),
                 tenant_id=tenant.id,
                 dataset_id=dataset.id,
                 document_id=document.id,

api/tests/test_containers_integration_tests/tasks/test_deal_dataset_vector_index_task.py

@@ -90,7 +90,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create dataset
         dataset = Dataset(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             name=fake.company(),
             description=fake.text(max_nb_chars=100),
@@ -150,7 +149,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create dataset
         dataset = Dataset(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             name=fake.company(),
             description=fake.text(max_nb_chars=100),
@@ -202,7 +200,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create segments
         segment = DocumentSegment(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             dataset_id=dataset.id,
             document_id=document.id,
@@ -253,7 +250,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create dataset with parent-child index
         dataset = Dataset(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             name=fake.company(),
             description=fake.text(max_nb_chars=100),
@@ -305,7 +301,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create segments
         segment = DocumentSegment(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             dataset_id=dataset.id,
             document_id=document.id,
@@ -371,7 +366,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create dataset without documents
         dataset = Dataset(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             name=fake.company(),
             description=fake.text(max_nb_chars=100),
@@ -403,7 +397,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create dataset
         dataset = Dataset(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             name=fake.company(),
             description=fake.text(max_nb_chars=100),
@@ -461,7 +454,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create dataset without documents
         dataset = Dataset(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             name=fake.company(),
             description=fake.text(max_nb_chars=100),
@@ -494,7 +486,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create dataset
         dataset = Dataset(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             name=fake.company(),
             description=fake.text(max_nb_chars=100),
@@ -546,7 +537,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create segments
         segment = DocumentSegment(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             dataset_id=dataset.id,
             document_id=document.id,
@@ -592,7 +582,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create dataset with custom index type
         dataset = Dataset(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             name=fake.company(),
             description=fake.text(max_nb_chars=100),
@@ -624,7 +613,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create segments
         segment = DocumentSegment(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             dataset_id=dataset.id,
             document_id=document.id,
@@ -670,7 +658,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create dataset without doc_form (should use default)
         dataset = Dataset(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             name=fake.company(),
             description=fake.text(max_nb_chars=100),
@@ -702,7 +689,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create segments
         segment = DocumentSegment(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             dataset_id=dataset.id,
             document_id=document.id,
@@ -748,7 +734,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create dataset
         dataset = Dataset(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             name=fake.company(),
             description=fake.text(max_nb_chars=100),
@@ -806,7 +791,6 @@ class TestDealDatasetVectorIndexTask:
         for i, document in enumerate(documents):
             for j in range(2):
                 segment = DocumentSegment(
-                    id=str(uuid.uuid4()),
                     tenant_id=tenant.id,
                     dataset_id=dataset.id,
                     document_id=document.id,
@@ -832,6 +816,7 @@ class TestDealDatasetVectorIndexTask:
             updated_document = db_session_with_containers.scalar(
                 select(Document).where(Document.id == document.id).limit(1)
             )
+            assert updated_document
             assert updated_document.indexing_status == IndexingStatus.COMPLETED
 
         # Verify index processor load was called multiple times
@@ -853,7 +838,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create dataset
         dataset = Dataset(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             name=fake.company(),
             description=fake.text(max_nb_chars=100),
@@ -905,7 +889,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create segments
         segment = DocumentSegment(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             dataset_id=dataset.id,
             document_id=document.id,
@@ -952,7 +935,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create dataset
         dataset = Dataset(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             name=fake.company(),
             description=fake.text(max_nb_chars=100),
@@ -1024,7 +1006,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create segments for enabled document only
         segment = DocumentSegment(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             dataset_id=dataset.id,
             document_id=enabled_document.id,
@@ -1075,7 +1056,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create dataset
         dataset = Dataset(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             name=fake.company(),
             description=fake.text(max_nb_chars=100),
@@ -1147,7 +1127,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create segments for active document only
         segment = DocumentSegment(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             dataset_id=dataset.id,
             document_id=active_document.id,
@@ -1198,7 +1177,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create dataset
         dataset = Dataset(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             name=fake.company(),
             description=fake.text(max_nb_chars=100),
@@ -1270,7 +1248,6 @@ class TestDealDatasetVectorIndexTask:
 
         # Create segments for completed document only
         segment = DocumentSegment(
-            id=str(uuid.uuid4()),
             tenant_id=tenant.id,
             dataset_id=dataset.id,
             document_id=completed_document.id,

api/tests/test_containers_integration_tests/tasks/test_delete_segment_from_index_task.py

@@ -209,26 +209,25 @@ class TestDeleteSegmentFromIndexTask:
         segments = []
 
         for i in range(count):
-            segment = DocumentSegment()
-            segment.id = fake.uuid4()
-            segment.tenant_id = document.tenant_id
-            segment.dataset_id = document.dataset_id
-            segment.document_id = document.id
-            segment.position = i + 1
-            segment.content = f"Test segment content {i + 1}: {fake.text(max_nb_chars=200)}"
-            segment.answer = f"Test segment answer {i + 1}: {fake.text(max_nb_chars=100)}"
-            segment.word_count = fake.random_int(min=10, max=100)
-            segment.tokens = fake.random_int(min=5, max=50)
-            segment.keywords = [fake.word() for _ in range(3)]
-            segment.index_node_id = f"node_{fake.uuid4()}"
-            segment.index_node_hash = fake.sha256()
-            segment.hit_count = 0
-            segment.enabled = True
-            segment.status = SegmentStatus.COMPLETED
-            segment.created_by = account.id
-            segment.created_at = fake.date_time_this_year()
-            segment.updated_by = account.id
-            segment.updated_at = segment.created_at
+            created_at = fake.date_time_this_year()
+            segment = DocumentSegment(
+                tenant_id=document.tenant_id,
+                dataset_id=document.dataset_id,
+                document_id=document.id,
+                position=i + 1,
+                content=f"Test segment content {i + 1}: {fake.text(max_nb_chars=200)}",
+                answer=f"Test segment answer {i + 1}: {fake.text(max_nb_chars=100)}",
+                word_count=fake.random_int(min=10, max=100),
+                tokens=fake.random_int(min=5, max=50),
+                keywords=[fake.word() for _ in range(3)],
+                index_node_id=f"node_{fake.uuid4()}",
+                index_node_hash=fake.sha256(),
+                hit_count=0,
+                enabled=True,
+                status=SegmentStatus.COMPLETED,
+                created_by=account.id,
+                updated_by=account.id,
+            )
 
             db_session_with_containers.add(segment)
             segments.append(segment)

api/tests/test_containers_integration_tests/tasks/test_disable_segment_from_index_task.py

@@ -159,7 +159,7 @@ class TestDisableSegmentFromIndexTask:
         dataset: Dataset,
         tenant: Tenant,
         account: Account,
-        status: str = "completed",
+        status: SegmentStatus = SegmentStatus.COMPLETED,
         enabled: bool = True,
     ) -> DocumentSegment:
         """

api/tests/test_containers_integration_tests/tasks/test_disable_segments_from_index_task.py

@@ -185,30 +185,31 @@ class TestDisableSegmentsFromIndexTask:
         segments = []
 
         for i in range(count):
-            segment = DocumentSegment()
-            segment.id = fake.uuid4()
-            segment.tenant_id = dataset.tenant_id
-            segment.dataset_id = dataset.id
-            segment.document_id = document.id
-            segment.position = i + 1
-            segment.content = f"Test segment content {i + 1}: {fake.text(max_nb_chars=200)}"
-            segment.answer = f"Test answer {i + 1}" if i % 2 == 0 else None
-            segment.word_count = fake.random_int(min=10, max=100)
-            segment.tokens = fake.random_int(min=5, max=50)
-            segment.keywords = [fake.word() for _ in range(3)]
-            segment.index_node_id = f"node_{segment.id}"
-            segment.index_node_hash = fake.sha256()
-            segment.hit_count = 0
-            segment.enabled = True
-            segment.disabled_at = None
-            segment.disabled_by = None
-            segment.status = SegmentStatus.COMPLETED
-            segment.created_by = account.id
-            segment.updated_by = account.id
-            segment.indexing_at = fake.date_time_this_year()
-            segment.completed_at = fake.date_time_this_year()
-            segment.error = None
-            segment.stopped_at = None
+            id = fake.uuid4()
+            segment = DocumentSegment(
+                tenant_id=dataset.tenant_id,
+                dataset_id=dataset.id,
+                document_id=document.id,
+                position=i + 1,
+                content=f"Test segment content {i + 1}: {fake.text(max_nb_chars=200)}",
+                answer=f"Test answer {i + 1}" if i % 2 == 0 else None,
+                word_count=fake.random_int(min=10, max=100),
+                tokens=fake.random_int(min=5, max=50),
+                keywords=[fake.word() for _ in range(3)],
+                index_node_id=f"node_{id}",
+                index_node_hash=fake.sha256(),
+                hit_count=0,
+                enabled=True,
+                disabled_at=None,
+                disabled_by=None,
+                status=SegmentStatus.COMPLETED,
+                created_by=account.id,
+                updated_by=account.id,
+                indexing_at=fake.date_time_this_year(),
+                completed_at=fake.date_time_this_year(),
+                error=None,
+                stopped_at=None,
+            )
 
             segments.append(segment)
 

api/tests/test_containers_integration_tests/tasks/test_duplicate_document_indexing_task.py

@@ -175,7 +175,6 @@ class TestDuplicateDocumentIndexingTasks:
         for document in documents:
             for i in range(segments_per_doc):
                 segment = DocumentSegment(
-                    id=fake.uuid4(),
                     tenant_id=dataset.tenant_id,
                     dataset_id=dataset.id,
                     document_id=document.id,

api/tests/test_containers_integration_tests/tasks/test_enable_segments_to_index_task.py

