20250329 简单的个人rag知识库构建

pom.xml

@@ -36,6 +36,11 @@
             <type>pom</type>
             <scope>import</scope>
         </dependency>
+        <dependency>
+            <groupId>dev.langchain4j</groupId>
+            <artifactId>langchain4j-easy-rag</artifactId>
+            <version>1.0.0-beta1</version>
+        </dependency>
 
         <dependency>
             <groupId>dev.langchain4j</groupId>

src/main/java/com/ai/config/AssistantInit.java

@@ -2,26 +2,40 @@ package com.ai.config;
 
 import com.ai.service.Assist;
 import dev.langchain4j.community.model.dashscope.QwenChatModel;
+import dev.langchain4j.data.document.Document;
+import dev.langchain4j.data.document.loader.FileSystemDocumentLoader;
+import dev.langchain4j.data.segment.TextSegment;
 import dev.langchain4j.memory.chat.MessageWindowChatMemory;
 import dev.langchain4j.model.chat.ChatLanguageModel;
+import dev.langchain4j.rag.content.retriever.EmbeddingStoreContentRetriever;
 import dev.langchain4j.service.AiServices;
+import dev.langchain4j.store.embedding.EmbeddingStoreIngestor;
+import dev.langchain4j.store.embedding.inmemory.InMemoryEmbeddingStore;
 import lombok.RequiredArgsConstructor;
 import org.springframework.context.annotation.Bean;
 import org.springframework.context.annotation.Configuration;
 
+import java.util.List;
+
 @Configuration
 @RequiredArgsConstructor
 public class AssistantInit {
 
-    ChatLanguageModel qwenModel = QwenChatModel.builder()
-            .apiKey("sk-2f703a41fff0488e9b6888013d2ee58a")
-            .modelName("deepseek-v3")
-            .build();
+    // All files in a directory, txt seems to be faster
 
     @Bean
     public Assist init() {
+        ChatLanguageModel qwenModel = QwenChatModel.builder()
+                .apiKey("sk-2f703a41fff0488e9b6888013d2ee58a")
+                .modelName("deepseek-v3")
+                .build();
+        List<Document> documents = FileSystemDocumentLoader.loadDocuments("E:\\ideaProject\\liang-ai");
+        // for simplicity, we will use an in-memory one:
+        InMemoryEmbeddingStore<TextSegment> embeddingStore = new InMemoryEmbeddingStore<>();
+        EmbeddingStoreIngestor.ingest(documents, embeddingStore);
         return AiServices.builder(Assist.class)
                 .chatMemoryProvider(memoryId -> MessageWindowChatMemory.withMaxMessages(10))
-                .chatLanguageModel(qwenModel).build();
+                .chatLanguageModel(qwenModel)
+                .contentRetriever(EmbeddingStoreContentRetriever.from(embeddingStore)).build();
     }
 }

src/main/java/com/ai/http/chat.http

@@ -2,7 +2,7 @@
 GET http://localhost:8080/langchain/chat?input=今天天气如何
 
 ### 测试 LangChainController 的 highlevel chat 接口
-GET http://localhost:8080/langchain/high/chat?input=评价一下路飞
+GET http://localhost:8080/langchain/high/chat?input=请推荐3件DWALK商城的商品
 
 ### 测试 LangChainController 的 highlevel memory chat 接口
 GET http://localhost:8080/langchain/high/memory-chat?memoryId=1&input=你好,我想要买电脑笔记本

个人知识库

@@ -0,0 +1,18 @@
+梁靖林个人博客halo
+地址：halo.liangjl.top
+账号：liangjinglin
+密码: 353912918a
+
+DWALK商城 商品推荐优先级规则
+1.根据本月销量推荐，销量越高，优先级越高，考量占比50%
+2.根据评分，评分越高，优先级越高 考量占比50%
+3.根据评论数，评论数越多，优先级越高 考量占比10%
+4.根据收藏数，收藏数越多，优先级越高 考量占比5%
+
+DWALK商城商品情况
+1.茅台 销量：100 评分：4 评论数：100 收藏数：100
+2.meta50 销量：200 评分：5 评论数：500 收藏数：5000
+3.机械革命 销量：50 评分：4.5 评论数：80 收藏数：100
+4.潮汕牛肉丸 销量：150 评分：3 评论数：100 收藏数：20
+5.风筒 销量：80 评分：3.5 评论数：50 收藏数：10
+6.口红 销量：30 评分：4.5 评论数：200 收藏数：200