Obtain media duratin like tree

media_tree.py

#!/usr/bin/env -S uv --quiet run --script
# /// script
# requires-python = ">=3.13"
# dependencies = [
#     "mutagen>=1.47.0",
#     "opencv-python>=4.11.0.86",
# ]
# ///

from pathlib import Path
from mutagen import File as MutagenFile
import cv2
import csv
import argparse

MEDIA_EXTS = {'.mp3', '.mp4', '.wav', '.mkv', '.flac', '.aac', '.mov', '.avi', '.wmv', '.ogg'}
AUDIO_EXTS = {'.mp3', '.wav', '.flac', '.aac', '.ogg'}
VIDEO_EXTS = {'.mp4', '.mkv', '.mov', '.avi', '.wmv'}

def get_media_duration(filepath):
    """Return duration in seconds for a media file using Python libs. Return 0 if not a media file or error."""
    ext = Path(filepath).suffix.lower()
    try:
        if ext in AUDIO_EXTS:
            audio = MutagenFile(filepath)
            if audio is not None and audio.info is not None:
                return audio.info.length
        elif ext in VIDEO_EXTS:
            # Use OpenCV to get video duration
            cap = cv2.VideoCapture(str(filepath))
            if not cap.isOpened():
                return 0.0
            fps = cap.get(cv2.CAP_PROP_FPS)
            frame_count = cap.get(cv2.CAP_PROP_FRAME_COUNT)
            cap.release()
            if fps > 0:
                return frame_count / fps
            else:
                return 0.0
    except Exception:
        return 0.0
    return 0.0

def scan_dir(path: Path, indent: str = "", is_last: bool = True, csv_rows=None, parent_path=""):
    """Recursively scan directory, print tree and duration, return total duration. Optionally collect csv_rows."""
    total_duration = 0.0
    entries = sorted(list(path.iterdir()), key=lambda x: (x.is_file(), x.name.lower()))
    for idx, entry in enumerate(entries):
        last = (idx == len(entries) - 1)
        prefix = indent + ("└── " if last else "├── ")
        rel_path = str(entry.relative_to(parent_path)) if parent_path else str(entry)
        if entry.is_dir():
            sub_duration = scan_dir(entry, indent + ("    " if last else "│   "), last, csv_rows, parent_path or str(path.parent))
            print(f"{prefix}{entry.name}/ [total: {format_duration(sub_duration)}]")
            if csv_rows is not None:
                csv_rows.append([rel_path + '/', 'dir', f"{sub_duration:.2f}"])
            total_duration += sub_duration
        elif entry.suffix.lower() in MEDIA_EXTS:
            duration = get_media_duration(entry)
            print(f"{prefix}{entry.name} [{format_duration(duration)}]")
            if csv_rows is not None:
                csv_rows.append([rel_path, 'file', f"{duration:.2f}"])
            total_duration += duration
        else:
            # Not a media file, skip
            continue
    return total_duration

def format_duration(seconds):
    if seconds < 1e-2:
        return "0s"
    m, s = divmod(int(seconds), 60)
    h, m = divmod(m, 60)
    if h:
        return f"{h}h{m}m{s}s"
    elif m:
        return f"{m}m{s}s"
    else:
        return f"{s}s"

def main():
    parser = argparse.ArgumentParser(description="Scan media files and print tree with durations.")
    parser.add_argument('directory', nargs='?', help='Directory to scan')
    parser.add_argument('--csv', dest='csv', help='Export result to CSV file')
    args = parser.parse_args()
    if not args.directory:
        print("Usage: python media_tree.py <directory> [--csv output.csv]")
        return
    root = Path(args.directory)
    if not root.is_dir():
        print(f"{root} is not a directory.")
        return
    print(f"{root.name}/")
    csv_rows = [] if args.csv else None
    total = scan_dir(root, csv_rows=csv_rows, parent_path=str(root.parent))
    print(f"\nTotal duration: {format_duration(total)}")
    if args.csv:
        with open(args.csv, 'w', newline='', encoding='utf-8') as f:
            writer = csv.writer(f)
            writer.writerow(['path', 'type', 'duration_seconds'])
            writer.writerows(csv_rows)
        print(f"CSV exported to {args.csv}")

if __name__ == "__main__":
    main()