@@ -139,7 +139,6 @@ class TestEnableSegmentsToIndexTask:
         for i in range(count):
             text = fake.text(max_nb_chars=200)
             segment = DocumentSegment(
-                id=fake.uuid4(),
                 tenant_id=document.tenant_id,
                 dataset_id=dataset.id,
                 document_id=document.id,

api/tests/unit_tests/services/test_dataset_service_segment.py

@@ -282,7 +282,6 @@ class TestSegmentServiceQueries:
 
     def test_get_segment_by_id_returns_only_document_segment_instances(self):
         segment = DocumentSegment(
-            id="segment-1",
             tenant_id="tenant-1",
             dataset_id="dataset-1",
             document_id="doc-1",
@@ -292,7 +291,7 @@ class TestSegmentServiceQueries:
             tokens=2,
             created_by="user-1",
         )
-
+        segment.id = "segment-1"
         with patch("services.dataset_service.db") as mock_db:
             mock_db.session.scalar.return_value = segment
             result = SegmentService.get_segment_by_id("segment-1", "tenant-1")
@@ -307,7 +306,6 @@ class TestSegmentServiceQueries:
 
     def test_get_segments_by_document_and_dataset_returns_scalars_result(self):
         segment = DocumentSegment(
-            id="segment-1",
             tenant_id="tenant-1",
             dataset_id="dataset-1",
             document_id="doc-1",
@@ -318,6 +316,7 @@ class TestSegmentServiceQueries:
             created_by="user-1",
         )
 
+        segment.id = "segment-1"
         with patch("services.dataset_service.db") as mock_db:
             mock_db.session.scalars.return_value.all.return_value = [segment]
 
@@ -461,6 +460,7 @@ class TestSegmentServiceMutations:
             vector_service.create_segments_vector.side_effect = RuntimeError("vector failed")
 
             result = SegmentService.multi_create_segment(segments, document, dataset)
+            assert result
 
         assert len(result) == 2
         assert [segment.position for segment in result] == [2, 3]

api/uv.lock

@@ -1379,7 +1379,7 @@ wheels = [
 
 [[package]]
 name = "dify-api"
-version = "1.14.0"
+version = "1.14.1"
 source = { virtual = "." }
 dependencies = [
     { name = "aliyun-log-python-sdk" },

docker/docker-compose-template.yaml

@@ -220,7 +220,7 @@ services:
   # API service
   api:
     <<: *shared-api-worker-config
-    image: langgenius/dify-api:1.14.0
+    image: langgenius/dify-api:1.14.1
     environment:
       MODE: api
       SENTRY_DSN: ${API_SENTRY_DSN:-}
@@ -264,7 +264,7 @@ services:
   # WebSocket service for workflow collaboration.
   api_websocket:
     <<: *shared-api-worker-config
-    image: langgenius/dify-api:1.14.0
+    image: langgenius/dify-api:1.14.1
     profiles:
       - collaboration
     environment:
@@ -290,7 +290,7 @@ services:
   # The Celery worker for processing all queues (dataset, workflow, mail, etc.)
   worker:
     <<: *shared-worker-config
-    image: langgenius/dify-api:1.14.0
+    image: langgenius/dify-api:1.14.1
     environment:
       MODE: worker
       SENTRY_DSN: ${API_SENTRY_DSN:-}
@@ -333,7 +333,7 @@ services:
   # Celery beat for scheduling periodic tasks.
   worker_beat:
     <<: *shared-worker-beat-config
-    image: langgenius/dify-api:1.14.0
+    image: langgenius/dify-api:1.14.1
     environment:
       MODE: beat
     depends_on:
@@ -366,7 +366,7 @@ services:
 
   # Frontend web application.
   web:
-    image: langgenius/dify-web:1.14.0
+    image: langgenius/dify-web:1.14.1
     restart: always
     env_file:
       - path: ./envs/core-services/web.env

docker/docker-compose.yaml

@@ -226,7 +226,7 @@ services:
   # API service
   api:
     <<: *shared-api-worker-config
-    image: langgenius/dify-api:1.14.0
+    image: langgenius/dify-api:1.14.1
     environment:
       MODE: api
       SENTRY_DSN: ${API_SENTRY_DSN:-}
@@ -270,7 +270,7 @@ services:
   # WebSocket service for workflow collaboration.
   api_websocket:
     <<: *shared-api-worker-config
-    image: langgenius/dify-api:1.14.0
+    image: langgenius/dify-api:1.14.1
     profiles:
       - collaboration
     environment:
@@ -296,7 +296,7 @@ services:
   # The Celery worker for processing all queues (dataset, workflow, mail, etc.)
   worker:
     <<: *shared-worker-config
-    image: langgenius/dify-api:1.14.0
+    image: langgenius/dify-api:1.14.1
     environment:
       MODE: worker
       SENTRY_DSN: ${API_SENTRY_DSN:-}
@@ -339,7 +339,7 @@ services:
   # Celery beat for scheduling periodic tasks.
   worker_beat:
     <<: *shared-worker-beat-config
-    image: langgenius/dify-api:1.14.0
+    image: langgenius/dify-api:1.14.1
     environment:
       MODE: beat
     depends_on:
@@ -372,7 +372,7 @@ services:
 
   # Frontend web application.
   web:
-    image: langgenius/dify-web:1.14.0
+    image: langgenius/dify-web:1.14.1
     restart: always
     env_file:
       - path: ./envs/core-services/web.env

web/__tests__/apps/create-app-flow.test.tsx

@@ -86,6 +86,9 @@ vi.mock('@tanstack/react-query', async (importOriginal) => {
   const actual = await importOriginal<typeof import('@tanstack/react-query')>()
   return {
     ...actual,
+    useQuery: () => ({
+      data: [],
+    }),
     useInfiniteQuery: () => ({
       data: { pages: mockPages },
       isLoading: mockIsLoading,

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

@@ -1,20 +1,19 @@
-import type { ReactNode } from 'react'
-import type { DataSet } from '@/models/datasets'
-import { render, screen } from '@testing-library/react'
-import { IndexingType } from '@/app/components/datasets/create/step-two'
-import { ChunkingMode, DatasetPermission, DataSourceType } from '@/models/datasets'
-import { RETRIEVE_METHOD } from '@/types/app'
+import { render, screen, waitFor } from '@testing-library/react'
+import { usePathname, useRouter } from '@/next/navigation'
+import { useDatasetDetail, useDatasetRelatedApps } from '@/service/knowledge/use-dataset'
 import DatasetDetailLayout from '../layout-main'
 
-let mockPathname = '/datasets/test-dataset-id/documents'
-let mockDataset: DataSet | undefined
-let mockCanAccessSnippetsAndEvaluation = true
-
+const mockReplace = vi.fn()
 const mockSetAppSidebarExpand = vi.fn()
-const mockMutateDatasetRes = vi.fn()
 
 vi.mock('@/next/navigation', () => ({
-  usePathname: () => mockPathname,
+  usePathname: vi.fn(),
+  useRouter: vi.fn(),
+}))
+
+vi.mock('@/service/knowledge/use-dataset', () => ({
+  useDatasetDetail: vi.fn(),
+  useDatasetRelatedApps: vi.fn(),
 }))
 
 vi.mock('@/app/components/app/store', () => ({
@@ -23,208 +22,130 @@ vi.mock('@/app/components/app/store', () => ({
   }),
 }))
 
-vi.mock('@/hooks/use-breakpoints', () => ({
-  default: () => 'desktop',
-  MediaType: {
-    mobile: 'mobile',
-    desktop: 'desktop',
-  },
-}))
-
-vi.mock('@/context/event-emitter', () => ({
-  useEventEmitterContextContext: () => ({
-    eventEmitter: {
-      useSubscription: vi.fn(),
-    },
-  }),
-}))
-
 vi.mock('@/context/app-context', () => ({
   useAppContext: () => ({
     isCurrentWorkspaceDatasetOperator: false,
   }),
 }))
 
-vi.mock('@/hooks/use-snippet-and-evaluation-plan-access', () => ({
-  useSnippetAndEvaluationPlanAccess: () => ({
-    canAccess: mockCanAccessSnippetsAndEvaluation,
-    isReady: true,
+vi.mock('@/context/event-emitter', () => ({
+  useEventEmitterContextContext: () => ({
+    eventEmitter: undefined,
   }),
 }))
 
+vi.mock('@/hooks/use-breakpoints', () => ({
+  default: () => 'desktop',
+  MediaType: {
+    mobile: 'mobile',
+  },
+}))
+
 vi.mock('@/hooks/use-document-title', () => ({
   default: vi.fn(),
 }))
 
-vi.mock('@/service/knowledge/use-dataset', () => ({
-  useDatasetDetail: () => ({
-    data: mockDataset,
-    error: null,
-    refetch: mockMutateDatasetRes,
-  }),
-  useDatasetRelatedApps: () => ({
-    data: [],
-  }),
-}))
-
 vi.mock('@/app/components/app-sidebar', () => ({
-  default: ({
-    navigation,
-    children,
-  }: {
-    navigation: Array<{ name: string, href: string, disabled?: boolean }>
-    children?: ReactNode
-  }) => (
-    <div data-testid="app-sidebar">
-      {navigation.map(item => (
-        <button
-          key={item.href}
-          type="button"
-          disabled={item.disabled}
-        >
-          {item.name}
-        </button>
-      ))}
-      {children}
-    </div>
-  ),
+  default: () => <aside aria-label="dataset navigation" />,
 }))
 
 vi.mock('@/app/components/datasets/extra-info', () => ({
-  default: () => <div data-testid="dataset-extra-info" />,
+  default: () => <div />,
 }))
 
