feat: Weaviate操作向量库功能优化

ruoyi-modules-api/ruoyi-knowledge-api/pom.xml

@@ -48,17 +48,17 @@
         </dependency>
 
         <!-- milvus java sdk -->
-        <dependency>
+<!--        <dependency>-->
-            <groupId>io.milvus</groupId>
+<!--            <groupId>io.milvus</groupId>-->
-            <artifactId>milvus-sdk-java</artifactId>
+<!--            <artifactId>milvus-sdk-java</artifactId>-->
-            <version>2.3.2</version>
+<!--            <version>2.3.2</version>-->
-        </dependency>
+<!--        </dependency>-->
 
-        <dependency>
+<!--        <dependency>-->
-            <groupId>io.weaviate</groupId>
+<!--            <groupId>io.weaviate</groupId>-->
-            <artifactId>client</artifactId>
+<!--            <artifactId>client</artifactId>-->
-            <version>4.0.0</version>
+<!--            <version>4.0.0</version>-->
-        </dependency>
+<!--        </dependency>-->
 
 
         <dependency>
@@ -86,7 +86,12 @@
 
         <dependency>
             <groupId>dev.langchain4j</groupId>
-            <artifactId>langchain4j-open-ai-spring-boot-starter</artifactId>
+            <artifactId>langchain4j-open-ai</artifactId>
+        </dependency>
+
+        <dependency>
+            <groupId>dev.langchain4j</groupId>
+            <artifactId>langchain4j-ollama</artifactId>
         </dependency>
 
     </dependencies>

ruoyi-modules-api/ruoyi-knowledge-api/src/main/java/org/ruoyi/domain/bo/QueryVectorBo.java

@@ -0,0 +1,43 @@
+package org.ruoyi.domain.bo;
+
+
+import lombok.Data;
+
+/**
+ * 查询向量所需参数
+ * @author ageer
+ */
+@Data
+public class QueryVectorBo {
+
+    /**
+     * 查询内容
+     */
+    private String query;
+
+    /**
+     * 知识库kid
+     */
+    private String kid;
+
+    /**
+     * 查询向量返回条数
+     */
+    private Integer maxResults;
+
+    /**
+     * 模型名称
+     */
+    private String modelName;
+
+    /**
+     * 请求key
+     */
+    private String apiKey;
+
+    /**
+     * 请求地址
+     */
+    private String baseUrl;
+
+}

ruoyi-modules-api/ruoyi-knowledge-api/src/main/java/org/ruoyi/domain/bo/StoreEmbeddingBo.java

@@ -0,0 +1,49 @@
+package org.ruoyi.domain.bo;
+
+import lombok.Data;
+
+import java.util.List;
+
+/**
+ * 保存向量所需参数
+ * @author ageer
+ */
+@Data
+public class StoreEmbeddingBo {
+
+    /**
+     * 切分文本块列表
+     */
+    private List<String> chunkList;
+
+    /**
+     * 知识库kid
+     */
+    private String kid;
+
+    /**
+     * 文档id
+     */
+    private String docId;
+
+    /**
+     * 知识块id列表
+     */
+    private List<String> fids;
+
+    /**
+     * 模型名称
+     */
+    private String modelName;
+
+    /**
+     * 请求key
+     */
+    private String apiKey;
+
+    /**
+     * 请求地址
+     */
+    private String baseUrl;
+
+}

ruoyi-modules-api/ruoyi-knowledge-api/src/main/java/org/ruoyi/service/VectorStoreService.java

@@ -1,20 +1,23 @@
 package org.ruoyi.service;
 
+import org.ruoyi.domain.bo.QueryVectorBo;
+import org.ruoyi.domain.bo.StoreEmbeddingBo;
+
 import java.util.List;
 
 /**
- * @author ageer
  * 向量库管理
+ * @author ageer
  */
 public interface VectorStoreService {
 
-    void storeEmbeddings(List<String> chunkList, String kid,String docId,List<String> fids);
+    void storeEmbeddings(StoreEmbeddingBo storeEmbeddingBo);
 
     void removeByDocId(String kid,String docId);
 
     void removeByKid(String kid);
 
-    List<String> getQueryVector(String query, String kid);
+    List<String> getQueryVector(QueryVectorBo queryVectorBo);
 
     void createSchema(String kid);
 

ruoyi-modules-api/ruoyi-knowledge-api/src/main/java/org/ruoyi/service/impl/WeaviateVectorStoreImpl.java

