personal_development/video_transcription/run_v2.py

#!/usr/bin/env python3
import os
import sys
import subprocess
import shutil

def clear_screen():
    os.system('cls' if os.name == 'nt' else 'clear')

def get_input(prompt, default=None):
    """Helper to get input with a default value."""
    if default:
        user_input = input(f"{prompt} [{default}]: ").strip()
        return user_input if user_input else default
    else:
        return input(f"{prompt}: ").strip()

def get_yes_no(prompt, default="y"):
    """Helper to get boolean input."""
    display_default = "Y/n" if default.lower() in ["y", "yes"] else "y/N"
    choice = get_input(f"{prompt} ({display_default})", default).lower()
    return choice in ["y", "yes", "true", "1"]

def print_header():
    print("==========================================")
    print("   AI Video Transcriber & Translator V2")
    print("      (Powered by Google GenAI SDK)")
    print("==========================================")
    print("")

def main():
    clear_screen()
    print_header()
    
    # Try to load the .env file so the wizard knows what's already configured
    try:
        from dotenv import load_dotenv
        script_dir = os.path.dirname(os.path.abspath(__file__))
        env_path = os.path.abspath(os.path.join(script_dir, '../.env_files/.env.aitranscribe'))
        if os.path.exists(env_path):
            load_dotenv(env_path)
    except ImportError:
        pass

    # 1. Input File/Folder (Multiple)
    input_paths = []
    while True:
        prompt_text = "Enter a path to a video file or folder"
        if input_paths:
            prompt_text += " (or press Enter to finish)"
        
        input_path = get_input(prompt_text)
        
        if not input_path:
            if input_paths:
                break
            else:
                print("Error: You must provide at least one path.")
                continue

        # Clean up input
        input_path = input_path.strip("\'"")
        input_path = input_path.replace(r'\ ', ' ')
        
        # Expand user (~) and resolve absolute path
        input_path = os.path.abspath(os.path.expanduser(input_path))
        
        if os.path.exists(input_path):
            input_paths.append(input_path)
            print(f"Added: {input_path}")
        else:
            print(f"Error: Path '{input_path}' does not exist. Please try again.\n")

    print("\nSelected Inputs:")
    for p in input_paths:
        print(f" - {p}")
    print("")

    # 2. Languages
    source_lang = get_input("Source Language (e.g., French, es)", default="auto")
    target_lang = get_input("Target Language for translation", default="English")
    print("")

    # 3. Model Size
    print("Model Size Options: tiny, base, small, medium, large, auto")
    model_size = get_input("Whisper Model Size", default="auto")
    print("")

    # 4. Features
    do_cleanup = get_yes_no("Cleanup temporary audio files after processing?", default="y")
    do_embed = get_yes_no("Embed subtitles into the video file (Soft Subs)?", default="y")
    do_diarize = get_yes_no("Enable Speaker Diarization (Identify speakers)?", default="n")
    
    do_delete_source = False
    if do_embed:
        print("\n⚠️  WARNING: Using this next option will PERMANENTLY DELETE the original video files.")
        print("   It will only run if the new subtitled video is successfully created.")
        do_delete_source = get_yes_no("Delete original source files after embedding?", default="n")
    
    do_prefer_deep = get_yes_no("Prefer DeepTranslate (Free) over Gemini API?", default="n")
    
    hf_token = None
    if do_diarize:
        if not os.getenv("HF_TOKEN"):
            print("\nSpeaker Diarization requires a HuggingFace Token.")
            hf_token = get_input("Enter your HuggingFace Token (hidden)", default="")
        else:
            print("Using HF_TOKEN from environment.")

    # 5. Build Command
    # Point to v2 main script
    script_dir = os.path.dirname(os.path.abspath(__file__))
    main_script = os.path.join(script_dir, "ai_transcriber_v2", "main.py")

    cmd = [sys.executable, main_script]
    cmd.extend(input_paths)
    
    cmd.extend(["--lang", target_lang])
    cmd.extend(["--model", model_size])
    
    if source_lang != "auto":
        cmd.extend(["--source-lang", source_lang])
    
    if do_cleanup:
        cmd.append("--cleanup")
    
    if do_embed:
        cmd.append("--embed")
        
    if do_delete_source:
        cmd.append("--delete-source")
        
    if do_diarize:
        cmd.append("--diarize")
        if hf_token:
            cmd.extend(["--hf-token", hf_token])

    if do_prefer_deep:
        cmd.append("--prefer-deep")

    # 6. Confirmation and Execution
    clear_screen()
    print_header()
    print("Configuration Complete!")
    print("-" * 30)
    print("Inputs:")
    for p in input_paths:
        print(f" - {p}")
    print(f"Source Lang:  {source_lang}")
    print(f"Target Lang:  {target_lang}")
    print(f"Model:        {model_size}")
    print(f"Cleanup:      {do_cleanup}")
    print(f"Embed Subs:   {do_embed}")
    print(f"Delete Src:   {do_delete_source}")
    print(f"Diarization:  {do_diarize}")
    print(f"Prefer Deep:  {do_prefer_deep}")
    print("-" * 30)
    
    if not get_yes_no("Run this job now?", default="y"):
        print("Aborted.")
        sys.exit(0)

    print("\nStarting Job (V2)...")
    
    try:
        # Pass environment variables including HF_TOKEN if set
        env = os.environ.copy()
        if hf_token:
            env["HF_TOKEN"] = hf_token
            
        subprocess.run(cmd, check=True, env=env)
        print("\n✅ Job Complete!")
    except subprocess.CalledProcessError as e:
        print(f"\n❌ Job Failed with error code {e.returncode}")
    except KeyboardInterrupt:
        print("\nJob interrupted by user.")

if __name__ == "__main__":
    main()