Package sk.iway.iwcm.rag.vectorjpa

Interface EmbeddingChunkRepository

All Superinterfaces:

org.springframework.data.repository.CrudRepository<EmbeddingChunkEntity,Long>, DomainIdRepository<EmbeddingChunkEntity,Long>, org.springframework.data.jpa.repository.JpaRepository<EmbeddingChunkEntity,Long>, org.springframework.data.jpa.repository.JpaSpecificationExecutor<EmbeddingChunkEntity>, org.springframework.data.repository.ListCrudRepository<EmbeddingChunkEntity,Long>, org.springframework.data.repository.ListPagingAndSortingRepository<EmbeddingChunkEntity,Long>, org.springframework.data.repository.PagingAndSortingRepository<EmbeddingChunkEntity,Long>, org.springframework.data.repository.query.QueryByExampleExecutor<EmbeddingChunkEntity>, org.springframework.data.repository.Repository<EmbeddingChunkEntity,Long>

@Repository
public interface EmbeddingChunkRepository
extends DomainIdRepository<EmbeddingChunkEntity,Long>

Spring Data repository for EmbeddingChunkEntity. Provides JPA query methods for managing embedding chunks in the rag_embedding_chunks table. Note: the actual embedding vector storage is managed via native SQL through VectorStore.

Nested Class Summary

Nested classes/interfaces inherited from interface org.springframework.data.jpa.repository.JpaSpecificationExecutor

org.springframework.data.jpa.repository.JpaSpecificationExecutor.SpecificationFluentQuery<T extends Object>

Method Summary

Modifier and Type	Method	Description
void	deleteByEntityTypeAndEntityIdAndDomainId(RagEntityType entityType, Long entityId, Integer domainId)	Deletes all chunks for a source in one storage domain, regardless of provider or model.
void	deleteByEntityTypeAndEntityIdAndEmbeddingProviderAndEmbeddingModelAndDomainId(RagEntityType entityType, Long entityId, String embeddingProvider, String embeddingModel, Integer domainId)	Deletes only the selected provider and model's chunks for a source in one storage domain.
List<EmbeddingChunkEntity>	findByEntityTypeAndDomainIdAndSourcePathStartingWith(RagEntityType entityType, Integer domainId, String sourcePath)
List<EmbeddingChunkEntity>	findByEntityTypeAndEntityIdAndDomainId(RagEntityType entityType, Long entityId, Integer domainId)
List<Long>	findDistinctDocumentEntityIdsWithIncompleteGroupData()	Finds documents whose stored chunks lack one or more group-scope values across all domains.
List<Integer>	findDistinctEntityIdsByEntityTypeAndDomainId(RagEntityType entityType, Integer domainId)	Lists distinct source IDs for document queue operations in one storage domain.
List<Integer>	findDistinctEntityIdsByEntityTypeAndEmbeddingProviderAndEmbeddingModelAndDomainId(RagEntityType entityType, String embeddingProvider, String embeddingModel, Integer domainId)	Lists source IDs with stored chunks from the selected provider and model.
List<RagEntityType>	findDistinctEntityTypes(Integer domainId)	Lists source types with stored chunks in the specified storage domain.
int	updateDocumentGroupData(Long entityId, Integer groupId, Integer rootGroupL1, Integer rootGroupL2, Integer rootGroupL3)	Updates group-scope metadata on all chunks of a document across providers, models, and domains.

Methods inherited from interface org.springframework.data.repository.CrudRepository

count, delete, deleteAll, deleteAll, deleteAllById, deleteById, existsById, findById, save

Methods inherited from interface sk.iway.iwcm.system.datatable.spring.DomainIdRepository

findAllByDomainId, findAllByDomainId, findAllByIdInAndDomainId, findFirstByIdAndDomainId

Methods inherited from interface org.springframework.data.jpa.repository.JpaRepository

deleteAllByIdInBatch, deleteAllInBatch, deleteAllInBatch, deleteInBatch, findAll, findAll, flush, getById, getOne, getReferenceById, saveAllAndFlush, saveAndFlush

Methods inherited from interface org.springframework.data.jpa.repository.JpaSpecificationExecutor

count, count, delete, delete, exists, exists, findAll, findAll, findAll, findAll, findAll, findBy, findBy, findOne, findOne, update

Methods inherited from interface org.springframework.data.repository.ListCrudRepository

findAll, findAllById, saveAll

Methods inherited from interface org.springframework.data.repository.ListPagingAndSortingRepository

findAll

Methods inherited from interface org.springframework.data.repository.PagingAndSortingRepository

findAll

Methods inherited from interface org.springframework.data.repository.query.QueryByExampleExecutor

count, exists, findAll, findBy, findOne

Method Details

findByEntityTypeAndDomainIdAndSourcePathStartingWith

List<EmbeddingChunkEntity> findByEntityTypeAndDomainIdAndSourcePathStartingWith(RagEntityType entityType,
 Integer domainId,
 String sourcePath)

findByEntityTypeAndEntityIdAndDomainId

List<EmbeddingChunkEntity> findByEntityTypeAndEntityIdAndDomainId(RagEntityType entityType,
 Long entityId,
 Integer domainId)

deleteByEntityTypeAndEntityIdAndDomainId

@Transactional
@Modifying
@Query("DELETE FROM EmbeddingChunkEntity c WHERE c.entityType = :entityType AND c.entityId = :entityId AND c.domainId = :domainId")
void deleteByEntityTypeAndEntityIdAndDomainId(@Param("entityType")
 RagEntityType entityType,
 @Param("entityId")
 Long entityId,
 @Param("domainId")
 Integer domainId)

Deletes all chunks for a source in one storage domain, regardless of provider or model.