-vi.mock('@/app/components/base/loading', () => ({
-  default: () => <div role="status">loading</div>,
-}))
-
-const createDataset = (overrides: Partial<DataSet> = {}): DataSet => ({
-  id: 'test-dataset-id',
-  name: 'Test Dataset',
-  indexing_status: 'completed',
-  icon_info: {
-    icon: 'book',
-    icon_background: '#fff',
-    icon_type: 'emoji',
-    icon_url: '',
-  },
-  description: '',
-  permission: DatasetPermission.onlyMe,
-  data_source_type: DataSourceType.FILE,
-  indexing_technique: IndexingType.QUALIFIED,
-  created_by: 'user-1',
-  updated_by: 'user-1',
-  updated_at: 0,
-  app_count: 0,
-  doc_form: ChunkingMode.text,
-  document_count: 0,
-  total_document_count: 0,
-  word_count: 0,
-  provider: 'vendor',
-  embedding_model: 'text-embedding',
-  embedding_model_provider: 'openai',
-  embedding_available: true,
-  retrieval_model_dict: {
-    search_method: RETRIEVE_METHOD.semantic,
-    reranking_enable: false,
-    reranking_model: {
-      reranking_provider_name: '',
-      reranking_model_name: '',
-    },
-    top_k: 3,
-    score_threshold_enabled: false,
-    score_threshold: 0.5,
-  },
-  retrieval_model: {
-    search_method: RETRIEVE_METHOD.semantic,
-    reranking_enable: false,
-    reranking_model: {
-      reranking_provider_name: '',
-      reranking_model_name: '',
-    },
-    top_k: 3,
-    score_threshold_enabled: false,
-    score_threshold: 0.5,
-  },
-  tags: [],
-  external_knowledge_info: {
-    external_knowledge_id: '',
-    external_knowledge_api_id: '',
-    external_knowledge_api_name: '',
-    external_knowledge_api_endpoint: '',
-  },
-  external_retrieval_model: {
-    top_k: 3,
-    score_threshold: 0.5,
-    score_threshold_enabled: false,
-  },
-  built_in_field_enabled: false,
-  pipeline_id: 'pipeline-1',
-  is_published: true,
-  runtime_mode: 'rag_pipeline',
-  enable_api: false,
-  is_multimodal: false,
-  ...overrides,
-})
+const mockUsePathname = vi.mocked(usePathname)
+const mockUseRouter = vi.mocked(useRouter)
+const mockUseDatasetDetail = vi.mocked(useDatasetDetail)
+const mockUseDatasetRelatedApps = vi.mocked(useDatasetRelatedApps)
 
 describe('DatasetDetailLayout', () => {
   beforeEach(() => {
     vi.clearAllMocks()
-    mockPathname = '/datasets/test-dataset-id/documents'
-    mockDataset = createDataset()
-    mockCanAccessSnippetsAndEvaluation = true
+    mockUsePathname.mockReturnValue('/datasets/dataset-1/pipeline')
+    mockUseRouter.mockReturnValue({
+      back: vi.fn(),
+      forward: vi.fn(),
+      refresh: vi.fn(),
+      push: vi.fn(),
+      replace: mockReplace,
+      prefetch: vi.fn(),
+    })
+    mockUseDatasetRelatedApps.mockReturnValue({ data: undefined } as ReturnType<typeof useDatasetRelatedApps>)
   })
 
-  describe('Evaluation navigation', () => {
-    it('should hide the evaluation menu when the dataset is not a rag pipeline', () => {
-      mockDataset = createDataset({
-        runtime_mode: 'general',
-        is_published: false,
+  describe('Access Errors', () => {
+    it.each([403, 404])('should redirect to datasets page when dataset detail returns %s', async (status) => {
+      // Arrange
+      mockUseDatasetDetail.mockReturnValue({
+        data: undefined,
+        error: new Response(null, { status }),
+        refetch: vi.fn(),
+      } as unknown as ReturnType<typeof useDatasetDetail>)
+
+      // Act
+      render(
+        <DatasetDetailLayout datasetId="dataset-1">
+          <div>Pipeline content</div>
+        </DatasetDetailLayout>,
+      )
+
+      // Assert
+      await waitFor(() => {
+        expect(mockReplace).toHaveBeenCalledWith('/datasets')
       })
+      expect(mockUseDatasetRelatedApps).toHaveBeenCalledWith('dataset-1', { enabled: false })
+      expect(screen.queryByText('Pipeline content')).not.toBeInTheDocument()
+    })
 
+    it('should redirect when the dataset detail error exposes status without being a Response', async () => {
+      // Arrange
+      mockUseDatasetDetail.mockReturnValue({
+        data: undefined,
+        error: { status: 403 },
+        refetch: vi.fn(),
+      } as unknown as ReturnType<typeof useDatasetDetail>)
+
+      // Act
       render(
-        <DatasetDetailLayout datasetId="test-dataset-id">
-          <div data-testid="dataset-detail-content">content</div>
+        <DatasetDetailLayout datasetId="dataset-1">
+          <div>Pipeline content</div>
         </DatasetDetailLayout>,
       )
 
-      expect(screen.queryByRole('button', { name: 'common.datasetMenus.evaluation' })).not.toBeInTheDocument()
-    })
-
-    it('should disable the evaluation menu when the rag pipeline is unpublished', () => {
-      mockDataset = createDataset({
-        is_published: false,
+      // Assert
+      await waitFor(() => {
+        expect(mockReplace).toHaveBeenCalledWith('/datasets')
       })
-
-      render(
-        <DatasetDetailLayout datasetId="test-dataset-id">
-          <div data-testid="dataset-detail-content">content</div>
-        </DatasetDetailLayout>,
-      )
-
-      expect(screen.getByRole('button', { name: 'common.datasetMenus.evaluation' })).toBeDisabled()
+      expect(screen.queryByText('Pipeline content')).not.toBeInTheDocument()
     })
+  })
 
-    it('should enable the evaluation menu when the rag pipeline is published', () => {
+  describe('Rendering', () => {
+    it('should render children when dataset detail is available', () => {
+      // Arrange
+      mockUseDatasetDetail.mockReturnValue({
+        data: {
+          id: 'dataset-1',
+          name: 'Dataset 1',
+          provider: 'vendor',
+          runtime_mode: 'rag_pipeline',
+          is_published: true,
+        },
+        error: null,
+        refetch: vi.fn(),
+      } as unknown as ReturnType<typeof useDatasetDetail>)
+
+      // Act
       render(
-        <DatasetDetailLayout datasetId="test-dataset-id">
-          <div data-testid="dataset-detail-content">content</div>
+        <DatasetDetailLayout datasetId="dataset-1">
+          <div>Pipeline content</div>
         </DatasetDetailLayout>,
       )
 
-      expect(screen.getByRole('button', { name: 'common.datasetMenus.evaluation' })).toBeEnabled()
-    })
-
-    it('should hide the evaluation menu when snippet and evaluation access is unavailable', () => {
-      mockCanAccessSnippetsAndEvaluation = false
-
-      render(
-        <DatasetDetailLayout datasetId="test-dataset-id">
-          <div data-testid="dataset-detail-content">content</div>
-        </DatasetDetailLayout>,
-      )
-
-      expect(screen.queryByRole('button', { name: 'common.datasetMenus.evaluation' })).not.toBeInTheDocument()
+      // Assert
+      expect(screen.getByText('Pipeline content')).toBeInTheDocument()
+      expect(mockUseDatasetRelatedApps).toHaveBeenCalledWith('dataset-1', { enabled: true })
+      expect(mockReplace).not.toHaveBeenCalled()
     })
   })
 })

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

@@ -9,8 +9,6 @@ import {
   RiFileTextLine,
   RiFocus2Fill,
   RiFocus2Line,
-  RiUserSettingsFill,
-  RiUserSettingsLine,
 } from '@remixicon/react'
 import * as React from 'react'
 import { useEffect, useMemo, useState } from 'react'
@@ -25,8 +23,7 @@ import DatasetDetailContext from '@/context/dataset-detail'
 import { useEventEmitterContextContext } from '@/context/event-emitter'
 import useBreakpoints, { MediaType } from '@/hooks/use-breakpoints'
 import useDocumentTitle from '@/hooks/use-document-title'
-import { useSnippetAndEvaluationPlanAccess } from '@/hooks/use-snippet-and-evaluation-plan-access'
-import { usePathname } from '@/next/navigation'
+import { usePathname, useRouter } from '@/next/navigation'
 import { useDatasetDetail, useDatasetRelatedApps } from '@/service/knowledge/use-dataset'
 
 type IAppDetailLayoutProps = {
@@ -34,8 +31,17 @@ type IAppDetailLayoutProps = {
   datasetId: string
 }
 
-const EvaluationIcon = ({ className }: { className?: string }) => {
-  return <span aria-hidden className={cn('i-custom-vender-line-others-evaluation', className)} />
+const getResponseStatus = (error: unknown) => {
+  if (error instanceof Response)
+    return error.status
+
+  if (typeof error === 'object' && error && 'status' in error && typeof error.status === 'number')
+    return error.status
+}
+
+const shouldRedirectToDatasetList = (error: unknown) => {
+  const status = getResponseStatus(error)
+  return status === 403 || status === 404
 }
 
 const DatasetDetailLayout: FC<IAppDetailLayoutProps> = (props) => {
@@ -44,6 +50,7 @@ const DatasetDetailLayout: FC<IAppDetailLayoutProps> = (props) => {
     datasetId,
   } = props
   const { t } = useTranslation()
+  const router = useRouter()
   const pathname = usePathname()
   const hideSideBar = pathname.endsWith('documents/create') || pathname.endsWith('documents/create-from-pipeline')
   const isPipelineCanvas = pathname.endsWith('/pipeline')
@@ -56,15 +63,14 @@ const DatasetDetailLayout: FC<IAppDetailLayoutProps> = (props) => {
       setHideHeader(v.payload)
   })
   const { isCurrentWorkspaceDatasetOperator } = useAppContext()
-  const { canAccess: canAccessSnippetsAndEvaluation } = useSnippetAndEvaluationPlanAccess()
 
   const media = useBreakpoints()
   const isMobile = media === MediaType.mobile
 
   const { data: datasetRes, error, refetch: mutateDatasetRes } = useDatasetDetail(datasetId)
+  const shouldRedirect = shouldRedirectToDatasetList(error)
 