@@ -3,6 +3,7 @@ package org.ruoyi.service.impl;
 import dev.langchain4j.data.embedding.Embedding;
 import dev.langchain4j.data.segment.TextSegment;
 import dev.langchain4j.model.embedding.EmbeddingModel;
+import dev.langchain4j.model.ollama.OllamaEmbeddingModel;
 import dev.langchain4j.model.openai.OpenAiEmbeddingModel;
 import dev.langchain4j.model.output.Response;
 import dev.langchain4j.store.embedding.EmbeddingMatch;
@@ -11,9 +12,12 @@ import dev.langchain4j.store.embedding.EmbeddingStore;
 import dev.langchain4j.store.embedding.filter.Filter;
 import dev.langchain4j.store.embedding.filter.comparison.IsEqualTo;
 import dev.langchain4j.store.embedding.weaviate.WeaviateEmbeddingStore;
+import jakarta.annotation.PostConstruct;
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
 import org.ruoyi.common.core.service.ConfigService;
+import org.ruoyi.domain.bo.QueryVectorBo;
+import org.ruoyi.domain.bo.StoreEmbeddingBo;
 import org.ruoyi.service.VectorStoreService;
 import org.springframework.stereotype.Service;
 
@@ -23,9 +27,11 @@ import java.util.HashMap;
 import java.util.List;
 import java.util.Map;
 
+
+
 /**
+ * Weaviate向量库管理
  * @author ageer
- * Weaviate 向量库管理
  */
 @Service
 @Slf4j
@@ -37,38 +43,7 @@ public class WeaviateVectorStoreImpl implements VectorStoreService {
     private final ConfigService configService;
 
     @Override
-    public List<String> getQueryVector(String query, String kid) {
+    @PostConstruct
-        EmbeddingModel embeddingModel = OpenAiEmbeddingModel.builder()
-                .apiKey("sk-xxx")
-                .baseUrl("https://api.pandarobot.chat/v1/")
-                .modelName(TEXT_EMBEDDING_3_SMALL)
-                .build();
-
-      //  Filter simpleFilter = new IsEqualTo("kid", kid);
-
-     //   createSchema(kid);
-
-        Embedding queryEmbedding = embeddingModel.embed("聊天补全模型").content();
-        EmbeddingSearchRequest embeddingSearchRequest = EmbeddingSearchRequest.builder()
-                .queryEmbedding(queryEmbedding)
-                .maxResults(2)
-                // 添加过滤条件
-             //   .filter(simpleFilter)
-                .build();
-        List<EmbeddingMatch<TextSegment>> matches = embeddingStore.search(embeddingSearchRequest).matches();
-
-
-
-        List<String> results = new ArrayList<>();
-
-        matches.forEach(embeddingMatch -> {
-            results.add(embeddingMatch.embedded().text());
-        });
-
-        return results;
-    }
-
-    @Override
     public void createSchema(String kid) {
         String protocol = configService.getConfigValue("weaviate", "protocol");
         String host = configService.getConfigValue("weaviate", "host");
@@ -84,24 +59,42 @@ public class WeaviateVectorStoreImpl implements VectorStoreService {
     }
 
     @Override
-    public void storeEmbeddings(List<String> chunkList,String kid,String docId,List<String> fids) {
+    public void storeEmbeddings(StoreEmbeddingBo storeEmbeddingBo) {
-        EmbeddingModel embeddingModel = OpenAiEmbeddingModel.builder()
+        EmbeddingModel embeddingModel = getEmbeddingModel(storeEmbeddingBo.getModelName(),
-                .apiKey("sk-xxxx")
+                storeEmbeddingBo.getApiKey(), storeEmbeddingBo.getBaseUrl());
-                .baseUrl("https://api.pandarobot.chat/v1/")
+        for (int i = 0; i < storeEmbeddingBo.getChunkList().size(); i++) {
-                .modelName(TEXT_EMBEDDING_3_SMALL)
-                .build();
-
-        chunkList.forEach(chunk -> {
             Map<String, Object> dataSchema = new HashMap<>();
-            dataSchema.put("kid", kid);
+            dataSchema.put("kid", storeEmbeddingBo.getKid());
-            dataSchema.put("docId", docId);
+            dataSchema.put("docId", storeEmbeddingBo.getKid());
-            dataSchema.put("fid", fids.get(0));
+            dataSchema.put("fid", storeEmbeddingBo.getFids().get(i));
-            Response<Embedding> response = embeddingModel.embed(chunk);
+            Response<Embedding> response = embeddingModel.embed(storeEmbeddingBo.getChunkList().get(i));
             Embedding embedding = response.content();
-            TextSegment segment = TextSegment.from(chunk);
+            TextSegment segment = TextSegment.from(storeEmbeddingBo.getChunkList().get(i));
             segment.metadata().putAll(dataSchema);
             embeddingStore.add(embedding,segment);
+        }
+    }
+
+    @Override
+    public List<String> getQueryVector(QueryVectorBo queryVectorBo) {
+        EmbeddingModel embeddingModel = getEmbeddingModel(queryVectorBo.getModelName(),
+                queryVectorBo.getApiKey(), queryVectorBo.getBaseUrl());
+        Filter simpleFilter = new IsEqualTo("kid", queryVectorBo.getKid());
+        Embedding queryEmbedding = embeddingModel.embed(queryVectorBo.getQuery()).content();
+        EmbeddingSearchRequest embeddingSearchRequest = EmbeddingSearchRequest.builder()
+                .queryEmbedding(queryEmbedding)
+                .maxResults(queryVectorBo.getMaxResults())
+                // 添加过滤条件
+                .filter(simpleFilter)
+                .build();
+        List<EmbeddingMatch<TextSegment>> matches = embeddingStore.search(embeddingSearchRequest).matches();
+
+        List<String> results = new ArrayList<>();
+
+        matches.forEach(embeddingMatch -> {
+            results.add(embeddingMatch.embedded().text());
         });
+        return results;
     }
 
