finish function split Audio

fasterWhisperGUIConfig.json

@@ -8,15 +8,15 @@
     "model_param": {
         "localModel": true,
         "onlineModel": false,
-        "model_path": "F:/WhisperModels/faster-whisper/large-v3-float32",
+        "model_path": "",
         "modelName": 0,
         "use_v3_model": true,
         "device": 1,
         "deviceIndex": "0",
         "preciese": 0,
         "thread_num": "4",
         "num_worker": "1",
-        "download_root": "C:/Users/12059/.cache/huggingface/hub",
+        "download_root": "",
         "local_files_only": false
     },
     "vad_param": {
@@ -67,8 +67,8 @@
         "tabShadowEnabled": false,
         "tabMaxWidth": 326,
         "closeDisplayMode": 0,
-        "whisperXMinSpeaker": 0,
-        "whisperXMaxSpeaker": 0,
+        "whisperXMinSpeaker": 3,
+        "whisperXMaxSpeaker": 3,
         "outputFormat": 0,
         "outputEncoding": 1
     }

faster_whisper_GUI/mainWindows.py

@@ -54,6 +54,7 @@
 from .subtitleFileRead import readSRTFileToSegments
 from .config import ENCODING_DICT
 from .util import outputWithDateTime
+from .split_audio import SplitAudioFileWithSpeakersWorker
 
 # =======================================================================================
 # SignalStore
@@ -1285,19 +1286,17 @@ def outputAudioPartWithSpeaker(self):
         output audio part with speaker
         """
         outputWithDateTime("SegmentAudioFileWithSpeaker")
-        for segments, file_path, info in self.current_result:
-            with av.open(file_path) as av_file:
-
-                stream_ = next(s for s in av_file.streams if s.codec_context.type == 'audio')
-                # stream_.
-                audio_channel_num = stream_.channels
-                if audio_channel_num < 2:
-                    print("单声道音频")
-                    split_setore = False
-                else:
-                    print("双声道音频")
-
-
+
+        self.page_output.outputAudioPartWithSpeakerButton.setEnabled(False)
+        output_path = self.page_output.outputGroupWidget.LineEdit_output_dir.text()
+        self.splitAudioFileWithSpeakerWorker = SplitAudioFileWithSpeakersWorker(self.current_result,output_path, self)
+        self.splitAudioFileWithSpeakerWorker.result_signal.connect(self.splitAudioFileWithSpeakerWorkerFinished)
+        self.splitAudioFileWithSpeakerWorker.start()
+
+    def splitAudioFileWithSpeakerWorkerFinished(self):
+        mes = MessageBox("over","ok", self)
+        mes.show()
+        self.page_output.outputAudioPartWithSpeakerButton.setEnabled(True)
 
     def singleAndSlotProcess(self):
         """

faster_whisper_GUI/seg_ment.py

@@ -16,6 +16,7 @@ def __init__(self, segment: Segment=None, start:float = 0, end: float = 0, text:
                     self.words = []
             except Exception:
                 self.words = []
+            self.speaker = None
 
         else:
             self.start = float(start)

faster_whisper_GUI/split_audio.py

@@ -1,6 +1,7 @@
 import os
 from PySide6.QtCore import (QThread, Signal)
 import subprocess
+from .transcribe import secondsToHMS
 
 class SplitAudioFileWithSpeakersWorker(QThread):
     # 定义一个信号，用于在处理完成后发送结果
@@ -12,12 +13,13 @@ def __init__(self, segments_path_info_list:list, output_path, parent=None):
         self.output_path = output_path
 
         # 检查输出目录
-        if not os.path.exists(self.output_path):
+        if output_path and not os.path.exists(self.output_path):
             os.makedirs(self.output_path)
 
     def creatCommandLine(self, start_time, end_time, fileName, output_path, speaker):
 
         output_fileName = self.getOutPutFileName(output_path, start_time, end_time, speaker)
+        # print(output_fileName)
         commandLine = []
         commandLine.append("ffmpeg")
         commandLine.append("-i")
@@ -27,6 +29,7 @@ def creatCommandLine(self, start_time, end_time, fileName, output_path, speaker)
         commandLine.append("-to")
         commandLine.append(end_time)
         commandLine.append(output_fileName)
+        return commandLine
 
     def getOutPutFileName(self, output_path:str, start_time:str, end_time:str, speaker:str):
         return os.path.join(output_path, f"{speaker}_{start_time.replace(':','_')}_{end_time.replace(':','_')}.wav")
@@ -36,28 +39,37 @@ def run(self):
 
         for result in self.segments_path_info_list:
             segments,path,info = result
-            _,file = os.path.split(path)
-            output_path = os.path.join(self.output_path, ".".join(file.split('.')[:-1]))
+            base_path,file = os.path.split(path)
+            print(f"    current task: {file}")
 
+            if not self.output_path:
+                output_path = base_path
+            else:
+                output_path = self.output_path
+            output_path = os.path.join(output_path, ".".join(file.split('.')[:-1]))
+
+            # print(output_path)
             # 检查输出路径
             if not os.path.exists(output_path):
                 os.makedirs(output_path)
 
             for segment in segments:
                 if not segment.speaker: continue
-                start_time = segment.start_time
-                end_time = segment.end_time
+                start_time = secondsToHMS(segment.start)
+                end_time = secondsToHMS(segment.end)
                 speaker = segment.speaker
 
                 commandLine = self.creatCommandLine(start_time.replace(',','.'),end_time.replace(',','.'),path,output_path,speaker)
 
+                # print(commandLine)
                 temp_process = subprocess.Popen(commandLine, shell=False, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, encoding="utf-8", text=True,
                                                 creationflags=subprocess.CREATE_NO_WINDOW)
                 temp_process.wait()
 
         # 完成后发送结果信号
-        result = "结束"
+        result = "over"
         self.result_signal.emit(result)
+        self.stop()
 
     def stop(self):
         self.is_running = False