AI 面试业务逻辑初始化

vetti-admin/src/main/java/com/vetti/socket/ChatWebSocketHandler.java

@@ -4,52 +4,34 @@ import com.vetti.common.ai.elevenLabs.ElevenLabsClient;
 import com.vetti.common.ai.gpt.ChatGPTClient;
 import com.vetti.common.ai.whisper.WhisperClient;
 import com.vetti.common.config.RuoYiConfig;
+import com.vetti.common.utils.spring.SpringUtils;
 import lombok.extern.slf4j.Slf4j;
-import org.springframework.beans.factory.annotation.Autowired;
 import org.springframework.stereotype.Component;
 
-import javax.sound.sampled.AudioFileFormat;
-import javax.sound.sampled.AudioFormat;
-import javax.sound.sampled.AudioInputStream;
-import javax.sound.sampled.AudioSystem;
 import javax.websocket.*;
 import javax.websocket.server.PathParam;
 import javax.websocket.server.ServerEndpoint;
-import java.io.*;
+import java.io.File;
+import java.io.FileOutputStream;
+import java.io.IOException;
+import java.io.RandomAccessFile;
 import java.nio.ByteBuffer;
 import java.nio.channels.FileChannel;
 
 /**
  * 语音面试 web处理器
- * 
  */
 @Slf4j
 @ServerEndpoint("/voice-websocket/{clientId}")
 @Component
 public class ChatWebSocketHandler {
 
-    // 音频参数配置 - 根据实际发送的音频流参数调整
-    private static final float SAMPLE_RATE = 16000.0f;    // 采样率
-    private static final int SAMPLE_SIZE_IN_BITS = 16;    // 采样位数
-    private static final int CHANNELS = 1;                // 声道数 (1=单声道, 2=立体声)
-    private static final boolean SIGNED = true;           // 是否有符号
-    private static final boolean BIG_ENDIAN = false;      // 字节序
-
     // 语音文件保存目录
     private static final String VOICE_STORAGE_DIR = "/voice_files/";
 
     // 语音结果文件保存目录
     private static final String VOICE_STORAGE_RESULT_DIR = "/voice_result_files/";
 
-    @Autowired
-    private ElevenLabsClient elevenLabsClient;
-
-    @Autowired
-    private ChatGPTClient chatGPTClient;
-
-    @Autowired
-    private WhisperClient whisperClient;
-
 