Parameters:

entityType - source entity type

entityId - source identifier

domainId - storage domain, including zero for shared Markdown

deleteByEntityTypeAndEntityIdAndEmbeddingProviderAndEmbeddingModelAndDomainId

@Transactional
@Modifying
@Query("DELETE FROM EmbeddingChunkEntity c WHERE c.entityType = :entityType AND c.entityId = :entityId AND c.embeddingProvider = :embeddingProvider AND c.embeddingModel = :embeddingModel AND c.domainId = :domainId")
void deleteByEntityTypeAndEntityIdAndEmbeddingProviderAndEmbeddingModelAndDomainId(@Param("entityType")
 RagEntityType entityType,
 @Param("entityId")
 Long entityId,
 @Param("embeddingProvider")
 String embeddingProvider,
 @Param("embeddingModel")
 String embeddingModel,
 @Param("domainId")
 Integer domainId)

Deletes only the selected provider and model's chunks for a source in one storage domain.

Parameters:

entityType - source entity type

entityId - source identifier

embeddingProvider - provider that generated the chunks

embeddingModel - model that generated the chunks

domainId - storage domain, including zero for shared Markdown

findDistinctEntityTypes

@Query("SELECT DISTINCT c.entityType FROM EmbeddingChunkEntity c WHERE c.domainId = :domainId")
List<RagEntityType> findDistinctEntityTypes(@Param("domainId")
 Integer domainId)

Lists source types with stored chunks in the specified storage domain.

Parameters:

domainId - storage domain, including zero for shared Markdown

Returns:

distinct stored entity types

findDistinctEntityIdsByEntityTypeAndDomainId

@Query("SELECT DISTINCT c.entityId FROM EmbeddingChunkEntity c WHERE c.entityType = :entityType AND c.domainId = :domainId")
List<Integer> findDistinctEntityIdsByEntityTypeAndDomainId(RagEntityType entityType,
 Integer domainId)

Lists distinct source IDs for document queue operations in one storage domain. The integer return type is intended for document IDs, not Markdown long identifiers.

Parameters:

entityType - source entity type, normally DOCUMENT

domainId - storage domain

Returns:

distinct source IDs represented as integers

findDistinctEntityIdsByEntityTypeAndEmbeddingProviderAndEmbeddingModelAndDomainId

@Query("SELECT DISTINCT c.entityId FROM EmbeddingChunkEntity c WHERE c.entityType = :entityType AND c.embeddingProvider = :embeddingProvider AND c.embeddingModel = :embeddingModel AND c.domainId = :domainId")
List<Integer> findDistinctEntityIdsByEntityTypeAndEmbeddingProviderAndEmbeddingModelAndDomainId(@Param("entityType")
 RagEntityType entityType,
 @Param("embeddingProvider")
 String embeddingProvider,
 @Param("embeddingModel")
 String embeddingModel,
 @Param("domainId")
 Integer domainId)

Lists source IDs with stored chunks from the selected provider and model. The integer return type is intended for document IDs, not Markdown long identifiers.

Parameters:

entityType - source entity type, normally DOCUMENT

embeddingProvider - provider that generated the chunks

embeddingModel - model that generated the chunks

domainId - storage domain

Returns:

distinct source IDs represented as integers

findDistinctDocumentEntityIdsWithIncompleteGroupData

@Query("SELECT DISTINCT c.entityId FROM EmbeddingChunkEntity c WHERE c.entityType = sk.iway.iwcm.rag.service.RagEntityType.DOCUMENT AND (c.groupId IS NULL OR c.rootGroupL1 IS NULL OR c.rootGroupL2 IS NULL OR c.rootGroupL3 IS NULL)")
List<Long> findDistinctDocumentEntityIdsWithIncompleteGroupData()

Finds documents whose stored chunks lack one or more group-scope values across all domains.

Returns:

distinct document IDs requiring group metadata backfill

updateDocumentGroupData

@Transactional
@Modifying
@Query("UPDATE EmbeddingChunkEntity c SET c.groupId = :groupId, c.rootGroupL1 = :rootGroupL1, c.rootGroupL2 = :rootGroupL2, c.rootGroupL3 = :rootGroupL3 WHERE c.entityType = sk.iway.iwcm.rag.service.RagEntityType.DOCUMENT AND c.entityId = :entityId")
int updateDocumentGroupData(@Param("entityId")
 Long entityId,
 @Param("groupId")
 Integer groupId,
 @Param("rootGroupL1")
 Integer rootGroupL1,
 @Param("rootGroupL2")
 Integer rootGroupL2,
 @Param("rootGroupL3")
 Integer rootGroupL3)

Updates group-scope metadata on all chunks of a document across providers, models, and domains.

Parameters:

entityId - document ID

groupId - immediate document group ID

rootGroupL1 - first-level root group ID

rootGroupL2 - second-level root group ID

rootGroupL3 - third-level root group ID

Returns:

number of chunk rows updated