-  const { data: relatedApps } = useDatasetRelatedApps(datasetId)
-  const isRagPipelineDataset = datasetRes?.runtime_mode === 'rag_pipeline'
+  const { data: relatedApps } = useDatasetRelatedApps(datasetId, { enabled: !!datasetRes && !shouldRedirect })
 
   const isButtonDisabledWithPipeline = useMemo(() => {
     if (!datasetRes)
@@ -92,46 +98,27 @@ const DatasetDetailLayout: FC<IAppDetailLayoutProps> = (props) => {
         selectedIcon: RiEqualizer2Fill,
         disabled: false,
       },
-      {
-        name: 'Access Config',
-        href: `/datasets/${datasetId}/access-config`,
-        icon: RiUserSettingsLine,
-        selectedIcon: RiUserSettingsFill,
-        disabled: false,
-      },
     ]
 
     if (datasetRes?.provider !== 'external') {
-      return [
-        {
-          name: t('datasetMenus.documents', { ns: 'common' }),
-          href: `/datasets/${datasetId}/documents`,
-          icon: RiFileTextLine,
-          selectedIcon: RiFileTextFill,
-          disabled: isButtonDisabledWithPipeline,
-        },
-        {
-          name: t('datasetMenus.pipeline', { ns: 'common' }),
-          href: `/datasets/${datasetId}/pipeline`,
-          icon: PipelineLine as RemixiconComponentType,
-          selectedIcon: PipelineFill as RemixiconComponentType,
-          disabled: false,
-        },
-        ...baseNavigation,
-        ...(isRagPipelineDataset && canAccessSnippetsAndEvaluation
-          ? [{
-              name: t('datasetMenus.evaluation', { ns: 'common' }),
-              href: `/datasets/${datasetId}/evaluation`,
-              icon: EvaluationIcon,
-              selectedIcon: EvaluationIcon,
-              disabled: isButtonDisabledWithPipeline,
-            }]
-          : []),
-      ]
+      baseNavigation.unshift({
+        name: t('datasetMenus.pipeline', { ns: 'common' }),
+        href: `/datasets/${datasetId}/pipeline`,
+        icon: PipelineLine as RemixiconComponentType,
+        selectedIcon: PipelineFill as RemixiconComponentType,
+        disabled: false,
+      })
+      baseNavigation.unshift({
+        name: t('datasetMenus.documents', { ns: 'common' }),
+        href: `/datasets/${datasetId}/documents`,
+        icon: RiFileTextLine,
+        selectedIcon: RiFileTextFill,
+        disabled: isButtonDisabledWithPipeline,
+      })
     }
 
     return baseNavigation
-  }, [canAccessSnippetsAndEvaluation, t, datasetId, isButtonDisabledWithPipeline, isRagPipelineDataset, datasetRes?.provider])
+  }, [t, datasetId, isButtonDisabledWithPipeline, datasetRes?.provider])
 
   useDocumentTitle(datasetRes?.name || t('menus.datasets', { ns: 'common' }))
 
@@ -143,9 +130,17 @@ const DatasetDetailLayout: FC<IAppDetailLayoutProps> = (props) => {
     setAppSidebarExpand(isMobile ? mode : localeMode)
   }, [isMobile, setAppSidebarExpand])
 
+  useEffect(() => {
+    if (shouldRedirect)
+      router.replace('/datasets')
+  }, [router, shouldRedirect])
+
   if (!datasetRes && !error)
     return <Loading type="app" />
 
+  if (shouldRedirect)
+    return <Loading type="app" />
+
   return (
     <div
       className={cn(
@@ -175,4 +170,4 @@ const DatasetDetailLayout: FC<IAppDetailLayoutProps> = (props) => {
     </div>
   )
 }
-export default React.memo(DatasetDetailLayout)
+export default React.memo(DatasetDetailLayout)

web/app/components/base/prompt-editor/__tests__/index.spec.tsx

@@ -36,6 +36,7 @@ const mocks = vi.hoisted(() => {
       })),
       parseEditorState: vi.fn(() => ({ state: 'parsed' })),
       setEditorState: vi.fn(),
+      setEditable: vi.fn(),
       focus: vi.fn(),
       update: vi.fn((fn: () => void) => fn()),
     },
@@ -71,6 +72,7 @@ vi.mock('lexical', async (importOriginal) => {
       })),
       getAllTextNodes: () => [],
     }),
+    $nodesOfType: () => [],
     TextNode: class TextNode {
       __text: string
       constructor(text = '') {
@@ -92,9 +94,8 @@ vi.mock('@lexical/react/LexicalComposer', () => ({
       try {
         initialConfig.onError(new Error('test error'))
       }
-      catch (e) {
-        // ignore error
-        console.error(e)
+      catch {
+        // Ignore the intentional throw from the mocked error boundary path.
       }
     }
     if (initialConfig?.nodes) {
@@ -328,6 +329,20 @@ describe('PromptEditor', () => {
       expect(screen.getByTestId('lexical-composer')).toBeInTheDocument()
     })
 
+    it('should sync editable changes to the lexical editor instance', async () => {
+      const { rerender } = render(<PromptEditor editable={true} />)
+
+      await waitFor(() => {
+        expect(mocks.editor.setEditable).toHaveBeenCalledWith(true)
+      })
+
+      rerender(<PromptEditor editable={false} />)
+
+      await waitFor(() => {
+        expect(mocks.editor.setEditable).toHaveBeenLastCalledWith(false)
+      })
+    })
+
     it('should render with isSupportFileVar=true', () => {
       render(<PromptEditor isSupportFileVar={true} />)
       expect(screen.getByTestId('lexical-composer')).toBeInTheDocument()

web/app/components/base/prompt-editor/index.tsx

@@ -3,10 +3,9 @@
 import type { InitialConfigType } from '@lexical/react/LexicalComposer'
 import type {
   EditorState,
-  LexicalCommand,
 } from 'lexical'
 import type { FC } from 'react'
-import type { Hotkey } from './plugins/shortcuts-popup-plugin'
+import type { Hotkey, ShortcutPopupInsertHandler } from './plugins/shortcuts-popup-plugin'
 import type {
   ContextBlockType,
   CurrentBlockType,
@@ -97,6 +96,16 @@ const ValueSyncPlugin: FC<{ value?: string }> = ({ value }) => {
   return null
 }
 
+const EditableSyncPlugin: FC<{ editable: boolean }> = ({ editable }) => {
+  const [editor] = useLexicalComposerContext()
+
+  useEffect(() => {
+    editor.setEditable(editable)
+  }, [editor, editable])
+
+  return null
+}
+
 export type PromptEditorProps = {
   instanceId?: string
   compact?: boolean
@@ -122,7 +131,7 @@ export type PromptEditorProps = {
   errorMessageBlock?: ErrorMessageBlockType
   lastRunBlock?: LastRunBlockType
   isSupportFileVar?: boolean
-  shortcutPopups?: Array<{ hotkey: Hotkey, Popup: React.ComponentType<{ onClose: () => void, onInsert: (command: LexicalCommand<unknown>, params: any[]) => void }> }>
+  shortcutPopups?: Array<{ hotkey: Hotkey, Popup: React.ComponentType<{ onClose: () => void, onInsert: ShortcutPopupInsertHandler }> }>
 }
 
 const PromptEditor: FC<PromptEditorProps> = ({
@@ -194,13 +203,13 @@ const PromptEditor: FC<PromptEditorProps> = ({
     eventEmitter?.emit({
       type: UPDATE_DATASETS_EVENT_EMITTER,
       payload: contextBlock?.datasets,
-    } as any)
+    })
   }, [eventEmitter, contextBlock?.datasets])
   useEffect(() => {
     eventEmitter?.emit({
       type: UPDATE_HISTORY_EVENT_EMITTER,
       payload: historyBlock?.history,
-    } as any)
+    })
   }, [eventEmitter, historyBlock?.history])
 
   const [floatingAnchorElem, setFloatingAnchorElem] = useState<HTMLDivElement | null>(null)
@@ -243,6 +252,7 @@ const PromptEditor: FC<PromptEditorProps> = ({
           onEditorChange={handleEditorChange}
         />
         <ValueSyncPlugin value={value} />
+        <EditableSyncPlugin editable={editable} />
       </div>
     </LexicalComposer>
   )

web/app/components/base/prompt-editor/plugins/hitl-input-block/__tests__/component-ui.spec.tsx

@@ -4,8 +4,10 @@ import type { FormInputItem, ParagraphFormInput } from '@/app/components/workflo
 import type { ValueSelector } from '@/app/components/workflow/types'
 
 import { LexicalComposer } from '@lexical/react/LexicalComposer'
-import { cleanup, fireEvent, render, screen } from '@testing-library/react'
-import { BlockEnum, InputVarType } from '@/app/components/workflow/types'
+import { act, cleanup, fireEvent, render, screen, waitFor } from '@testing-library/react'
+import { useEffect, useState } from 'react'
+import { BlockEnum, InputVarType, SupportUploadFileTypes } from '@/app/components/workflow/types'
+import { TransferMethod } from '@/types/app'
 import HITLInputComponentUI from '../component-ui'
 import { HITLInputNode } from '../node'
 
@@ -113,6 +115,57 @@ describe('HITLInputComponentUI', () => {
       expect(screen.queryByRole('button', { name: 'common.operation.remove' })).not.toBeInTheDocument()
     })
 
