ai回复（切换到qdrant向量库）

fs-ai-api/src/main/java/com/fs/ai/rag/controller/LlmController.java

@@ -0,0 +1,104 @@
+package com.fs.ai.rag.controller;
+
+import cn.hutool.http.HttpRequest;
+import com.alibaba.fastjson.JSON;
+import com.alibaba.fastjson.JSONArray;
+import com.alibaba.fastjson.JSONObject;
+import com.fs.common.core.domain.R;
+import lombok.extern.slf4j.Slf4j;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.web.bind.annotation.PostMapping;
+import org.springframework.web.bind.annotation.RequestBody;
+import org.springframework.web.bind.annotation.RequestMapping;
+import org.springframework.web.bind.annotation.RestController;
+
+import java.util.List;
+import java.util.Map;
+
+@Slf4j
+@RestController
+@RequestMapping("/ai/llm")
+public class LlmController {
+
+    @Value("${ai.rag.llm-url}")
+    private String llmUrl;
+
+    @Value("${ai.rag.llm-api-key:}")
+    private String apiKey;
+
+    @Value("${ai.rag.llm-model}")
+    private String model;
+
+    @PostMapping("/chat")
+    public R chat(@RequestBody Map<String, Object> request) {
+        String modelName = (String) request.getOrDefault("model", model);
+        @SuppressWarnings("unchecked")
+        List<Map<String, String>> messages = (List<Map<String, String>>) request.get("messages");
+        double temperature = request.get("temperature") != null
+                ? ((Number) request.get("temperature")).doubleValue() : 0.7;
+        int maxTokens = request.get("max_tokens") != null
+                ? ((Number) request.get("max_tokens")).intValue() : 2048;
+
+        if (messages == null || messages.isEmpty()) {
+            return R.error().put("msg", "messages 不能为空");
+        }
+
+        // 构建 OpenAI 兼容请求体
+        JSONObject reqBody = new JSONObject();
+        reqBody.put("model", modelName);
+
+        JSONArray msgArray = new JSONArray();
+        for (Map<String, String> msg : messages) {
+            JSONObject msgObj = new JSONObject();
+            msgObj.put("role", msg.get("role"));
+            msgObj.put("content", msg.get("content"));
+            msgArray.add(msgObj);
+        }
+        reqBody.put("messages", msgArray);
+        reqBody.put("temperature", temperature);
+        reqBody.put("max_tokens", maxTokens);
+
+        log.info("LLM 请求: model={}, messages.size={}", modelName, messages.size());
+
+        try {
+            String result = HttpRequest.post(llmUrl)
+                    .header("Authorization", "Bearer " + apiKey)
+                    .header("Content-Type", "application/json")
+                    .body(reqBody.toJSONString())
+                    .timeout(60000)
+                    .execute()
+                    .body();
+
+            JSONObject resp = JSON.parseObject(result);
+
+            // 检查是否有错误
+            if (resp.containsKey("error")) {
+                log.error("LLM 返回错误: {}", resp);
+                return R.error().put("msg", resp.getJSONObject("error").getString("message"));
+            }
+
+            JSONArray choices = resp.getJSONArray("choices");
+            if (choices == null || choices.isEmpty()) {
+                return R.error().put("msg", "LLM 返回为空");
+            }
+
+            String content = choices.getJSONObject(0)
+                    .getJSONObject("message")
+                    .getString("content");
+
+            JSONObject data = new JSONObject();
+            data.put("content", content);
+            data.put("model", resp.getString("model"));
+            data.put("usage", resp.getJSONObject("usage"));
+
+            log.info("LLM 回复成功: content.length={}, usage={}",
+                    content != null ? content.length() : 0, resp.getJSONObject("usage"));
+
+            return R.ok().put("data", data);
+
+        } catch (Exception e) {
+            log.error("LLM 调用异常: {}", e.getMessage(), e);
+            return R.error().put("msg", "LLM 调用失败: " + e.getMessage());
+        }
+    }
+}