模型切换deepseek 使用deepseek的思考

ask-data-ai/ask-data-ai-biz/pom.xml

@@ -79,6 +79,10 @@
         </dependency>
 
         <!-- Spring AI -->
+        <dependency>
+            <groupId>org.springframework.ai</groupId>
+            <artifactId>spring-ai-starter-model-deepseek</artifactId>
+        </dependency>
         <dependency>
             <groupId>org.springframework.ai</groupId>
             <artifactId>spring-ai-starter-model-openai</artifactId>

ask-data-ai/ask-data-ai-biz/src/main/java/com/ask/config/CommonConfiguration.java

@@ -9,6 +9,7 @@ import org.springframework.ai.chat.memory.ChatMemoryRepository;
 import org.springframework.ai.chat.memory.MessageWindowChatMemory;
 import org.springframework.ai.chat.memory.repository.jdbc.JdbcChatMemoryRepository;
 import org.springframework.ai.chat.prompt.PromptTemplate;
+import org.springframework.ai.deepseek.DeepSeekChatModel;
 import org.springframework.ai.openai.OpenAiChatModel;
 import org.springframework.ai.rag.advisor.RetrievalAugmentationAdvisor;
 import org.springframework.ai.rag.generation.augmentation.ContextualQueryAugmenter;
@@ -38,7 +39,14 @@ public class CommonConfiguration {
     }
 
     @Bean
-    public ChatClient chatClient(OpenAiChatModel model) {
+    public ChatClient openAiChatClient(OpenAiChatModel model) {
+        return ChatClient.builder(model)
+                .defaultAdvisors()
+                .build();
+    }
+
+    @Bean
+    public ChatClient deepseekChatClient(DeepSeekChatModel model) {
         return ChatClient.builder(model)
                 .defaultAdvisors()
                 .build();

ask-data-ai/ask-data-ai-biz/src/main/java/com/ask/controller/ChatController.java

@@ -8,29 +8,37 @@ import com.ask.common.core.R;
 import com.ask.service.ChatConversationService;
 import com.ask.service.impl.ChatService;
 import com.ask.service.impl.RagPromptService;
+import com.ask.utils.FluxUtils;
 import io.swagger.v3.oas.annotations.Operation;
 import io.swagger.v3.oas.annotations.Parameter;
 import io.swagger.v3.oas.annotations.tags.Tag;
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
+import org.apache.commons.lang3.StringUtils;
 import org.springframework.ai.chat.client.ChatClient;
+import org.springframework.ai.chat.client.ChatClientResponse;
 import org.springframework.ai.chat.client.advisor.MessageChatMemoryAdvisor;
 import org.springframework.ai.chat.client.advisor.PromptChatMemoryAdvisor;
 import org.springframework.ai.chat.memory.ChatMemory;
 import org.springframework.ai.chat.messages.*;
+import org.springframework.ai.chat.model.ChatResponse;
 import org.springframework.ai.chat.prompt.Prompt;
+import org.springframework.ai.deepseek.DeepSeekAssistantMessage;
 import org.springframework.ai.document.Document;
+import org.springframework.ai.openai.OpenAiChatModel;
 import org.springframework.ai.rag.advisor.RetrievalAugmentationAdvisor;
 import org.springframework.ai.rag.retrieval.search.VectorStoreDocumentRetriever;
 import org.springframework.ai.vectorstore.SearchRequest;
 import org.springframework.ai.vectorstore.VectorStore;
 import org.springframework.ai.vectorstore.filter.Filter;
 import org.springframework.ai.vectorstore.filter.FilterExpressionBuilder;
+import org.springframework.http.MediaType;
 import org.springframework.web.bind.annotation.*;
 import reactor.core.publisher.Flux;
 import reactor.core.publisher.Mono;
 
 import java.util.*;
+import java.util.concurrent.atomic.AtomicBoolean;
 
 @Slf4j
 @RestController
@@ -39,7 +47,9 @@ import java.util.*;
 @Tag(description = "ai", name = "AI对话模块")
 public class ChatController {
 
-    private final ChatClient chatClient;
+    private final ChatClient openAiChatClient;
+
+    private final ChatClient deepseekChatClient;
 
     private final ChatConversationService chatConversationService;
 
@@ -56,6 +66,7 @@ public class ChatController {
     private final RagPromptService ragPromptService;
 
     private final ChatMemory chatMemory;
+    private final OpenAiChatModel openAiChatModel;
 