+    it('should close the edit modal when readonly becomes true', async () => {
+      let setReadonlyValue: ((readonly: boolean) => void) | undefined
+      const Harness = () => {
+        const [readonly, setReadonly] = useState(false)
+        const [namespace] = useState(() => `hitl-input-test-${crypto.randomUUID()}`)
+
+        useEffect(() => {
+          setReadonlyValue = setReadonly
+          return () => {
+            setReadonlyValue = undefined
+          }
+        }, [])
+
+        return (
+          <LexicalComposer
+            initialConfig={{
+              namespace,
+              onError: (error: Error) => {
+                throw error
+              },
+              nodes: [HITLInputNode],
+            }}
+          >
+            <HITLInputComponentUI
+              nodeId="node-1"
+              varName="customer_name"
+              workflowNodesMap={createWorkflowNodesMap()}
+              onChange={vi.fn()}
+              onRename={vi.fn()}
+              onRemove={vi.fn()}
+              readonly={readonly}
+            />
+          </LexicalComposer>
+        )
+      }
+
+      render(<Harness />)
+
+      fireEvent.click(await screen.findByRole('button', { name: 'common.operation.edit' }))
+
+      expect(await screen.findByRole('textbox')).toBeInTheDocument()
+
+      act(() => {
+        setReadonlyValue?.(true)
+      })
+
+      await waitFor(() => {
+        expect(screen.queryByRole('textbox')).not.toBeInTheDocument()
+      })
+    })
+
     it('should render select option summary for constant options', () => {
       const { getByText } = renderComponent({
         formInput: {
@@ -212,10 +265,33 @@ describe('HITLInputComponentUI', () => {
 
       expect(queryByRole('textbox')).not.toBeInTheDocument()
     })
+
+    it('should prevent renaming to an existing variable name', async () => {
+      const {
+        findByRole,
+        onChange,
+        onRename,
+      } = renderComponent({
+        unavailableVariableNames: ['existing_name'],
+      })
+
+      fireEvent.click(await screen.findByRole('button', { name: 'common.operation.edit' }))
+
+      const textbox = await findByRole('textbox')
+      fireEvent.change(textbox, { target: { value: 'existing_name' } })
+
+      expect(screen.getByText('workflow.nodes.humanInput.insertInputField.variableNameDuplicated')).toBeInTheDocument()
+      expect(screen.getByRole('button', { name: 'common.operation.save' })).toBeDisabled()
+
+      fireEvent.click(screen.getByRole('button', { name: 'common.operation.save' }))
+
+      expect(onChange).not.toHaveBeenCalled()
+      expect(onRename).not.toHaveBeenCalled()
+    })
   })
 
   describe('Default formInput', () => {
-    it('should pass default payload to InputField when formInput is undefined', async () => {
+    it('should open an empty default editor when formInput is undefined', async () => {
       const { findByRole } = renderComponent({
         formInput: undefined,
       })
@@ -223,10 +299,10 @@ describe('HITLInputComponentUI', () => {
       fireEvent.click(await screen.findByRole('button', { name: 'common.operation.edit' }))
 
       const textbox = await findByRole('textbox')
+      const saveButton = await screen.findByRole('button', { name: 'common.operation.save' })
 
-      fireEvent.click(await screen.findByRole('button', { name: 'common.operation.save' }))
-
-      expect(textbox).toHaveValue('customer_name')
+      expect(textbox).toHaveValue('')
+      expect(saveButton).toBeDisabled()
     })
 
     it('should render variable selector when workflowNodesMap fallback is used', () => {

web/app/components/base/prompt-editor/plugins/hitl-input-block/__tests__/component.spec.tsx

@@ -129,6 +129,31 @@ describe('HITLInputComponent', () => {
     expect(onChange.mock.calls[0][0][0].output_variable_name).toBe('renamed_name')
   })
 
+  it('should ignore rename when the target variable name already exists', async () => {
+    const user = userEvent.setup()
+    const onChange = vi.fn()
+
+    render(
+      <HITLInputComponent
+        nodeKey="node-key-duplicate"
+        nodeId="node-duplicate"
+        varName="user_name"
+        formInputs={[
+          createInput(),
+          createInput({ output_variable_name: 'renamed_name' }),
+        ]}
+        onChange={onChange}
+        onRename={vi.fn()}
+        onRemove={vi.fn()}
+        workflowNodesMap={{}}
+      />,
+    )
+
+    await user.click(screen.getByRole('button', { name: 'emit-rename' }))
+
+    expect(onChange).not.toHaveBeenCalled()
+  })
+
   it('should update existing payload when variable name stays the same', async () => {
     const user = userEvent.setup()
     const onChange = vi.fn()

web/app/components/base/prompt-editor/plugins/hitl-input-block/__tests__/index.spec.tsx

@@ -1,11 +1,13 @@
+import type { LexicalEditor } from 'lexical'
 import type { FormInputItem } from '@/app/components/workflow/nodes/human-input/types'
 import { LexicalComposer } from '@lexical/react/LexicalComposer'
 import { useLexicalComposerContext } from '@lexical/react/LexicalComposerContext'
 import { act, render, waitFor } from '@testing-library/react'
 import {
+  $nodesOfType,
   COMMAND_PRIORITY_EDITOR,
 } from 'lexical'
-import { useEffect } from 'react'
+import { useEffect, useState } from 'react'
 import {
   BlockEnum,
   InputVarType,
@@ -13,6 +15,7 @@ import {
 import { CustomTextNode } from '../../custom-text/node'
 import {
   getNodeCount,
+  readEditorStateValue,
   readRootTextContent,
   renderLexicalEditor,
   selectRootEnd,
@@ -76,6 +79,12 @@ const createInsertPayload = () => ({
   onFormInputItemRemove: vi.fn(),
 })
 
+const readHITLReadonlyValues = (editor: LexicalEditor): boolean[] => {
+  return readEditorStateValue(editor, () => {
+    return $nodesOfType(HITLInputNode).map(node => node.getReadonly())
+  })
+}
+
 const renderHITLInputBlock = (props?: {
   onInsert?: () => void
   onDelete?: () => void
@@ -169,6 +178,65 @@ describe('HITLInputBlock', () => {
       expect(getNodeCount(editor, HITLInputNode)).toBe(1)
     })
 
+    it('should update existing and newly inserted nodes when readonly changes', async () => {
+      let setReadonlyValue: ((readonly: boolean) => void) | undefined
+      const ReadonlyHarness = () => {
+        const [readonly, setReadonly] = useState(false)
+
+        useEffect(() => {
+          setReadonlyValue = setReadonly
+          return () => {
+            setReadonlyValue = undefined
+          }
+        }, [])
+
+        return (
+          <HITLInputBlock
+            nodeId="node-1"
+            formInputs={[createFormInput()]}
+            onFormInputItemRename={vi.fn()}
+            onFormInputItemRemove={vi.fn()}
+            workflowNodesMap={createWorkflowNodesMap('First Node')}
+            readonly={readonly}
+          />
+        )
+      }
+
+      const { getEditor } = renderLexicalEditor({
+        namespace: 'hitl-input-block-readonly-update-test',
+        nodes: [CustomTextNode, HITLInputNode],
+        children: <ReadonlyHarness />,
+      })
+
+      const editor = await waitForEditorReady(getEditor)
+
+      selectRootEnd(editor)
+      act(() => {
+        editor.dispatchCommand(INSERT_HITL_INPUT_BLOCK_COMMAND, createInsertPayload())
+      })
+
+      await waitFor(() => {
+        expect(readHITLReadonlyValues(editor)).toEqual([false])
+      })
+
+      act(() => {
+        setReadonlyValue?.(true)
+      })
+
+      await waitFor(() => {
+        expect(readHITLReadonlyValues(editor)).toEqual([true])
+      })
+
+      selectRootEnd(editor)
+      act(() => {
+        editor.dispatchCommand(INSERT_HITL_INPUT_BLOCK_COMMAND, createInsertPayload())
+      })
+
+      await waitFor(() => {
+        expect(readHITLReadonlyValues(editor)).toEqual([true, true])
+      })
+    })
+
     it('should call onDelete when delete command is dispatched', async () => {
       const onDelete = vi.fn()
       const { getEditor } = renderHITLInputBlock({ onDelete })

web/app/components/base/prompt-editor/plugins/hitl-input-block/__tests__/input-field.spec.tsx

@@ -116,6 +116,31 @@ describe('InputField', () => {
     expect(onChange).not.toHaveBeenCalled()
   })
 
+  it('should disable save and show validation error when variable name already exists', async () => {
+    const user = userEvent.setup()
+    const onChange = vi.fn()
+
+    render(
+      <InputField
+        nodeId="node-duplicate-name"
+        isEdit={false}
+        payload={createPayload()}
+        unavailableVariableNames={['existing_name']}
+        onChange={onChange}
+        onCancel={vi.fn()}
+      />,
+    )
+
+    const inputs = screen.getAllByRole('textbox')
+    await user.clear(inputs[0]!)
+    await user.type(inputs[0]!, 'existing_name')
+
+    expect(screen.getByText('workflow.nodes.humanInput.insertInputField.variableNameDuplicated')).toBeInTheDocument()
+    expect(screen.getByRole('button', { name: /workflow\.nodes\.humanInput\.insertInputField\.insert/i })).toBeDisabled()
+    await user.keyboard('{Control>}{Enter}{/Control}')
+    expect(onChange).not.toHaveBeenCalled()
+  })
+
   it('should call onChange when saving a valid payload in edit mode', async () => {
     const user = userEvent.setup()
     const onChange = vi.fn()

web/app/components/base/prompt-editor/plugins/hitl-input-block/__tests__/node.spec.tsx