 
@@ -128,4 +121,25 @@ public class WeaviateVectorStoreImpl implements VectorStoreService {
         embeddingStore.removeAll(simpleFilterByAnd);
     }
 
+    /**
+     * 获取向量模型
+     */
+    public EmbeddingModel getEmbeddingModel(String modelName,String apiKey,String baseUrl) {
+        EmbeddingModel embeddingModel = OpenAiEmbeddingModel.builder().build();
+        if(TEXT_EMBEDDING_3_SMALL.toString().equals(modelName)) {
+             embeddingModel = OpenAiEmbeddingModel.builder()
+                    .apiKey(apiKey)
+                    .baseUrl(baseUrl)
+                    .modelName(TEXT_EMBEDDING_3_SMALL)
+                    .build();
+        // TODO 添加枚举
+        }else if("quentinz/bge-large-zh-v1.5".equals(modelName)) {
+            embeddingModel = OllamaEmbeddingModel.builder()
+                    .baseUrl(baseUrl)
+                    .modelName(modelName)
+                    .build();
+        }
+        return embeddingModel;
+    }
+
 }

ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/chat/service/chat/impl/SseServiceImpl.java

@@ -25,6 +25,7 @@ import org.ruoyi.common.core.utils.file.FileUtils;
 import org.ruoyi.common.core.utils.file.MimeTypeUtils;
 import org.ruoyi.common.redis.utils.RedisUtils;
 import org.ruoyi.domain.bo.ChatSessionBo;
+import org.ruoyi.domain.bo.QueryVectorBo;
 import org.ruoyi.domain.vo.ChatModelVo;
 import org.ruoyi.service.VectorStoreService;
 import org.ruoyi.service.IChatModelService;
@@ -166,7 +167,10 @@ public class SseServiceImpl implements ISseService {
         // 获取对话消息列表
         List<Message> messages = chatRequest.getMessages();
         String sysPrompt = chatModelVo.getSystemPrompt();
+
+
         if(StringUtils.isEmpty(sysPrompt)){
+            // TODO 系统默认提示词,后续会增加提示词管理
             sysPrompt ="你是一个由RuoYI-AI开发的人工智能助手，名字叫熊猫助手。你擅长中英文对话，能够理解并处理各种问题，提供安全、有帮助、准确的回答。" +
                     "当前时间："+ DateUtils.getDate()+
                     "#注意：回复之前注意结合上下文和工具返回内容进行回复。";
@@ -180,11 +184,20 @@ public class SseServiceImpl implements ISseService {
         if(StringUtils.isNotEmpty(chatRequest.getKid())){
             List<Message> knMessages = new ArrayList<>();
             String content = messages.get(messages.size() - 1).getContent().toString();
-            List<String> nearestList = vectorStoreService.getQueryVector(content, chatRequest.getKid());
+            QueryVectorBo queryVectorBo = new QueryVectorBo();
+            queryVectorBo.setQuery(content);
+            queryVectorBo.setKid(chatRequest.getKid());
+            queryVectorBo.setApiKey(chatModelVo.getApiKey());
+            queryVectorBo.setBaseUrl(chatModelVo.getApiHost());
+            queryVectorBo.setModelName(chatModelVo.getModelName());
+            // TODO 查询向量返回条数,这里应该查询知识库配置
+            queryVectorBo.setMaxResults(3);
+            List<String> nearestList = vectorStoreService.getQueryVector(queryVectorBo);
             for (String prompt : nearestList) {
                 Message userMessage = Message.builder().content(prompt).role(Message.Role.USER).build();
                 knMessages.add(userMessage);
             }
+            // TODO 提示词,这里应该查询知识库配置
             Message userMessage = Message.builder().content(content + (!nearestList.isEmpty() ? "\n\n注意：回答问题时，须严格根据我给你的系统上下文内容原文进行回答，请不要自己发挥,回答时保持原来文本的段落层级" : "")).role(Message.Role.USER).build();
             knMessages.add(userMessage);
             messages.addAll(knMessages);

ruoyi-modules/ruoyi-chat/src/main/java/org/ruoyi/chat/service/knowledge/KnowledgeInfoServiceImpl.java

