[fix]: typo

transcribe/strategy.py

@@ -34,17 +34,17 @@ class TranscriptResult:
 class TranscriptToken:
     """表示一个转录片段，包含文本和时间信息"""
     text: str  # 转录的文本内容
-    t0: float  # 开始时间（百分之一秒）
-    t1: float  # 结束时间（百分之一秒）
+    t0: int  # 开始时间（百分之一秒）
+    t1: int  # 结束时间（百分之一秒）
 
     def is_punctuation(self):
         """检查文本是否包含标点符号"""
         return REGEX_MARKERS.search(self.text.strip()) is not  None
-    
+
     def is_end(self):
         """检查文本是否为句子结束标记"""
         return SENTENCE_END_PATTERN.search(self.text.strip())  is not  None
-    
+
     def is_pause(self):
         """检查文本是否为暂停标记"""
         return PAUSEE_END_PATTERN.search(self.text.strip()) is not  None
@@ -86,13 +86,13 @@ class TranscriptChunk:
             if not ck.only_punctuation()
         ]
 
-    
+
     def get_split_first_rest(self,  mode: SplitMode):
         chunks = self.split_by(mode)
         fisrt_chunk = chunks[0] if chunks else self
         rest_chunks = chunks[1:] if chunks else None
         return fisrt_chunk, rest_chunks
-    
+
     def puncation_numbers(self) -> int:
         """计算片段中标点符号的数量"""
         return sum(1 for seg in self.items if seg.is_punctuation())
@@ -104,35 +104,35 @@ class TranscriptChunk:
     def join(self) -> str:
         """将片段连接为一个字符串"""
         return self.separator.join(seg.text for seg in self.items)
-    
+
     def compare(self, chunk: Optional['TranscriptChunk'] = None) -> float:
         """比较当前片段与另一个片段的相似度"""
         if not chunk:
             return 0
-    
+
         score =  self._calculate_similarity(self.join(), chunk.join())
         logger.debug(f"Compare: {self.join()} vs {chunk.join()} : {score}")
         return score
-    
+
     def only_punctuation(self)->bool:
         return all(seg.is_punctuation() for seg in self.items)
-    
+
     def has_punctuation(self) -> bool:
         return any(seg.is_punctuation() for seg in self.items)
-    
+
     def get_buffer_index(self) -> int:
         return self.items[-1].buffer_index()
-    
+
     def is_end_sentence(self) ->bool:
         return self.items[-1].is_end()
- 
+
 
 class TranscriptHistory:
     """管理转录片段的历史记录"""
 
     def __init__(self) -> None:
         self.history = collections.deque(maxlen=2)  # 存储最近的两个片段
-    
+
     def add(self, chunk: TranscriptChunk):
         """添加新的片段到历史记录"""
         self.history.appendleft(chunk)
@@ -144,7 +144,7 @@ class TranscriptHistory:
     def lastest_chunk(self):
         """获取最后一个片段"""
         return self.history[-1]
-    
+
     def clear(self):
         self.history.clear()
 
@@ -168,7 +168,7 @@ class TranscriptBuffer:
 
     def get_seg_id(self) -> int:
         return self._current_seg_id
-    
+
     @property
     def current_sentences_length(self) -> int:
         count = 0
@@ -178,7 +178,7 @@ class TranscriptBuffer:
             else:
                 count += len(item)
         return count
-    
+
     def update_pending_text(self, text: str) -> None:
         """更新临时缓冲字符串"""
         self._buffer = text
@@ -192,11 +192,11 @@ class TranscriptBuffer:
     def commit_paragraph(self) -> None:
         """
         提交当前短句为完整段落（如句子结束）
-        
+
         Args:
             end_of_sentence: 是否为句子结尾（如检测到句号）
         """
-        
+
         count = 0
         current_sentences = []
         while len(self._sentences): # and count < 20:
@@ -219,13 +219,13 @@ class TranscriptBuffer:
         output = self.split_and_join(
                     text.replace(
                         self._separator, ""))
-        
+
         logger.debug("==== rebuild string ====")
         logger.debug(text)
         logger.debug(output)
 
         return output
-    
+
     @staticmethod
     def split_and_join(text):
         tokens = []
@@ -264,7 +264,7 @@ class TranscriptBuffer:
             for stable_str in stable_strings:
                 self.update_pending_text(stable_str)
                 self.commit_line()
-                
+
             current_text_len = len(self.current_not_commit_text.split(self._separator)) if self._separator else len(self.current_not_commit_text)
             # current_text_len = len(self.current_not_commit_text.split(self._separator))
             self.update_pending_text(remaining_string)
@@ -279,7 +279,7 @@ class TranscriptBuffer:
             self.update_pending_text(remaining_string)
         return False
 
-       
+
     @property
     def un_commit_paragraph(self) -> str:
         """当前短句组合"""
@@ -298,7 +298,7 @@ class TranscriptBuffer:
     @property
     def current_not_commit_text(self) -> str:
         return self.un_commit_paragraph + self.pending_text
-    
+
 
 
 class TranscriptStabilityAnalyzer:
@@ -311,8 +311,8 @@ class TranscriptStabilityAnalyzer:
     def merge_chunks(self, chunks: List[TranscriptChunk])->str:
         output =  list(r.join() for r in chunks if r)
         return output
-    
-    
+
+
 
     def analysis(self, current: TranscriptChunk, buffer_duration: float) -> Iterator[TranscriptResult]:
         current = TranscriptChunk(items=current, separator=self._separator)
@@ -344,13 +344,13 @@ class TranscriptStabilityAnalyzer:
         # logger.debug("==========================")
 
         if curr_first and prev_first:
-          
+
             core = curr_first.compare(prev_first)
             has_punctuation = curr_first.has_punctuation()
             if core >= 0.8 and has_punctuation:
                 yield from self._yield_commit_results(curr_first, curr_rest, curr_first.is_end_sentence())
                 return
-            
+
         yield TranscriptResult(
             seg_id=self._transcript_buffer.get_seg_id(),
             context=self._transcript_buffer.current_not_commit_text
@@ -377,7 +377,7 @@ class TranscriptStabilityAnalyzer:
         stable_str_list = [stable_chunk.join()] if hasattr(stable_chunk, "join") else self.merge_chunks(stable_chunk)
         remaining_str_list = self.merge_chunks(remaining_chunks)
         frame_cut_index = stable_chunk[-1].get_buffer_index() if isinstance(stable_chunk, list) else stable_chunk.get_buffer_index()
-        
+
         prev_seg_id = self._transcript_buffer.get_seg_id()
         commit_paragraph = self._transcript_buffer.update_and_commit(stable_str_list, remaining_str_list, is_end_sentence)
         logger.debug(f"current buffer: {self._transcript_buffer.__dict__}")
@@ -401,4 +401,4 @@ class TranscriptStabilityAnalyzer:
                 cut_index=frame_cut_index,
                 context=self._transcript_buffer.current_not_commit_text,
             )
-  
+