     /**
      * 获取会话ID
@@ -84,21 +95,24 @@ public class ChatController {
      *
      * @return
      */
-    @Operation(summary = "普通对话", description = "普通对话")
-    @GetMapping(value = "/chat", produces = "text/html;charset=utf-8")
-    public Flux<String> chat(@Parameter(description = "对话内容") @RequestParam String message, @Parameter(description = "会话ID") @RequestParam String conversationId) {
-        // 创建系统消息，告诉大模型只返回工具名和参数
-        Message systemMessage = new SystemMessage("你是一个AI客服助手。请严格按照以下格式回答每个问题：");
-        // 用户消息
-        String question = "请严格按以下格式回答：\n" + "<think>\n" + "[你的逐步推理过程]\n" + "</think>\n" + "<answer>\n" + "[最终答案]\n" + "</answer>\n" + "推理过程不要设计`<think>` 和 `<answer>` \n" + "问题：" + message + "\n";
-        Message userMessage = new UserMessage(question);
-
-        // 创建提示，包含系统消息和用户消息
-        Prompt prompt = new Prompt(Arrays.asList(systemMessage, userMessage));
-        // 使用修改后的提示获取响应
-        return chatClient.prompt(prompt).advisors(messageChatMemoryAdvisor).advisors(a -> a.param(ChatMemory.CONVERSATION_ID, conversationId)).stream().content();
-    }
+    @GetMapping(value = "/chat", produces = MediaType.TEXT_EVENT_STREAM_VALUE)
+    public Flux<String> chat(@RequestParam String message,
+                             @RequestParam String conversationId) {
+
+        Message systemMessage = new SystemMessage("你是一个AI问答助手，请用回答用户问题");
+        Message userMessage = new UserMessage("问题：" + message + "\n回答要求:请使用markdown格式输出");
+        Prompt prompt = new Prompt(List.of(systemMessage, userMessage));
+
+        AtomicBoolean reasoningStarted = new AtomicBoolean(false);
+        AtomicBoolean answerStarted = new AtomicBoolean(false);
+
+        return FluxUtils.wrapDeepSeekStream(deepseekChatClient.prompt(prompt)
+                .advisors(messageChatMemoryAdvisor)
+                .advisors(a -> a.param(ChatMemory.CONVERSATION_ID, conversationId))
+                .stream()
+                .chatResponse());
 
+    }
     /**
      * 知识库对话
      * <p>
@@ -107,7 +121,7 @@ public class ChatController {
      * @return
      */
     @Operation(summary = "知识库对话", description = "知识库对话")
-    @GetMapping(value = "/rag/chat", produces = "text/html;charset=utf-8")
+    @GetMapping(value = "/rag/chat", produces = MediaType.TEXT_EVENT_STREAM_VALUE)
     public Flux<String> ragChat(@RequestParam @Parameter(description = "对话内容") String message, @RequestParam @Parameter(description = "会话ID") String conversationId) {
 
         //获取对话历史
@@ -125,21 +139,15 @@ public class ChatController {
         String context = chatService.convertDocumentsToString(documents);
         //创建提示词
         String userPrompt = ragPromptService.createRagPrompt(message, context, historyMemory);
-        StringBuilder contentBuilder = new StringBuilder();
 
-        return chatClient.prompt()
-                .user(userPrompt)
-                .system("你是一个智能助手，基于以下上下文和历史对话回答问题,请用简洁的语言回答问题，并确保答案准确,要求" +
-                        "1.以 Markdown 格式输出" +
-                        "2.请务必将你的思考过程放在 <think></think> 标签内" +
-                        "3.请务必将生成最终答案放在 <answer></answer> 标签内")
-                .stream()
-                .content()
-                .concatWith(Mono.just(reference))
-                .doOnNext(chunk -> {
-                    // 实时收集每个流片段
-                    contentBuilder.append(chunk);
-                })
+        StringBuilder contentBuilder = new StringBuilder();
+        return FluxUtils.wrapDeepSeekStream(deepseekChatClient.prompt()
+                        .user(userPrompt)
+                        .system("你是一个智能助手，基于以下上下文和历史对话回答问题,请用简洁的语言回答问题，并确保答案准确,要求" +
+                                "1.以 Markdown 格式输出" )
+                        .stream()
+                        .chatResponse(),contentBuilder)
+                .concatWith(Flux.just(reference))
                 .doOnComplete(() -> {
                     // 流结束时获取完整内容
                     String fullResponse = contentBuilder.toString();

