update:密钥自动生成

Author: openrz

README.md

@@ -58,7 +58,23 @@ CREATE TABLE voiceprints (
 ```
 - 复制 `voiceprint.yaml` 为 `data/.voiceprint.yaml`
 
-  4. 启动
+4. 修改配置
+修改`data/.voiceprint.yaml`连接数据库的IP、用户名和密码
+```
+mysql:
+  database: voiceprint_db
+  # 你的mysql所在的局域网ip
+  host: "127.0.0.1"
+  # 密码
+  password: 123456
+  # 端口
+  port: 3306
+  # 用户名
+  user: root 
+```
+
+
+5. 启动
 ```
 python app.py
 ```

app.py

@@ -1,7 +1,9 @@
 import os
+import socket
 import yaml
 import numpy as np
 import torch
+import uuid
 from fastapi import FastAPI, File, UploadFile, Form, Header, HTTPException
 from fastapi.responses import JSONResponse
 from modelscope.pipelines import pipeline
@@ -14,37 +16,58 @@
 import tempfile
 
 # 设置日志
-logging.basicConfig(
-    level=logging.INFO,
-    format="%(asctime)s %(levelname)s %(message)s"
-)
+logging.basicConfig(level=logging.INFO, format="%(asctime)s %(levelname)s %(message)s")
 logger = logging.getLogger(__name__)
 
 # 创建临时目录用于存放上传的音频文件
 TMP_DIR = "tmp"
 os.makedirs(TMP_DIR, exist_ok=True)
 
+
 def load_config():
     """
     加载配置文件，优先读取环境变量（适合Docker部署），否则读取本地yaml。
+    如果authorization不足32位或为空，自动生成UUID并更新配置文件。
     """
     config_path = os.path.join("data", ".voiceprint.yaml")
     if not os.path.exists(config_path):
         logger.error("配置文件 data/.voiceprint.yaml 未找到，请先配置。")
         raise RuntimeError("请先配置 data/.voiceprint.yaml")
+
     with open(config_path, "r", encoding="utf-8") as f:
-        return yaml.safe_load(f)
+        config = yaml.safe_load(f)
+
+    # 检查authorization字段
+    if "server" not in config:
+        config["server"] = {}
+
+    authorization = config["server"].get("authorization", "")
+
+    # 如果authorization为空或长度不足32位，生成新的UUID
+    if not authorization or len(str(authorization)) < 32:
+        new_authorization = str(uuid.uuid4())
+        config["server"]["authorization"] = new_authorization
+
+        # 更新配置文件
+        with open(config_path, "w", encoding="utf-8") as f:
+            yaml.dump(config, f, default_flow_style=False, allow_unicode=True)
+
+        logger.info(f"已自动生成新的authorization密钥: {new_authorization}")
+        logger.info("配置文件已更新，请妥善保管此密钥")
+
+    return config
+
 
 try:
     config = load_config()
-    API_TOKEN = config['server']['authorization']
+    API_TOKEN = config["server"]["authorization"]
 except Exception as e:
     logger.error(f"配置加载失败: {e}")
     raise
 
 # 初始化数据库连接
 try:
-    db = VoiceprintDB(config['mysql'])
+    db = VoiceprintDB(config["mysql"])
     logger.info("数据库连接成功。")
 except Exception as e:
     logger.error(f"数据库连接失败: {e}")
@@ -53,32 +76,48 @@ def load_config():
 # 初始化声纹模型（线程安全，建议单进程部署，或用gunicorn单进程模式）
 try:
     sv_pipeline = pipeline(
-        task=Tasks.speaker_verification, model="iic/speech_campplus_sv_zh-cn_3dspeaker_16k"
+        task=Tasks.speaker_verification,
+        model="iic/speech_campplus_sv_zh-cn_3dspeaker_16k",
     )
     logger.info("声纹模型加载成功。")
 except Exception as e:
     logger.error(f"声纹模型加载失败: {e}")
     raise
 
+
 def _to_numpy(x):
     """
     将torch tensor或其他类型转为numpy数组
     """
     return x.cpu().numpy() if torch.is_tensor(x) else np.asarray(x)
 
+
 app = FastAPI(
-    title="3D-Speaker 声纹识别API",
-    description="基于3D-Speaker的声纹注册与识别服务"
+    title="3D-Speaker 声纹识别API", description="基于3D-Speaker的声纹注册与识别服务"
 )
 
+
 def check_token(token: str = Header(...)):
     """
     校验接口令牌
     """
-    if token != API_TOKEN:
+    if token != "Bearer " + API_TOKEN:
         logger.warning("无效的接口令牌。")
         raise HTTPException(status_code=401, detail="无效的接口令牌")
 
+
+def get_local_ip():
+    try:
+        s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
+        # Connect to Google's DNS servers
+        s.connect(("8.8.8.8", 80))
+        local_ip = s.getsockname()[0]
+        s.close()
+        return local_ip
+    except Exception as e:
+        return "127.0.0.1"
+
+
 def ensure_16k_wav(audio_bytes):
     """
     将任意采样率的wav bytes转为16kHz wav临时文件，返回文件路径
@@ -93,15 +132,21 @@ def ensure_16k_wav(audio_bytes):
         if data.ndim == 1:
             data_rs = librosa.resample(data, orig_sr=sr, target_sr=16000)
         else:
-            data_rs = np.vstack([librosa.resample(data[:, ch], orig_sr=sr, target_sr=16000) for ch in range(data.shape[1])]).T
