Package com.bytedesk.kbase.llm_webpage.vector

Class WebpageVectorService

java.lang.Object

com.bytedesk.kbase.llm_webpage.vector.WebpageVectorService

@Service
@ConditionalOnBean(org.springframework.ai.vectorstore.elasticsearch.ElasticsearchVectorStore.class)
public class WebpageVectorService
extends Object

网页向量检索服务 用于处理网页的向量存储和相似度搜索

Author:

jackning

Field Summary

Fields

Modifier and Type	Field	Description
private final org.springframework.ai.vectorstore.elasticsearch.ElasticsearchVectorStore	vectorStore
private final WebpageRestService	webpageRestService

Constructor Summary

Constructors

Constructor	Description
WebpageVectorService()

Method Summary

Modifier and Type	Method	Description
private WebpageVector	createSimpleWebpageVectorFromDocument(org.springframework.ai.document.Document doc)	从文档创建简化版的WebpageVector对象 用于在找不到对应的网页实体时提供基础信息
Boolean	deleteWebpageVector(WebpageEntity webpage)	从向量存储中删除网页向量文档 修改为不更新实体的方式，只进行向量删除操作，避免实体并发修改冲突
void	indexWebpageVector(WebpageEntity webpage)	将网页内容添加到向量存储中
List<WebpageVectorSearchResult>	searchWebpageVector(String query, String kbUid, String categoryUid, String orgUid, int limit)	在向量存储中进行语义搜索
void	updateAllVectorIndex(WebpageRequest request)	更新所有网页的向量索引
void	updateVectorIndex(WebpageRequest request)	更新网页的向量索引

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Details

vectorStore

private final org.springframework.ai.vectorstore.elasticsearch.ElasticsearchVectorStore vectorStore

webpageRestService

private final WebpageRestService webpageRestService

Constructor Details

WebpageVectorService

public WebpageVectorService()

Method Details

indexWebpageVector

@Transactional
public void indexWebpageVector(WebpageEntity webpage)

将网页内容添加到向量存储中

Parameters:

webpage - 网页实体

updateVectorIndex

public void updateVectorIndex(WebpageRequest request)

更新网页的向量索引

Parameters:

request - 网页请求对象

updateAllVectorIndex

public void updateAllVectorIndex(WebpageRequest request)

更新所有网页的向量索引

Parameters:

request - 网页请求对象，包含知识库ID

deleteWebpageVector

@Transactional(readOnly=true)
public Boolean deleteWebpageVector(WebpageEntity webpage)

从向量存储中删除网页向量文档 修改为不更新实体的方式，只进行向量删除操作，避免实体并发修改冲突

Parameters:

webpage - 网页实体

Returns:

删除成功返回true，否则返回false

searchWebpageVector

public List<WebpageVectorSearchResult> searchWebpageVector(String query,
 String kbUid,
 String categoryUid,
 String orgUid,
 int limit)

在向量存储中进行语义搜索

Parameters:

query - 搜索关键词

kbUid - 知识库UID（可选）

categoryUid - 分类UID（可选）

orgUid - 组织UID（可选）

limit - 返回结果数量限制

Returns:

相似度搜索结果列表

createSimpleWebpageVectorFromDocument

private WebpageVector createSimpleWebpageVectorFromDocument(org.springframework.ai.document.Document doc)

从文档创建简化版的WebpageVector对象 用于在找不到对应的网页实体时提供基础信息

Parameters:

doc - 文档对象

Returns:

简化的WebpageVector对象