@@ -98,6 +98,8 @@ describe('HITLInputNode', () => {
         expect(node.getConversationVariables()).toEqual(props.conversationVariables)
         expect(node.getRagVariables()).toEqual(props.ragVariables)
         expect(node.getReadonly()).toBe(true)
+        node.setReadonly(false)
+        expect(node.getReadonly()).toBe(false)
         expect(node.getTextContent()).toBe('{{#$output.user_name#}}')
       })
     })

web/app/components/base/prompt-editor/plugins/hitl-input-block/component-ui.tsx

@@ -26,6 +26,7 @@ type HITLInputComponentUIProps = {
   nodeId: string
   varName: string
   formInput?: FormInputItem
+  unavailableVariableNames?: string[]
   onChange: (input: FormInputItem) => void
   onRename: (payload: FormInputItem, oldName: string) => void
   onRemove: (varName: string) => void
@@ -44,6 +45,7 @@ const HITLInputComponentUI: FC<HITLInputComponentUIProps> = ({
   nodeId,
   varName,
   formInput,
+  unavailableVariableNames = [],
   onChange,
   onRename,
   onRemove,
@@ -64,6 +66,11 @@ const HITLInputComponentUI: FC<HITLInputComponentUIProps> = ({
     setFalse: hideEditModal,
   }] = useBoolean(false)
 
+  useEffect(() => {
+    if (readonly)
+      hideEditModal()
+  }, [hideEditModal, readonly])
+
   // Lexical delegate the click make it unable to add click by the method of react
   const editBtnRef = useRef<HTMLDivElement>(null)
   useEffect(() => {
@@ -91,12 +98,15 @@ const HITLInputComponentUI: FC<HITLInputComponentUIProps> = ({
   }, [onRemove, varName])
 
   const handleChange = useCallback((newPayload: FormInputItem) => {
+    if (newPayload.output_variable_name !== varName && unavailableVariableNames.includes(newPayload.output_variable_name))
+      return
+
     if (varName === newPayload.output_variable_name)
       onChange(newPayload)
     else
       onRename(newPayload, varName)
     hideEditModal()
-  }, [hideEditModal, onChange, onRename, varName])
+  }, [hideEditModal, onChange, onRename, unavailableVariableNames, varName])
 
   const isDefaultValueVariable = useMemo(() => {
     return paragraphDefault?.type === 'variable'
@@ -203,6 +213,7 @@ const HITLInputComponentUI: FC<HITLInputComponentUIProps> = ({
               nodeId={nodeId}
               isEdit
               payload={formInput}
+              unavailableVariableNames={unavailableVariableNames}
               onChange={handleChange}
               onCancel={hideEditModal}
             />

web/app/components/base/prompt-editor/plugins/hitl-input-block/component.tsx

@@ -45,8 +45,14 @@ const HITLInputComponent: FC<HITLInputComponentProps> = ({
 }) => {
   const [ref] = useSelectOrDelete(nodeKey, DELETE_HITL_INPUT_BLOCK_COMMAND)
   const payload = formInputs.find(item => item.output_variable_name === varName)
+  const unavailableVariableNames = formInputs
+    .map(item => item.output_variable_name)
+    .filter(name => name !== varName)
 
   const handleChange = useCallback((newPayload: FormInputItem) => {
+    if (newPayload.output_variable_name !== varName && unavailableVariableNames.includes(newPayload.output_variable_name))
+      return
+
     if (!payload) {
       onChange([...formInputs, newPayload])
       return
@@ -58,7 +64,7 @@ const HITLInputComponent: FC<HITLInputComponentProps> = ({
       return
     }
     onChange(formInputs.map(item => item.output_variable_name === varName ? newPayload : item))
-  }, [formInputs, onChange, payload, varName])
+  }, [formInputs, onChange, payload, unavailableVariableNames, varName])
 
   return (
     <div
@@ -69,6 +75,7 @@ const HITLInputComponent: FC<HITLInputComponentProps> = ({
         nodeId={nodeId}
         varName={varName}
         formInput={payload}
+        unavailableVariableNames={unavailableVariableNames}
         onChange={handleChange}
         onRename={onRename}
         onRemove={onRemove}

web/app/components/base/prompt-editor/plugins/hitl-input-block/hitl-input-block-replacement-block.tsx

@@ -1,5 +1,6 @@
 import type { TextNode } from 'lexical'
 import type { HITLInputBlockType } from '../../types'
+import type { Var } from '@/app/components/workflow/types'
 import { useLexicalComposerContext } from '@lexical/react/LexicalComposerContext'
 import { mergeRegister } from '@lexical/utils'
 import { $applyNodeReplacement } from 'lexical'
@@ -31,7 +32,7 @@ const HITLInputReplacementBlock = ({
 
   const environmentVariables = useMemo(() => variables?.find(o => o.nodeId === 'env')?.vars || [], [variables])
   const conversationVariables = useMemo(() => variables?.find(o => o.nodeId === 'conversation')?.vars || [], [variables])
-  const ragVariables = useMemo(() => variables?.reduce<any[]>((acc, curr) => {
+  const ragVariables = useMemo(() => variables?.reduce<Var[]>((acc, curr) => {
     if (curr.nodeId === 'rag')
       acc.push(...curr.vars)
     else
@@ -81,7 +82,7 @@ const HITLInputReplacementBlock = ({
     return mergeRegister(
       editor.registerNodeTransform(CustomTextNode, textNode => decoratorTransform(textNode, getMatch, createHITLInputBlockNode)),
     )
-  }, [])
+  }, [editor, getMatch, createHITLInputBlockNode])
 
   return null
 }

web/app/components/base/prompt-editor/plugins/hitl-input-block/index.tsx

@@ -6,6 +6,7 @@ import { useLexicalComposerContext } from '@lexical/react/LexicalComposerContext
 import { mergeRegister } from '@lexical/utils'
 import {
   $insertNodes,
+  $nodesOfType,
   COMMAND_PRIORITY_EDITOR,
   createCommand,
 } from 'lexical'
@@ -43,6 +44,14 @@ const HITLInputBlock = memo(({
     })
   }, [editor, workflowNodesMap, workflowAvailableVariables])
 
+  useEffect(() => {
+    editor.update(() => {
+      $nodesOfType(HITLInputNode).forEach((node) => {
+        node.setReadonly(readonly)
+      })
+    })
+  }, [editor, readonly])
+
   useEffect(() => {
     if (!editor.hasNodes([HITLInputNode]))
       throw new Error('HITLInputBlockPlugin: HITLInputBlock not registered on editor')
@@ -95,7 +104,7 @@ const HITLInputBlock = memo(({
         COMMAND_PRIORITY_EDITOR,
       ),
     )
-  }, [editor, onInsert, onDelete])
+  }, [editor, onInsert, onDelete, workflowNodesMap, getVarType, readonly])
 
   return null
 })

web/app/components/base/prompt-editor/plugins/hitl-input-block/input-field.tsx

@@ -31,6 +31,7 @@ type InputFieldProps = {
   nodeId: string
   isEdit: boolean
   payload?: FormInputItem
+  unavailableVariableNames?: string[]
   onChange: (newPayload: FormInputItem) => void
   onCancel: () => void
 }
@@ -38,6 +39,7 @@ const InputField: React.FC<InputFieldProps> = ({
   nodeId,
   isEdit,
   payload,
+  unavailableVariableNames = [],
   onChange,
   onCancel,
 }) => {
@@ -73,14 +75,24 @@ const InputField: React.FC<InputFieldProps> = ({
 
     return createDefaultParagraphFormInput(tempPayload.output_variable_name)
   }, [tempPayload])
-  const nameValid = useMemo(() => {
+  const unavailableVariableNameSet = useMemo(() => {
+    return new Set(unavailableVariableNames.map(name => name.trim()).filter(Boolean))
+  }, [unavailableVariableNames])
+  const variableNameError = useMemo(() => {
     const name = tempPayload.output_variable_name.trim()
     if (!name)
-      return false
+      return null
     if (name.includes(' '))
-      return false
-    return /^[a-z_]\w{0,29}$/.test(name)
-  }, [tempPayload.output_variable_name])
+      return 'variableNameInvalid'
+    if (!/^[a-z_]\w{0,29}$/.test(name))
+      return 'variableNameInvalid'
+    if (unavailableVariableNameSet.has(name))
+      return 'variableNameDuplicated'
+    return null
+  }, [tempPayload.output_variable_name, unavailableVariableNameSet])
+  const nameValid = useMemo(() => {
+    return !!tempPayload.output_variable_name.trim() && !variableNameError
+  }, [tempPayload.output_variable_name, variableNameError])
   const handleSave = useCallback(() => {
     if (!nameValid)
       return
@@ -223,9 +235,9 @@ const InputField: React.FC<InputFieldProps> = ({
           }}
           autoFocus
         />
-        {tempPayload.output_variable_name && !nameValid && (
+        {tempPayload.output_variable_name && variableNameError && (
           <div className="mt-1 px-1 system-xs-regular text-text-destructive-secondary">
-            {t(`${i18nPrefix}.variableNameInvalid`, { ns: 'workflow' })}
+            {t(`${i18nPrefix}.${variableNameError}`, { ns: 'workflow' })}
           </div>
         )}
       </div>

web/app/components/base/prompt-editor/plugins/hitl-input-block/node.tsx

@@ -109,6 +109,11 @@ export class HITLInputNode extends DecoratorNode<React.JSX.Element> {
     return self.__readonly || false
   }
 
+  setReadonly(readonly?: boolean): void {
+    const self = this.getWritable()
+    self.__readonly = readonly
+  }
+
   static override clone(node: HITLInputNode): HITLInputNode {
     return new HITLInputNode(
       node.__variableName,

web/app/components/workflow/nodes/human-input/components/__tests__/form-content.spec.tsx

@@ -64,6 +64,7 @@ vi.mock('@/app/components/base/prompt-editor', () => ({
 vi.mock('../add-input-field', () => ({
   __esModule: true,
   default: (props: {
+    unavailableVariableNames?: string[]
     onSave: (payload: {
       type: string
       output_variable_name: string
@@ -231,6 +232,41 @@ describe('FormContent', () => {
     expect(container.firstChild).toHaveClass('pointer-events-none')
   })
 