ask-data-ai/ask-data-ai-biz/src/main/java/com/ask/utils/FluxUtils.java

+package com.ask.utils;
+
+import com.baomidou.mybatisplus.core.toolkit.StringUtils;
+import org.springframework.ai.chat.model.ChatResponse;
+import org.springframework.ai.deepseek.DeepSeekAssistantMessage;
+import reactor.core.publisher.Flux;
+
+import java.util.List;
+import java.util.concurrent.atomic.AtomicBoolean;
+
+public class FluxUtils {
+
+    /**
+     * 将 DeepSeek 的 Flux<ChatResponse> 转换成带 <think>/<answer> 的 Flux<String>
+     * @param upstream 原始 SSE 流
+     * @return 带标签的逐块流
+     */
+    public static Flux<String> wrapDeepSeekStream(Flux<ChatResponse> upstream) {
+        AtomicBoolean reasoningStarted = new AtomicBoolean(false);
+        AtomicBoolean answerStarted    = new AtomicBoolean(false);
+
+        return upstream
+                .flatMapIterable(resp -> {
+                    DeepSeekAssistantMessage msg =
+                            (DeepSeekAssistantMessage) resp.getResult().getOutput();
+
+                    StringBuilder sb = new StringBuilder();
+
+                    // 推理阶段：第一次出现推理内容时输出 <think>
+                    if (StringUtils.isNotBlank(msg.getReasoningContent())) {
+                        if (reasoningStarted.compareAndSet(false, true)) {
+                            sb.append("<think>");
+                        }
+                        sb.append(msg.getReasoningContent());
+                    }
+
+                    // 回答阶段：第一次出现答案时输出 </think><answer>
+                    if (StringUtils.isNotBlank(msg.getText())) {
+                        if (answerStarted.compareAndSet(false, true)) {
+                            sb.append("</think><answer>");
+                        }
+                        sb.append(msg.getText());
+                    }
+
+                    return List.of(sb.toString());
+                })
+                .concatWith(Flux.just("</answer>")); // 末尾补一次关闭标签
+    }
+
+    public static Flux<String> wrapDeepSeekStream(Flux<ChatResponse> upstream,StringBuilder
+                                                  stringBuilder) {
+        AtomicBoolean reasoningStarted = new AtomicBoolean(false);
+        AtomicBoolean answerStarted    = new AtomicBoolean(false);
+
+        return upstream
+                .flatMapIterable(resp -> {
+                    DeepSeekAssistantMessage msg =
+                            (DeepSeekAssistantMessage) resp.getResult().getOutput();
+
+                    StringBuilder sb = new StringBuilder();
+
+                    // 推理阶段：第一次出现推理内容时输出 <think>
+                    if (StringUtils.isNotBlank(msg.getReasoningContent())) {
+                        if (reasoningStarted.compareAndSet(false, true)) {
+                            sb.append("<think>");
+                        }
+                        sb.append(msg.getReasoningContent());
+                    }
+
+                    // 回答阶段：第一次出现答案时输出 </think><answer>
+                    if (StringUtils.isNotBlank(msg.getText())) {
+                        stringBuilder.append(msg.getText());
+                        if (answerStarted.compareAndSet(false, true)) {
+                            sb.append("</think><answer>");
+                        }
+                        sb.append(msg.getText());
+                    }
+
+                    return List.of(sb.toString());
+                })
+                .concatWith(Flux.just("</answer>")); // 末尾补一次关闭标签
+    }
+}

ask-data-ai/ask-data-ai-boot/src/main/resources/application.yml

@@ -37,12 +37,19 @@ spring:
       api-key: sk-ae96ff281ff644c992843c64a711a950
       chat:
         options:
-          model: qwen-plus
+          model: deepseek-r1
       embedding:
         base-url: https://dashscope.aliyuncs.com/compatible-mode
         api-key: sk-ae96ff281ff644c992843c64a711a950
         options:
           model: text-embedding-v4
+    deepseek:
+      base-url: https://dashscope.aliyuncs.com/compatible-mode/v1
+      api-key: sk-ae96ff281ff644c992843c64a711a950
+      chat:
+        enabled: true
+        options:
+          model: deepseek-r1
 
 mybatis-plus:
   mapper-locations: classpath*:/mapper/*Mapper.xml # mapper文件位置