update media process

course.py

@@ -10,12 +10,10 @@ from threading import Thread
 
 import requests
 
-from fileconvert import mp4_to_mp3, mp4_to_wav
 from headers import headers
 
 from logging_config import setup_logging
-# from video_voice_process import process_audio_file
-from sense_voice_process import process_audio_file
+from transcribe_media import convert_media
 
 logger = setup_logging()
 
@@ -225,14 +223,7 @@ def download_course_contents(course_ids, course_ids_dict):
                         # 构建完整的 MP4 文件路径
                         mp4_file = os.path.join(root, file)
         if (not exist_md_file) and mp4_file is not None:
-            mp4_to_mp3(mp4_file)
-            # 调用 mp4_to_wav 函数进行转换
-            wav_file = mp4_to_wav(mp4_file)
-            if wav_file is not None:
-                try:
-                    process_audio_file(wav_file)
-                except:
-                    print('process_audio_file fail')
+            convert_media(mp4_file)
 
 
 def get_course():

markdown_generator.py

@@ -12,6 +12,7 @@ from gradio_client import Client, handle_file
 import json
 from logging_config import setup_logging
 from sense_voice_process import short_audio_process
+from transcribe_media import convert_media
 
 # 读取配置文件
 config = configparser.ConfigParser()
@@ -84,7 +85,7 @@ def voice2txt(voice_path):
     #     api_name="/model_inference"
     # )
     # logger.info(text)
-    text = short_audio_process(voice_path)
+    text = convert_media(voice_path)
     return text
 
 

transcribe_media.py

@@ -48,7 +48,7 @@ def transcribe_audio_funasr(audio_path, device="cuda:0"):
     return text
 
 
-def main(file_path):
+def convert_media(file_path):
     try:
         audio_file = extract_or_convert_audio(file_path)
         transcript = transcribe_audio_funasr(audio_file)
@@ -56,13 +56,15 @@ def main(file_path):
         print(transcript)
 
         # ✅ Save transcript to disk
-        output_path = os.path.splitext(file_path)[0] + "_transcript.txt"
+        output_path = os.path.splitext(file_path)[0] + "_transcript.md"
         with open(output_path, "w", encoding="utf-8") as f:
             f.write(transcript)
 
         print(f"✅ Transcript saved to: {output_path}")
+        return transcript
     finally:
         if os.path.exists("processed_audio.wav"):
             os.remove("processed_audio.wav")
 
-main("./course/676/mp4/20250413142836-第36期茶话会-视频-1.mp4")
+if __name__ == '__main__':
+    convert_media("./course/676/mp4/20250413142836-第36期茶话会-视频-1.mp4")