Enable debug mode with global configuration and detailed task logging when active

main.py

@@ -45,6 +45,7 @@ if lib_path.exists() and lib_path.as_posix() not in sys.path:
     sys.path.insert(0, lib_path.as_posix())
 
 from voice_dialogue.core.launcher import launch_system
+from voice_dialogue.core.constants import set_debug_mode
 from voice_dialogue.cli.args import create_argument_parser
 from voice_dialogue.api.server import launch_api_server
 
@@ -62,11 +63,14 @@ def main():
     parser = create_argument_parser()
     args = parser.parse_args()
 
+    set_debug_mode(args.debug)
+
     print(f"""
 {"=" * 80}
 VoiceDialogue - 语音对话系统
 {"=" * 80}
 运行模式: {args.mode.upper()}
+调试模式: {'启用' if args.debug else '禁用'}
 {"=" * 80}
     """)
 

src/voice_dialogue/cli/args.py

@@ -47,6 +47,13 @@ def create_argument_parser():
         help='运行模式: cli=命令行模式, api=API服务器模式 (默认: cli)'
     )
 
+    parser.add_argument(
+        '--debug',
+        action='store_true',
+        default=False,
+        help='启动debug模式'
+    )
+
     # 命令行模式参数
     cli_group = parser.add_argument_group('命令行模式参数')
     cli_group.add_argument(

src/voice_dialogue/core/constants.py

@@ -7,6 +7,20 @@ from voice_dialogue.utils.cache import LRUCacheDict
 from .session_manager import SessionIdManager
 from .state_manager import VoiceStateManager
 
+# ======================= 应用配置常量 =======================
+
+# 全局调试模式状态
+DEBUG_MODE = False
+
+def set_debug_mode(enabled: bool):
+    """设置全局调试模式"""
+    global DEBUG_MODE
+    DEBUG_MODE = enabled
+
+def is_debug_mode() -> bool:
+    """检查是否启用了调试模式"""
+    return DEBUG_MODE
+
 # ======================= 音频配置常量 =======================
 
 # 音频采样率与窗口大小映射配置

src/voice_dialogue/services/audio/player.py

@@ -9,7 +9,7 @@ from playsound import playsound
 from voice_dialogue.core.base import BaseThread
 from voice_dialogue.core.constants import (
     user_still_speaking_event, voice_state_manager, dropped_audio_cache, chat_history_cache,
-    silence_over_threshold_event, session_manager
+    silence_over_threshold_event, session_manager, is_debug_mode
 )
 from voice_dialogue.models.voice_task import VoiceTask, AnswerDisplayMessage
 from voice_dialogue.utils.logger import logger
@@ -70,26 +70,8 @@ class AudioStreamPlayer(BaseThread):
                         )
                     )
 
-                # import librosa
-                # asr_duration = voice_task.whisper_end_time - voice_task.whisper_start_time
-                # llm_duration = voice_task.llm_end_time - voice_task.llm_start_time
-                # tts_duration = voice_task.tts_end_time - voice_task.tts_start_time
-                # audio_data, sample_rate = voice_task.tts_generated_sentence_audio
-                # audio_duration = librosa.get_duration(y=audio_data, sr=sample_rate)
-                #
-                # logger.info(
-                #     f"\n"
-                #     f"┌───────────────────────── 任务信息  ───────────────────────┐\n"
-                #     f"│ 任务ID: {voice_task.id}\n"
-                #     f"├───────────────────────── 性能统计 ────────────────────────┤\n"
-                #     f"│ ASR 耗时: {asr_duration:.2f}s\n"
-                #     f"│ LLM 耗时: {llm_duration:.2f}s\n"
-                #     f"│ TTS 耗时: {tts_duration:.2f}s\n"
-                #     f"│ 音频长度: {audio_duration:.2f}s\n"
-                #     f"├───────────────────────── 生成内容 ────────────────────────┤\n"
-                #     f"│-> {voice_task.answer_sentence}\n"
-                #     f"└──────────────────────────────────────────────────────────┘"
-                # )
+                if is_debug_mode():
+                    self._log_task_info(voice_task)
 
                 self.update_chat_history(voice_task)
 
@@ -105,6 +87,27 @@ class AudioStreamPlayer(BaseThread):
 
                 break
 
+    def _log_task_info(self, voice_task):
+        import librosa
+        asr_duration = voice_task.whisper_end_time - voice_task.whisper_start_time
+        llm_duration = voice_task.llm_end_time - voice_task.llm_start_time
+        tts_duration = voice_task.tts_end_time - voice_task.tts_start_time
+        audio_data, sample_rate = voice_task.tts_generated_sentence_audio
+        audio_duration = librosa.get_duration(y=audio_data, sr=sample_rate)
+        logger.info(
+            f"\n"
+            f"┌───────────────────────── 任务信息  ───────────────────────┐\n"
+            f"│ 任务ID: {voice_task.id}\n"
+            f"├───────────────────────── 性能统计 ────────────────────────┤\n"
+            f"│ ASR 耗时: {asr_duration:.2f}s\n"
+            f"│ LLM 耗时: {llm_duration:.2f}s\n"
+            f"│ TTS 耗时: {tts_duration:.2f}s\n"
+            f"│ 音频长度: {audio_duration:.2f}s\n"
+            f"├───────────────────────── 生成内容 ────────────────────────┤\n"
+            f"│-> {voice_task.answer_sentence}\n"
+            f"└──────────────────────────────────────────────────────────┘"
+        )
+
     def update_chat_history(self, voice_task):
         chat_history = chat_history_cache.get(voice_task.session_id, OrderedDict())
         task_answer_id = voice_task.answer_id