add build_videos file

transcribe_video.py

@@ -8,15 +8,16 @@ from whisper.utils import get_writer
 
 model = whisper.load_model("base")
 
-def extract_audio(video_path: str, audio_temp_path: str):
+def extract_audio(video_path: str):
     # Create a temporary path for a sanitized copy of the video
     sanitized_video_path = video_path.replace(".mp4", "_clean.mp4")
+    audio_temp_path = video_path.replace(".mp4", ".wav")
     
     print("Sanitizing video metadata for MoviePy parser...")
     # -map_chapters -1 removes chapter layouts that break the parser.
     # -sn strips text/subtitle streams that crash MoviePy.
     # -c copy copies video and audio instantly without quality loss.
-    linux.run_command(f"ffmpeg -y -i {video_path} -map_chapters -1 -sn -c copy {sanitized_video_path}")
+    linux.run_command_ffmpeg(f"ffmpeg -y -i {video_path} -map_chapters -1 -sn -c copy {sanitized_video_path}")
     
     print("Extracting uncompressed WAV audio...")
     try:
@@ -34,15 +35,20 @@ def extract_audio(video_path: str, audio_temp_path: str):
             os.remove(sanitized_video_path)
 
 
-def transcribe_to_srt(audio_path: str, output_directory: str, output_filename: str):    
+def transcribe_to_srt(video_path: str):
+
+    extract_audio(video_path)
+
+    transcribe_path = video_path.replace(".mp4", ".str")
+    audio_path = video_path.replace(".mp4", ".wav")
     print("Transcribing audio...")
     result = model.transcribe(audio_path)
     
     print("Creating SRT file...")
-    srt_writer = get_writer("srt", output_directory)
-    srt_writer(result, output_filename, {})
+    srt_writer = get_writer("srt", transcribe_path)
+    srt_writer(result, transcribe_path, {})
     
-    print(f"SRT subtitle file saved in: {output_directory}")
+    print(f"SRT subtitle file saved in: {transcribe_path}")
     if os.path.exists(audio_path):
         os.remove(audio_path)