Add pause and resume functionality to voice dialogue system

- Implement `pause` and `resume` methods in audio capture service.
- Add `/pause` and `/resume` API endpoints to manage system state.
- Update system schema to include `paused` and `resuming` statuses.

src/voice_dialogue/api/routes/system_routes.py

@@ -205,3 +205,159 @@ async def _start_system_background(request: Request):
     except Exception as e:
         logger.error(f"后台启动系统失败: {e}", exc_info=True)
         _system_status["status"] = "stopped"
+
+
+@router.post("/pause", response_model=SystemResponse, summary="暂停系统")
+async def pause_system(request: Request):
+    """
+    暂停语音对话系统
+    """
+    try:
+        # 检查当前状态是否允许暂停
+        if _system_status["status"] == "stopped":
+            return SystemResponse(
+                success=False,
+                message="系统未启动，无法暂停"
+            )
+        
+        if _system_status["status"] == "paused":
+            return SystemResponse(
+                success=False,
+                message="系统已经暂停"
+            )
+        
+        if _system_status["status"] in ["starting", "stopping"]:
+            return SystemResponse(
+                success=False,
+                message="系统正在启动或停止中，请稍后再试"
+            )
+
+        # 获取服务管理器
+        service_manager = getattr(request.app.state, "service_manager", None)
+        if not service_manager:
+            return SystemResponse(
+                success=False,
+                message="服务管理器未初始化"
+            )
+
+        # 获取音频捕获服务
+        audio_capture_service = service_manager.get_service("audio_capture")
+        if not audio_capture_service:
+            return SystemResponse(
+                success=False,
+                message="音频捕获服务未找到"
+            )
+
+        # 检查服务是否正在运行
+        if not service_manager.is_service_running("audio_capture"):
+            return SystemResponse(
+                success=False,
+                message="音频捕获服务未运行"
+            )
+
+        # 暂停音频捕获服务
+        try:
+            audio_capture_service.pause()
+            logger.info("音频捕获服务已暂停")
+        except Exception as e:
+            logger.error(f"暂停音频捕获服务失败: {e}", exc_info=True)
+            return SystemResponse(
+                success=False,
+                message=f"暂停音频捕获服务失败: {str(e)}"
+            )
+
+        # 更新系统状态
+        _system_status["status"] = "paused"
+
+        return SystemResponse(
+            success=True,
+            message="语音对话系统已成功暂停"
+        )
+
+    except Exception as e:
+        logger.error(f"暂停语音对话系统失败: {e}", exc_info=True)
+        # 恢复状态
+        if _system_status["status"] == "paused":
+            _system_status["status"] = "running"
+        raise HTTPException(status_code=500, detail=f"暂停语音对话系统失败: {str(e)}")
+
+
+@router.post("/resume", response_model=SystemResponse, summary="恢复系统")
+async def resume_system(request: Request):
+    """
+    恢复语音对话系统
+    """
+    try:
+        # 检查当前状态是否允许恢复
+        if _system_status["status"] == "stopped":
+            return SystemResponse(
+                success=False,
+                message="系统未启动，请先启动系统"
+            )
+        
+        if _system_status["status"] == "running":
+            return SystemResponse(
+                success=False,
+                message="系统已经在运行中"
+            )
+        
+        if _system_status["status"] in ["starting", "stopping"]:
+            return SystemResponse(
+                success=False,
+                message="系统正在启动或停止中，请稍后再试"
+            )
+        
+        if _system_status["status"] != "paused":
+            return SystemResponse(
+                success=False,
+                message="只有暂停状态的系统才能恢复"
+            )
+
+        # 获取服务管理器
+        service_manager = getattr(request.app.state, "service_manager", None)
+        if not service_manager:
+            return SystemResponse(
+                success=False,
+                message="服务管理器未初始化"
+            )
+
+        # 获取音频捕获服务
+        audio_capture_service = service_manager.get_service("audio_capture")
+        if not audio_capture_service:
+            return SystemResponse(
+                success=False,
+                message="音频捕获服务未找到"
+            )
+
+        # 检查服务是否存在（可能已被停止）
+        if not service_manager.is_service_running("audio_capture"):
+            return SystemResponse(
+                success=False,
+                message="音频捕获服务未运行，请重新启动系统"
+            )
+
+        # 恢复音频捕获服务
+        try:
+            audio_capture_service.resume()
+            logger.info("音频捕获服务已恢复运行")
+        except Exception as e:
+            logger.error(f"恢复音频捕获服务失败: {e}", exc_info=True)
+            return SystemResponse(
+                success=False,
+                message=f"恢复音频捕获服务失败: {str(e)}"
+            )
+
+        # 更新系统状态为运行中
+        _system_status["status"] = "running"
+
+        return SystemResponse(
+            success=True,
+            message="语音对话系统已成功恢复运行"
+        )
+
+    except Exception as e:
+        logger.error(f"恢复语音对话系统失败: {e}", exc_info=True)
+        # 恢复状态
+        if _system_status["status"] == "running":
+            _system_status["status"] = "paused"
+        raise HTTPException(status_code=500, detail=f"恢复语音对话系统失败: {str(e)}")

src/voice_dialogue/api/schemas/system_schemas.py

@@ -5,7 +5,7 @@ from pydantic import BaseModel, Field
 
 class SystemStatusResponse(BaseModel):
     """系统状态响应"""
-    status: Literal['running', 'stopped', 'starting', 'stopping'] = Field(..., description="系统状态")
+    status: Literal['running', 'stopped', 'paused', 'starting', 'stopping', 'resuming'] = Field(..., description="系统状态")
     uptime: Optional[float] = Field(None, description="运行时间(秒)")
     active_sessions: int = Field(default=0, description="活跃会话数")
     system_running: bool = Field(default=False, description="系统是否运行中")

src/voice_dialogue/services/audio/capture.py

@@ -5,6 +5,7 @@
 
 import ctypes
 import time
+import threading
 
 import numpy as np
 
@@ -23,6 +24,17 @@ class EchoCancellingAudioCapture(BaseThread):
         super().__init__(group, target, name, args, kwargs, daemon=daemon)
 
         self.audio_frames_queue = audio_frames_queue
+        self.pause_event = threading.Event()
+    
+    @property
+    def is_paused(self):
+        return self.pause_event.is_set()
+
+    def pause(self):
+        self.pause_event.set()
+    
+    def resume(self):
+        self.pause_event.clear()
 
     def run(self):
         """主运行循环，持续获取音频数据"""
@@ -45,7 +57,8 @@ class EchoCancellingAudioCapture(BaseThread):
                     audio_data = bytes(data_ptr[: size.value])
                     audio_frame = np.frombuffer(audio_data, dtype=np.int16).astype(np.float32) / np.iinfo(np.int16).max
 
-                    self.audio_frames_queue.put((audio_frame, is_voice_active.value))
+                    if not self.is_paused:
+                        self.audio_frames_queue.put((audio_frame, is_voice_active.value))
 
                     # 使用完数据后释放内存
                     audio_recorder.freeAudioData(data_ptr)