Merge branches 'main' and 'main' of hf.co:MoYoYoTech/VoiceDialogue

src/voice_dialogue/services/audio/generator.py

@@ -103,4 +103,4 @@ class TTSAudioGenerator(BaseThread, TaskStatusMixin):
         voice_task.tts_generated_sentence_audio = tts_generated_sentence_audio
         voice_task.tts_end_time = time.time()
 
-        self.audio_output_queue.put(voice_task)
+        self.audio_output_queue.put(voice_task.model_copy())

src/voice_dialogue/services/speech/monitor.py

@@ -291,7 +291,7 @@ class SpeechStateMonitor(BaseThread):
                 # 8. 检查是否需要发送语音任务
                 if self._should_send_voice_task(is_audio_sent_for_processing):
                     voice_task = self._create_voice_task(audio_frames)
-                    self.user_voice_queue.put(voice_task)
+                    self.user_voice_queue.put(voice_task.model_copy(deep=True))
 
                     # 更新状态
                     is_audio_sent_for_processing = True

src/voice_dialogue/services/speech/recognizer.py

@@ -38,7 +38,6 @@ class ASRWorker(BaseThread, PerformanceLogMixin):
             except Empty:
                 continue
 
-
             voice_task.language = self.language
             voice_task.whisper_start_time = time.time()
 
@@ -47,16 +46,16 @@ class ASRWorker(BaseThread, PerformanceLogMixin):
             if not transcribed_text.strip():
                 voice_state_manager.reset_task_id()
                 continue
-                
+
             self.log_task_user_question(voice_task)
 
             voice_task.whisper_end_time = time.time()
 
             task_id = voice_task.id
-            cached_user_question = self.cached_user_questions.get(task_id, [])
-            cached_user_question.append(transcribed_text)
 
+            cached_user_question = self.cached_user_questions.get(task_id, [])
             if voice_task.is_over_audio_frames_threshold:
+                cached_user_question.append(transcribed_text)
                 self.cached_user_questions[task_id] = cached_user_question
 
             answer_id = voice_task.answer_id
@@ -72,4 +71,4 @@ class ASRWorker(BaseThread, PerformanceLogMixin):
             voice_task.transcribed_text = ' '.join(cached_user_question) if cached_user_question else transcribed_text
 
             voice_task.user_voice = []
-            self.transcribed_text_queue.put(voice_task)
+            self.transcribed_text_queue.put(voice_task.model_copy())

src/voice_dialogue/services/text/generator.py

@@ -1,5 +1,6 @@
 import copy
 import time
+import unicodedata
 from queue import Queue, Empty
 
 from langchain.memory import ConversationBufferWindowMemory
@@ -99,12 +100,30 @@ class LLMResponseGenerator(BaseThread):
         """重置 chunks 列表"""
         return [remain_content] if remain_content else []
 
+    def _is_punctuation(self, char: str) -> bool:
+        """判断一个字符是否是标点符号"""
+        if not char or len(char) != 1:
+            return False
+        # 检查字符的 Unicode 类别是否为标点符号 (Punctuation)
+        return unicodedata.category(char).startswith('P')
+
     def _process_chunk_content(self, chunk_content: str) -> tuple:
-        """处理 chunk 内容，分离句子结束标记和剩余内容"""
-        if len(chunk_content) > 1:
-            return chunk_content[0], chunk_content[1:]
-        else:
-            return chunk_content, ''
+        """处理 chunk 内容，从右到左找标点符号并分割成三部分"""
+        if not chunk_content:
+            return '', '', ''
+
+        # 从右到左迭代，找到第一个标点符号
+        for i in range(len(chunk_content) - 1, -1, -1):
+            char = chunk_content[i]
+            if self._is_punctuation(char):
+                # 找到标点符号，分割成三部分
+                before_punct = chunk_content[:i]  # 标点之前的部分
+                punct = char  # 标点符号本身
+                after_punct = chunk_content[i + 1:]  # 标点之后的部分
+                return before_punct, punct, after_punct
+
+        # 如果没有找到标点符号，返回整个内容作为第一部分
+        return chunk_content, '', ''
 
     def _process_voice_task(self, voice_task: VoiceTask) -> None:
         """处理单个语音任务"""
@@ -140,8 +159,11 @@ class LLMResponseGenerator(BaseThread):
 
                 chunk_content = f'{chunk.content}'
 
-                sentence_end_mark, remain_content = self._process_chunk_content(chunk_content)
-                chunks.append(sentence_end_mark)
+                before_punct, sentence_end_mark, remain_content = self._process_chunk_content(chunk_content)
+                if before_punct:
+                    chunks.append(before_punct)
+                if sentence_end_mark:
+                    chunks.append(sentence_end_mark)
 
                 sentence = preprocess_sentence_text(chunks)
                 if not sentence:

src/voice_dialogue/services/text/processor.py

@@ -9,9 +9,6 @@ from langchain_core.prompts import (
 from langchain_core.runnables import RunnableWithMessageHistory
 
 from voice_dialogue.utils.logger import logger
-from voice_dialogue.utils.strings import (
-    remove_emojis, convert_comma_separated_numbers, convert_uppercase_words_to_lowercase
-)
 
 
 def create_langchain_chat_llamacpp_instance(
@@ -54,9 +51,6 @@ def warmup_langchain_pipeline(pipeline):
 
 def preprocess_sentence_text(sentences):
     sentence_text = ''.join(sentences)
-    sentence_text = remove_emojis(sentence_text)
-    sentence_text = convert_comma_separated_numbers(sentence_text)
-    sentence_text = convert_uppercase_words_to_lowercase(sentence_text)
     if sentence_text:
         sentence_mark = sentence_text[-1]
         sentence_content = sentence_text[:-1].replace('!', ',').replace('?', ',').replace('.', ',')