+  it('should not insert a new input when the variable name already exists', () => {
+    render(
+      <FormContent
+        nodeId="node-2"
+        value="Initial content"
+        onChange={onChange}
+        formInputs={[{
+          type: 'paragraph',
+          output_variable_name: 'approval',
+          default: {
+            type: 'constant',
+            selector: [],
+            value: '',
+          },
+        } as never]}
+        onFormInputsChange={onFormInputsChange}
+        onFormInputItemRename={onFormInputItemRename}
+        onFormInputItemRemove={onFormInputItemRemove}
+        editorKey={1}
+        isExpand={false}
+        availableVars={[]}
+        availableNodes={[]}
+      />,
+    )
+
+    expect(mockAddInputField).toHaveBeenCalledWith(expect.objectContaining({
+      unavailableVariableNames: ['approval'],
+    }))
+
+    fireEvent.click(screen.getByText('save-input'))
+
+    expect(mockOnInsert).not.toHaveBeenCalled()
+    expect(onFormInputsChange).not.toHaveBeenCalled()
+  })
+
   it('should render the mac hotkey hint when focused on macOS', () => {
     mockIsMac.mockReturnValue(true)
 

web/app/components/workflow/nodes/human-input/components/__tests__/single-run-form.spec.tsx

@@ -1,4 +1,5 @@
-import type { ReactNode } from 'react'
+import type { HumanInputFieldValue } from '@/app/components/base/chat/chat/answer/human-input-content/field-renderer'
+import type { FormInputItem } from '@/app/components/workflow/nodes/human-input/types'
 import type { HumanInputFormData } from '@/types/workflow'
 import { render, screen, waitFor } from '@testing-library/react'
 import userEvent from '@testing-library/user-event'
@@ -12,25 +13,54 @@ vi.mock('react-i18next', () => ({
   }),
 }))
 
-vi.mock('@langgenius/dify-ui/button', () => ({
-  Button: ({
-    children,
-    disabled,
-    onClick,
-  }: {
-    children?: ReactNode
-    disabled?: boolean
-    onClick?: () => void
-  }) => (
-    <button type="button" disabled={disabled} onClick={onClick}>
-      {children}
-    </button>
-  ),
-}))
-
 vi.mock('@/app/components/base/chat/chat/answer/human-input-content/content-item', () => ({
   __esModule: true,
-  default: ({ content }: { content: string }) => <div>{content}</div>,
+  default: ({
+    content,
+    formInputFields,
+    inputs,
+    onInputChange,
+  }: {
+    content: string
+    formInputFields: FormInputItem[]
+    inputs: Record<string, HumanInputFieldValue>
+    onInputChange: (name: string, value: HumanInputFieldValue) => void
+  }) => {
+    const fieldName = /\{\{#\$output\.([^#]+)#\}\}/.exec(content)?.[1]
+    if (!fieldName)
+      return <div>{content}</div>
+
+    const field = formInputFields.find(field => field.output_variable_name === fieldName)
+    if (!field)
+      return null
+
+    if (field.type === 'select') {
+      return (
+        <select
+          aria-label={fieldName}
+          value={typeof inputs[fieldName] === 'string' ? inputs[fieldName] : ''}
+          onChange={event => onInputChange(fieldName, event.target.value)}
+        >
+          <option value="">Select</option>
+          {field.option_source.value.map(option => (
+            <option key={option} value={option}>{option}</option>
+          ))}
+        </select>
+      )
+    }
+
+    if (field.type === 'paragraph') {
+      return (
+        <textarea
+          aria-label={fieldName}
+          value={typeof inputs[fieldName] === 'string' ? inputs[fieldName] : ''}
+          onChange={event => onInputChange(fieldName, event.target.value)}
+        />
+      )
+    }
+
+    return <div>{fieldName}</div>
+  },
 }))
 
 const createFormData = (overrides: Partial<HumanInputFormData> = {}): HumanInputFormData => ({
@@ -60,6 +90,10 @@ const createFormData = (overrides: Partial<HumanInputFormData> = {}): HumanInput
 })
 
 describe('SingleRunForm', () => {
+  beforeEach(() => {
+    vi.clearAllMocks()
+  })
+
   it('renders the back action as a named button and forwards clicks', async () => {
     const user = userEvent.setup()
     const handleBack = vi.fn()
@@ -99,4 +133,99 @@ describe('SingleRunForm', () => {
       })
     })
   })
+
+  it('submits updated paragraph input values', async () => {
+    const user = userEvent.setup()
+    const onSubmit = vi.fn().mockResolvedValue(undefined)
+
+    render(
+      <SingleRunForm
+        nodeName="Review"
+        data={createFormData()}
+        onSubmit={onSubmit}
+      />,
+    )
+
+    await user.clear(screen.getByRole('textbox', { name: 'review' }))
+    await user.type(screen.getByRole('textbox', { name: 'review' }), 'updated review')
+    await user.click(screen.getByRole('button', { name: 'Approve' }))
+
+    await waitFor(() => {
+      expect(onSubmit).toHaveBeenCalledWith({
+        inputs: { review: 'updated review' },
+        action: 'approve',
+      })
+    })
+  })
+
+  it('uses resolved default values for variable paragraph inputs', async () => {
+    const user = userEvent.setup()
+    const onSubmit = vi.fn().mockResolvedValue(undefined)
+
+    render(
+      <SingleRunForm
+        nodeName="Review"
+        data={createFormData({
+          inputs: [{
+            type: InputVarType.paragraph,
+            output_variable_name: 'review',
+            default: {
+              selector: ['source', 'answer'],
+              type: 'variable',
+              value: 'fallback review',
+            },
+          }],
+          resolved_default_values: {
+            review: 'resolved review',
+          },
+        })}
+        onSubmit={onSubmit}
+      />,
+    )
+
+    await user.click(screen.getByRole('button', { name: 'Approve' }))
+
+    await waitFor(() => {
+      expect(onSubmit).toHaveBeenCalledWith({
+        inputs: { review: 'resolved review' },
+        action: 'approve',
+      })
+    })
+  })
+
+  it('disables submit actions until a select input has a value', async () => {
+    const user = userEvent.setup()
+    const onSubmit = vi.fn().mockResolvedValue(undefined)
+
+    render(
+      <SingleRunForm
+        nodeName="Review"
+        data={createFormData({
+          form_content: 'Choose {{#$output.choice#}}',
+          inputs: [{
+            type: InputVarType.select,
+            output_variable_name: 'choice',
+            option_source: {
+              selector: [],
+              type: 'constant',
+              value: ['approve', 'reject'],
+            },
+          }],
+        })}
+        onSubmit={onSubmit}
+      />,
+    )
+
+    expect(screen.getByRole('button', { name: 'Approve' })).toBeDisabled()
+
+    await user.selectOptions(screen.getByRole('combobox', { name: 'choice' }), 'approve')
+    await user.click(screen.getByRole('button', { name: 'Approve' }))
+
+    await waitFor(() => {
+      expect(onSubmit).toHaveBeenCalledWith({
+        inputs: { choice: 'approve' },
+        action: 'approve',
+      })
+    })
+  })
 })

web/app/components/workflow/nodes/human-input/components/add-input-field.tsx

@@ -6,12 +6,14 @@ import InputField from '@/app/components/base/prompt-editor/plugins/hitl-input-b
 
 type Props = {
   nodeId: string
+  unavailableVariableNames?: string[]
   onSave: (newPayload: FormInputItem) => void
   onCancel: () => void
 }
 
 const AddInputField: FC<Props> = ({
   nodeId,
+  unavailableVariableNames,
   onSave,
   onCancel,
 }) => {
@@ -19,6 +21,7 @@ const AddInputField: FC<Props> = ({
     <InputField
       nodeId={nodeId}
       isEdit={false}
+      unavailableVariableNames={unavailableVariableNames}
       onChange={onSave}
       onCancel={onCancel}
     />

web/app/components/workflow/nodes/human-input/components/form-content.tsx

@@ -1,12 +1,13 @@
 'use client'
-import type { LexicalCommand } from 'lexical'
 import type { FC } from 'react'
 import type { FormInputItem } from '../types'
+import type { ShortcutPopupInsertHandler } from '@/app/components/base/prompt-editor/plugins/shortcuts-popup-plugin'
+import type { WorkflowNodesMap } from '@/app/components/base/prompt-editor/types'
 import type { Node, NodeOutPutVar } from '@/app/components/workflow/types'
 import { cn } from '@langgenius/dify-ui/cn'
 import { useBoolean } from 'ahooks'
 import * as React from 'react'
-import { useEffect, useState } from 'react'
+import { useEffect, useRef } from 'react'
 import { Trans, useTranslation } from 'react-i18next'
 import PromptEditor from '@/app/components/base/prompt-editor'
 import { INSERT_HITL_INPUT_BLOCK_COMMAND } from '@/app/components/base/prompt-editor/plugins/hitl-input-block'
@@ -56,11 +57,20 @@ const FormContent: FC<FormContentProps> = ({
 
   const getVarType = useWorkflowVariableType()
 
-  const [needToAddFormInput, setNeedToAddFormInput] = useState(false)
-  const [newFormInputs, setNewFormInputs] = useState<FormInputItem[]>([])
-  const handleInsertHITLNode = (onInsert: (command: LexicalCommand<unknown>, params: any) => void) => {
+  const pendingFormInputsRef = useRef<{
+    value: string
+    formInputs: FormInputItem[]
+  } | null>(null)
+  const handleInsertHITLNode = (onInsert: ShortcutPopupInsertHandler) => {
     return (payload: FormInputItem) => {
+      if (formInputs.some(input => input.output_variable_name === payload.output_variable_name))
+        return
+
       const newFormInputs = [...(formInputs || []), payload]
+      pendingFormInputsRef.current = {
+        value,
+        formInputs: newFormInputs,
+      }
       onInsert(INSERT_HITL_INPUT_BLOCK_COMMAND, {
         variableName: payload.output_variable_name,
         nodeId,
@@ -69,25 +79,25 @@ const FormContent: FC<FormContentProps> = ({
         onFormInputItemRename,
         onFormInputItemRemove,
       })
-      setNewFormInputs(newFormInputs)
-      setNeedToAddFormInput(true)
     }
   }
 
   // avoid update formInputs would overwrite the value just inserted
   useEffect(() => {
-    if (needToAddFormInput) {
-      onFormInputsChange(newFormInputs)
-      setNeedToAddFormInput(false)
-    }
-  }, [value])
+    const pendingFormInputs = pendingFormInputsRef.current
+    if (!pendingFormInputs || pendingFormInputs.value === value)
+      return
+
+    onFormInputsChange(pendingFormInputs.formInputs)
+    pendingFormInputsRef.current = null
+  }, [onFormInputsChange, value])
 
   const [isFocus, {
     setTrue: setFocus,
     setFalse: setBlur,
   }] = useBoolean(false)
 
-  const workflowNodesMap = availableNodes.reduce((acc: any, node) => {
+  const workflowNodesMap = availableNodes.reduce<WorkflowNodesMap>((acc, node) => {
     acc[node.id] = {
       title: node.data.title,
       type: node.data.type,
@@ -137,7 +147,7 @@ const FormContent: FC<FormContentProps> = ({
           workflowVariableBlock={{
             show: true,
             variables: availableVars || [],
-            getVarType: getVarType as any,
+            getVarType,
             workflowNodesMap,
           }}
           editable={!readonly}
@@ -145,10 +155,12 @@ const FormContent: FC<FormContentProps> = ({
             ? []
             : [{
                 hotkey: ['mod', '/'],
+                // eslint-disable-next-line react/component-hook-factories, react/no-nested-component-definitions
                 Popup: ({ onClose, onInsert }) => (
                   <AddInputField
                     nodeId={nodeId}
-                    onSave={handleInsertHITLNode(onInsert!)}
+                    unavailableVariableNames={formInputs.map(input => input.output_variable_name)}
+                    onSave={handleInsertHITLNode(onInsert)}
                     onCancel={onClose}
                   />
                 ),

web/app/components/workflow/nodes/human-input/hooks/__tests__/use-form-content.spec.ts

@@ -96,6 +96,26 @@ describe('human-input/use-form-content', () => {
     expect(result.current.editorKey).toBe(1)
   })
 
+  it('should not rename an input to an existing variable name', () => {
+    currentInputs = createPayload({
+      inputs: [
+        createFormInput(),
+        createFormInput({ output_variable_name: 'existing_name' }),
+      ],
+    })
+    const { result } = renderHook(() => useFormContent('human-input-node', currentInputs))
+
+    act(() => {
+      result.current.handleFormInputItemRename(createFormInput({
+        output_variable_name: 'existing_name',
+      }), 'old_name')
+    })
+
+    expect(mockSetInputs).not.toHaveBeenCalled()
+    expect(mockHandleOutVarRenameChange).not.toHaveBeenCalled()
+    expect(result.current.editorKey).toBe(0)
+  })
+
   it('should remove an input placeholder and its form input metadata', () => {
     const { result } = renderHook(() => useFormContent('human-input-node', currentInputs))
 

web/app/components/workflow/nodes/human-input/hooks/use-form-content.ts

@@ -29,6 +29,13 @@ const useFormContent = (id: string, payload: HumanInputNodeType) => {
 
   const handleFormInputItemRename = useCallback((payload: FormInputItem, oldName: string) => {
     const inputs = inputsRef.current
+    if (
+      oldName !== payload.output_variable_name
+      && inputs.inputs.some(item => item.output_variable_name === payload.output_variable_name)
+    ) {
+      return
+    }
+
     const newInputs = produce(inputs, (draft) => {
       draft.form_content = draft.form_content.replaceAll(`{{#$output.${oldName}#}}`, `{{#$output.${payload.output_variable_name}#}}`)
       draft.inputs = draft.inputs.map(item => item.output_variable_name === oldName ? payload : item)

web/i18n/en-US/workflow.json

@@ -654,6 +654,7 @@
   "nodes.humanInput.insertInputField.useConstantInstead": "Use Constant Instead",
   "nodes.humanInput.insertInputField.useVarInstead": "Use Variable Instead",
   "nodes.humanInput.insertInputField.variable": "variable",
+  "nodes.humanInput.insertInputField.variableNameDuplicated": "Variable name already exists",
   "nodes.humanInput.insertInputField.variableNameInvalid": "Variable name can only contain letters, numbers, and underscores, and cannot start with a number",
   "nodes.humanInput.log.backstageInputURL": "Backstage input URL:",
   "nodes.humanInput.log.reason": "Reason:",

web/i18n/zh-Hans/workflow.json

@@ -654,6 +654,7 @@
   "nodes.humanInput.insertInputField.useConstantInstead": "使用常量代替",
   "nodes.humanInput.insertInputField.useVarInstead": "使用变量代替",
   "nodes.humanInput.insertInputField.variable": "变量",
+  "nodes.humanInput.insertInputField.variableNameDuplicated": "变量名已存在",
   "nodes.humanInput.insertInputField.variableNameInvalid": "只能包含字母、数字和下划线，且不能以数字开头",
   "nodes.humanInput.log.backstageInputURL": "表单输入 URL：",
   "nodes.humanInput.log.reason": "原因：",

web/package.json

@@ -1,7 +1,7 @@
 {
   "name": "dify-web",
   "type": "module",
-  "version": "1.14.0",
+  "version": "1.14.1",
   "private": true,
   "imports": {
     "#i18n": {

web/service/knowledge/use-dataset.spec.ts

@@ -0,0 +1,94 @@
+import { useQuery } from '@tanstack/react-query'
+import { get } from '../base'
+import { useDatasetDetail, useDatasetRelatedApps } from './use-dataset'
+
+vi.mock('@tanstack/react-query', () => ({
+  keepPreviousData: Symbol('keepPreviousData'),
+  useInfiniteQuery: vi.fn(),
+  useMutation: vi.fn(),
+  useQuery: vi.fn(),
+  useQueryClient: vi.fn(),
+}))
+
+vi.mock('../base', () => ({
+  get: vi.fn(),
+  post: vi.fn(),
+}))
+
+vi.mock('../use-base', () => ({
+  useInvalid: vi.fn(),
+}))
+
+const mockUseQuery = vi.mocked(useQuery)
+const mockGet = vi.mocked(get)
+
+type QueryOptions = Parameters<typeof useQuery>[0]
+type RetryFn = (failureCount: number, error: unknown) => boolean
+
+const getLastQueryOptions = () => {
+  return mockUseQuery.mock.calls.at(-1)?.[0] as QueryOptions
+}
+
+const getRetryFn = () => {
+  return getLastQueryOptions().retry as RetryFn
+}
+
+describe('knowledge dataset hooks', () => {
+  beforeEach(() => {
+    vi.clearAllMocks()
+    mockUseQuery.mockReturnValue({} as ReturnType<typeof useQuery>)
+  })
+
+  describe('useDatasetDetail', () => {
+    it('should not retry forbidden or missing dataset detail errors', () => {
+      // Arrange & Act
+      useDatasetDetail('dataset-1')
+      const retry = getRetryFn()
+
+      // Assert
+      expect(retry(0, new Response(null, { status: 403 }))).toBe(false)
+      expect(retry(0, new Response(null, { status: 404 }))).toBe(false)
+    })
+
+    it('should retry other dataset detail errors fewer than three times', () => {
+      // Arrange & Act
+      useDatasetDetail('dataset-1')
+      const retry = getRetryFn()
+
+      // Assert
+      expect(retry(2, new Error('temporary failure'))).toBe(true)
+      expect(retry(3, new Error('temporary failure'))).toBe(false)
+    })
+
+    it('should fetch dataset detail without silent mode', () => {
+      // Arrange
+      mockGet.mockResolvedValue({ id: 'dataset-1' })
+
+      // Act
+      useDatasetDetail('dataset-1')
+      const queryFn = getLastQueryOptions().queryFn as () => unknown
+      queryFn()
+
+      // Assert
+      expect(mockGet).toHaveBeenCalledWith('/datasets/dataset-1')
+    })
+  })
+
+  describe('useDatasetRelatedApps', () => {
+    it('should use explicit enabled option when provided', () => {
+      // Arrange & Act
+      useDatasetRelatedApps('dataset-1', { enabled: false })
+
+      // Assert
+      expect(getLastQueryOptions().enabled).toBe(false)
+    })
+
+    it('should enable related apps query when dataset id exists and no option is provided', () => {
+      // Arrange & Act
+      useDatasetRelatedApps('dataset-1')
+
+      // Assert
+      expect(getLastQueryOptions().enabled).toBe(true)
+    })
+  })
+})

web/service/knowledge/use-dataset.ts

@@ -110,13 +110,20 @@ export const useDatasetDetail = (datasetId: string) => {
     queryKey: [...datasetDetailQueryKeyPrefix, datasetId],
     queryFn: () => get<DataSet>(`/datasets/${datasetId}`),
     enabled: !!datasetId,
+    retry: (failureCount, error) => {
+      if (error instanceof Response && [403, 404].includes(error.status))
+        return false
+
+      return failureCount < 3
+    },
   })
 }
 
-export const useDatasetRelatedApps = (datasetId: string) => {
+export const useDatasetRelatedApps = (datasetId: string, options?: { enabled?: boolean }) => {
   return useQuery({
     queryKey: [NAME_SPACE, 'related-apps', datasetId],
     queryFn: () => get<RelatedAppResponse>(`/datasets/${datasetId}/related-apps`),
+    enabled: options?.enabled ?? !!datasetId,
   })
 }