@@ -3,6 +3,7 @@ package org.ruoyi.chat.service.knowledge;
 import cn.hutool.core.collection.CollUtil;
 import cn.hutool.core.util.RandomUtil;
 import com.baomidou.mybatisplus.core.conditions.query.LambdaQueryWrapper;
+import com.baomidou.mybatisplus.core.conditions.query.QueryWrapper;
 import com.baomidou.mybatisplus.core.toolkit.Wrappers;
 import com.baomidou.mybatisplus.extension.plugins.pagination.Page;
 import lombok.RequiredArgsConstructor;
@@ -14,15 +15,19 @@ import org.ruoyi.common.core.utils.StringUtils;
 import org.ruoyi.common.satoken.utils.LoginHelper;
 import org.ruoyi.core.page.PageQuery;
 import org.ruoyi.core.page.TableDataInfo;
+import org.ruoyi.domain.ChatModel;
 import org.ruoyi.domain.KnowledgeAttach;
 import org.ruoyi.domain.KnowledgeFragment;
 import org.ruoyi.domain.KnowledgeInfo;
 import org.ruoyi.domain.bo.KnowledgeInfoBo;
 import org.ruoyi.domain.bo.KnowledgeInfoUploadBo;
+import org.ruoyi.domain.bo.StoreEmbeddingBo;
+import org.ruoyi.domain.vo.ChatModelVo;
 import org.ruoyi.domain.vo.KnowledgeInfoVo;
 import org.ruoyi.mapper.KnowledgeAttachMapper;
 import org.ruoyi.mapper.KnowledgeFragmentMapper;
 import org.ruoyi.mapper.KnowledgeInfoMapper;
+import org.ruoyi.service.IChatModelService;
 import org.ruoyi.service.VectorStoreService;
 import org.ruoyi.service.IKnowledgeInfoService;
 import org.slf4j.Logger;
@@ -55,6 +60,8 @@ public class KnowledgeInfoServiceImpl implements IKnowledgeInfoService {
 
     private final KnowledgeAttachMapper attachMapper;
 
+    private final IChatModelService chatModelService;
+
     /**
      * 查询知识库
      */
@@ -219,10 +226,31 @@ public class KnowledgeInfoServiceImpl implements IKnowledgeInfoService {
         knowledgeAttach.setContent(content);
         knowledgeAttach.setCreateTime(new Date());
         attachMapper.insert(knowledgeAttach);
-        vectorStoreService.storeEmbeddings(chunkList,kid,docId,fids);
+
+        // 通过kid查询知识库信息
+        KnowledgeInfoVo knowledgeInfoVo = baseMapper.selectVoOne(Wrappers.<KnowledgeInfo>lambdaQuery()
+                .eq(KnowledgeInfo::getKid, kid));
+
+        // 通过向量模型查询模型信息
+        ChatModelVo chatModelVo = chatModelService.selectModelByName(knowledgeInfoVo.getVectorModel());
+
+        StoreEmbeddingBo storeEmbeddingBo = new StoreEmbeddingBo();
+        storeEmbeddingBo.setKid(kid);
+        storeEmbeddingBo.setDocId(docId);
+        storeEmbeddingBo.setFids(fids);
+        storeEmbeddingBo.setChunkList(chunkList);
+        storeEmbeddingBo.setModelName(knowledgeInfoVo.getVectorModel());
+        storeEmbeddingBo.setApiKey(chatModelVo.getApiKey());
+        storeEmbeddingBo.setBaseUrl(chatModelVo.getApiHost());
+        vectorStoreService.storeEmbeddings(storeEmbeddingBo);
     }
 
 
+    /**
+     * 检查用户是否有删除知识库权限
+     *
+     * @param knowledgeInfoList 知识库列表
+     */
     public void check(List<KnowledgeInfoVo> knowledgeInfoList){
         LoginUser loginUser = LoginHelper.getLoginUser();
         for (KnowledgeInfoVo knowledgeInfoVo : knowledgeInfoList) {