+            data_rs = np.vstack(
+                [
+                    librosa.resample(data[:, ch], orig_sr=sr, target_sr=16000)
+                    for ch in range(data.shape[1])
+                ]
+            ).T
         sf.write(tmp_path, data_rs, 16000)
     return tmp_path
 
+
 @app.post("/register", summary="声纹注册")
 async def register(
     authorization: str = Header(..., description="接口令牌", alias="authorization"),
     speaker_id: str = Form(..., description="说话人ID"),
-    file: UploadFile = File(..., description="WAV音频文件")
+    file: UploadFile = File(..., description="WAV音频文件"),
 ):
     """
     注册声纹接口
@@ -129,11 +174,12 @@ async def register(
         if audio_path and os.path.exists(audio_path):
             os.remove(audio_path)
 
+
 @app.post("/identify", summary="声纹识别")
 async def identify(
     authorization: str = Header(..., description="接口令牌", alias="authorization"),
     speaker_ids: str = Form(..., description="候选说话人ID，逗号分隔"),
-    file: UploadFile = File(..., description="WAV音频文件")
+    file: UploadFile = File(..., description="WAV音频文件"),
 ):
     """
     声纹识别接口
@@ -160,14 +206,16 @@ async def identify(
             logger.info("未找到候选说话人声纹。")
             return {"speaker_id": "", "score": 0.0}
         similarities = {
-            name: float(np.dot(test_emb, emb) / (np.linalg.norm(test_emb) * np.linalg.norm(emb)))
+            name: float(
+                np.dot(test_emb, emb) / (np.linalg.norm(test_emb) * np.linalg.norm(emb))
+            )
             for name, emb in voiceprints.items()
         }
         match_name = max(similarities, key=similarities.get)
         match_score = similarities[match_name]
         if match_score < 0.2:
             logger.info(f"未识别到说话人，最高分: {match_score}")
-            return 
+            return
         logger.info(f"识别到说话人: {match_name}, 分数: {match_score}")
         return {"speaker_id": match_name, "score": match_score}
     except Exception as e:
@@ -177,26 +225,31 @@ async def identify(
         if audio_path and os.path.exists(audio_path):
             os.remove(audio_path)
 
+
 @app.get("/", include_in_schema=False)
 def root():
     """
     根路径，返回服务运行信息
     """
     return JSONResponse({"msg": "3D-Speaker voiceprint API service running."})
 
+
 if __name__ == "__main__":
     try:
         logger.info(
             f"服务启动中，监听地址: {config['server']['host']}:{config['server']['port']}，"
             f"文档: http://{config['server']['host']}:{config['server']['port']}/docs"
         )
-        print("="*60)
-        print(f"3D-Speaker 声纹API服务已启动，访问: http://{config['server']['host']}:{config['server']['port']}/docs")
-        print("="*60)
+        print("=" * 60)
+        local_ip = get_local_ip()
+        print(
+            f"3D-Speaker 声纹API服务已启动，访问: http://{local_ip}:{config['server']['port']}/docs"
+        )
+        print("=" * 60)
         uvicorn.run(
             "app:app",
-            host=config['server']['host'],
-            port=config['server']['port'],
+            host=config["server"]["host"],
+            port=config["server"]["port"],
         )
     except KeyboardInterrupt:
-        logger.info("收到中断信号，正在退出服务。")
+        logger.info("收到中断信号，正在退出服务。")

db.py

@@ -12,11 +12,14 @@ def __init__(self, config):
 
         :param config: dict，包含数据库连接信息（host, port, user, password, database）
         """
+        # 确保密码字段是字符串类型
+        password = str(config['password']) if config['password'] is not None else ""
+        
         self.conn = pymysql.connect(
             host=config['host'],
             port=config['port'],
             user=config['user'],
-            password=config['password'],
+            password=password,
             database=config['database'],
             charset='utf8mb4',
             autocommit=True

voiceprint.yaml

@@ -2,13 +2,13 @@ server:
   # 服务监听地址，0.0.0.0 表示所有网卡
   host: 0.0.0.0
   # 服务监听端口   
-  port: 8004
-  # 接口访问令牌，调用API时需在header中携带          
-  authorization: "Bearer ac1ab7b959989135c030157ee5b73eb5"
+  port: 8005
+  # 接口访问令牌，会随机生成，如果为空，会自动生成
+  authorization: 
 
 mysql:
   # MySQL数据库主机地址
-  host: "localhost"
+  host: "127.0.0.1"
   # 端口  
   port: 3306
   # 用户名