refactor(api): deduplicate RAG index entities and consolidate import paths (#34690)

api/core/rag/entities/__init__.py

@@ -1,12 +1,20 @@
 from core.rag.entities.citation_metadata import RetrievalSourceMetadata
 from core.rag.entities.context_entities import DocumentContext
+from core.rag.entities.event import DatasourceCompletedEvent, DatasourceErrorEvent, DatasourceProcessingEvent
+from core.rag.entities.index_entities import EconomySetting, EmbeddingSetting, IndexMethod
 from core.rag.entities.metadata_entities import Condition, MetadataFilteringCondition, SupportedComparisonOperator
 from core.rag.entities.processing_entities import ParentMode, PreProcessingRule, Rule, Segmentation
-from core.rag.entities.retrieval_settings import KeywordSetting, VectorSetting
+from core.rag.entities.retrieval_settings import KeywordSetting, VectorSetting, WeightedScoreConfig
 
 __all__ = [
     "Condition",
+    "DatasourceCompletedEvent",
+    "DatasourceErrorEvent",
+    "DatasourceProcessingEvent",
     "DocumentContext",
+    "EconomySetting",
+    "EmbeddingSetting",
+    "IndexMethod",
     "KeywordSetting",
     "MetadataFilteringCondition",
     "ParentMode",
@@ -16,4 +24,5 @@ __all__ = [
     "Segmentation",
     "SupportedComparisonOperator",
     "VectorSetting",
+    "WeightedScoreConfig",
 ]

api/core/rag/entities/index_entities.py

@@ -0,0 +1,30 @@
+from typing import Literal
+
+from pydantic import BaseModel
+
+
+class EmbeddingSetting(BaseModel):
+    """
+    Embedding Setting.
+    """
+
+    embedding_provider_name: str
+    embedding_model_name: str
+
+
+class EconomySetting(BaseModel):
+    """
+    Economy Setting.
+    """
+
+    keyword_number: int
+
+
+class IndexMethod(BaseModel):
+    """
+    Knowledge Index Setting.
+    """
+
+    indexing_technique: Literal["high_quality", "economy"]
+    embedding_setting: EmbeddingSetting
+    economy_setting: EconomySetting

api/core/rag/entities/retrieval_settings.py

@@ -17,3 +17,12 @@ class KeywordSetting(BaseModel):
     """
 
     keyword_weight: float
+
+
+class WeightedScoreConfig(BaseModel):
+    """
+    Weighted score Config.
+    """
+
+    vector_setting: VectorSetting
+    keyword_setting: KeywordSetting

api/core/workflow/nodes/knowledge_index/entities.py

@@ -1,10 +1,10 @@
-from typing import Literal, Union
+from typing import Union
 
 from graphon.entities.base_node_data import BaseNodeData
 from graphon.enums import NodeType
 from pydantic import BaseModel
 
-from core.rag.entities import KeywordSetting, VectorSetting
+from core.rag.entities import WeightedScoreConfig
 from core.rag.index_processor.index_processor_base import SummaryIndexSettingDict
 from core.rag.retrieval.retrieval_methods import RetrievalMethod
 from core.workflow.nodes.knowledge_index import KNOWLEDGE_INDEX_NODE_TYPE
@@ -19,32 +19,6 @@ class RerankingModelConfig(BaseModel):
     reranking_model_name: str
 
 
-class WeightedScoreConfig(BaseModel):
-    """
-    Weighted score Config.
-    """
-
-    vector_setting: VectorSetting
-    keyword_setting: KeywordSetting
-
-
-class EmbeddingSetting(BaseModel):
-    """
-    Embedding Setting.
-    """
-
-    embedding_provider_name: str
-    embedding_model_name: str
-
-
-class EconomySetting(BaseModel):
-    """
-    Economy Setting.
-    """
-
-    keyword_number: int
-
-
 class RetrievalSetting(BaseModel):
     """
     Retrieval Setting.
@@ -60,16 +34,6 @@ class RetrievalSetting(BaseModel):
     weights: WeightedScoreConfig | None = None
 
 
-class IndexMethod(BaseModel):
-    """
-    Knowledge Index Setting.
-    """
-
-    indexing_technique: Literal["high_quality", "economy"]
-    embedding_setting: EmbeddingSetting
-    economy_setting: EconomySetting
-
-
 class FileInfo(BaseModel):
     """
     File Info.

api/core/workflow/nodes/knowledge_retrieval/entities.py

@@ -5,7 +5,7 @@ from graphon.enums import BuiltinNodeTypes, NodeType
 from graphon.nodes.llm.entities import ModelConfig, VisionConfig
 from pydantic import BaseModel, Field
 
-from core.rag.entities import Condition, KeywordSetting, MetadataFilteringCondition, VectorSetting
+from core.rag.entities import Condition, MetadataFilteringCondition, WeightedScoreConfig
 
 __all__ = ["Condition"]
 
@@ -19,15 +19,6 @@ class RerankingModelConfig(BaseModel):
     model: str
 
 
-class WeightedScoreConfig(BaseModel):
-    """
-    Weighted score Config.
-    """
-
-    vector_setting: VectorSetting
-    keyword_setting: KeywordSetting
-
-
 class MultipleRetrievalConfig(BaseModel):
     """
     Multiple Retrieval Config.

api/services/entities/knowledge_entities/rag_pipeline_entities.py

@@ -46,23 +46,6 @@ class WeightedScoreConfig(BaseModel):
     keyword_setting: KeywordSetting | None
 
 
-class EmbeddingSetting(BaseModel):
-    """
-    Embedding Setting.
-    """
-
-    embedding_provider_name: str
-    embedding_model_name: str
-
-
-class EconomySetting(BaseModel):
-    """
-    Economy Setting.
-    """
-
-    keyword_number: int
-
-
 class RetrievalSetting(BaseModel):
     """
     Retrieval Setting.
@@ -78,16 +61,6 @@ class RetrievalSetting(BaseModel):
     weights: WeightedScoreConfig | None = None
 
 
-class IndexMethod(BaseModel):
-    """
-    Knowledge Index Setting.
-    """
-
-    indexing_technique: Literal["high_quality", "economy"]
-    embedding_setting: EmbeddingSetting
-    economy_setting: EconomySetting
-
-
 class KnowledgeConfiguration(BaseModel):
     """
     Knowledge Base Configuration.

api/services/rag_pipeline/rag_pipeline.py

@@ -38,11 +38,7 @@ from core.datasource.online_document.online_document_plugin import OnlineDocumen
 from core.datasource.online_drive.online_drive_plugin import OnlineDriveDatasourcePlugin
 from core.datasource.website_crawl.website_crawl_plugin import WebsiteCrawlDatasourcePlugin
 from core.helper import marketplace
-from core.rag.entities.event import (
-    DatasourceCompletedEvent,
-    DatasourceErrorEvent,
-    DatasourceProcessingEvent,
-)
+from core.rag.entities import DatasourceCompletedEvent, DatasourceErrorEvent, DatasourceProcessingEvent
 from core.repositories.factory import DifyCoreRepositoryFactory, OrderConfig
 from core.repositories.sqlalchemy_workflow_node_execution_repository import SQLAlchemyWorkflowNodeExecutionRepository
 from core.workflow.node_factory import LATEST_VERSION, get_node_type_classes_mapping