     public ChatWebSocketHandler() {
         // 初始化存储目录
@@ -91,30 +73,36 @@ public class ChatWebSocketHandler {
         try {
             log.info("客户端ID为:{}", clientId);
             // 处理二进制流数据
-            byte[] data = new byte[byteBuffer.remaining()];
+            byte[] bytes = new byte[byteBuffer.remaining()];
+            //从缓冲区中读取数据并存储到指定的字节数组中
+            byteBuffer.get(bytes);
             // 生成唯一文件名
-            String fileName = clientId + "_" + System.currentTimeMillis() + ".mp3";
+            String fileName = clientId + "_" + System.currentTimeMillis() + ".webm";
             String pathUrl = RuoYiConfig.getProfile()+VOICE_STORAGE_DIR + fileName;
+//            String pathUrl = "/Users/wangxiangshun/Desktop/0.8733346782733291.webm";
             log.info("文件路径为:{}", pathUrl);
-            saveAsWebM(data,pathUrl);
-//            //拿到文件进行文字转换
-//            String resultText = whisperClient.handleVoiceToText(pathUrl);
-//            //把提问的文字发送给CPT
-//            String resultMsg = chatGPTClient.handleAiChat(resultText);
-//            //把结果文字转成语音文件
-//            //生成文件
-//            // 生成唯一文件名
-//            String resultFileName = clientId + "_" + System.currentTimeMillis() + ".webm";
-//            String resultPathUrl = RuoYiConfig.getProfile()+VOICE_STORAGE_DIR + resultFileName;
-//            elevenLabsClient.handleTextToVoice(resultMsg,resultPathUrl);
+            saveAsWebM(bytes, pathUrl);
+            //拿到文件进行文字转换
+            WhisperClient whisperClient = SpringUtils.getBean(WhisperClient.class);
+            String resultText = whisperClient.handleVoiceToText(pathUrl);
+            //把提问的文字发送给CPT
+            ChatGPTClient chatGPTClient = SpringUtils.getBean(ChatGPTClient.class);
+            String resultMsg = chatGPTClient.handleAiChat(resultText);
+            //把结果文字转成语音文件
+            //生成文件
+            // 生成唯一文件名
+            String resultFileName = clientId + "_" + System.currentTimeMillis() + ".webm";
+            String resultPathUrl = RuoYiConfig.getProfile() + VOICE_STORAGE_RESULT_DIR + resultFileName;
+            ElevenLabsClient elevenLabsClient = SpringUtils.getBean(ElevenLabsClient.class);
+            elevenLabsClient.handleTextToVoice(resultMsg, resultPathUrl);
             //把语音文件转换成流,发送给前端
-            System.out.println("接收到二进制数据，长度: " + data.length + " bytes");
+            System.out.println("接收到二进制数据，长度: " + bytes.length + " bytes");
             try {
                 //文件转换成文件流
-//                ByteBuffer outByteBuffer = convertFileToByteBuffer(resultPathUrl);
-//                session.getBasicRemote().sendBinary(outByteBuffer);
+                ByteBuffer outByteBuffer = convertFileToByteBuffer(resultPathUrl);
+                session.getBasicRemote().sendBinary(outByteBuffer);
                 // 发送响应确认
-                session.getBasicRemote().sendText("已收到二进制数据，长度: " + data.length);
+                session.getBasicRemote().sendText("已收到二进制数据，长度: " + bytes.length);
             } catch (IOException e) {
                 e.printStackTrace();
             }
@@ -138,6 +126,7 @@ public class ChatWebSocketHandler {
 
     /**
      * 将字节数组保存为WebM文件
+     *
      * @param byteData 包含WebM数据的字节数组
      * @param filePath 目标文件路径
      * @return 操作是否成功
@@ -181,39 +170,9 @@ public class ChatWebSocketHandler {
         return false;
     }
 
-    /**
-     * 将接收的音频数据保存为WAV文件
-     */
-    private void saveAudioToFile(byte[] audioData,String pathUrl) {
-        if (audioData == null || audioData.length == 0) {
-            System.out.println("没有接收到音频数据，不保存文件");
-            return;
-        }
-        try {
-            // 合并所有音频数据
-            // 创建音频格式
-            AudioFormat format = new AudioFormat(
-                    SAMPLE_RATE,
-                    SAMPLE_SIZE_IN_BITS,
-                    CHANNELS,
-                    SIGNED,
-                    BIG_ENDIAN
-            );
-            // 创建音频输入流
-            ByteArrayInputStream bais = new ByteArrayInputStream(audioData);
-            AudioInputStream ais = new AudioInputStream(bais, format, audioData.length / format.getFrameSize());
-            // 保存为WAV文件
-            File outputFile = new File(pathUrl);
-            AudioSystem.write(ais, AudioFileFormat.Type.WAVE, outputFile);
-            System.out.println("音频文件保存成功: " + outputFile.getAbsolutePath());
-        } catch (IOException e) {
-            System.err.println("保存音频文件失败: " + e.getMessage());
-            e.printStackTrace();
-        }
-    }
-
     /**
      * File 转换成 ByteBuffer
+     *
      * @param fileUrl 文件路径
      * @return
      */

vetti-admin/src/main/resources/application-druid.yml

@@ -152,7 +152,7 @@ whisper:
     apiUrl: https://api.openai.com/v1/audio/transcriptions
     model: whisper-1
     apiKey: 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
-    language: zh
+    language: en
 
 # AI 聊天
 chatGpt:

vetti-admin/target/classes/application-druid.yml

@@ -152,7 +152,7 @@ whisper:
     apiUrl: https://api.openai.com/v1/audio/transcriptions
     model: whisper-1
     apiKey: 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
-    language: zh
+    language: en
 
 # AI 聊天
 chatGpt: