#!/usr/bin/env python3
"""
Build a JSON index for the NIU skylapse archive.

Run this from the /skylapses directory on the server, for example:
    cd /var/www/html/skylapses
    python3 build_video_index.py

It writes ./video_index.json, which skylapse_search.html uses.
"""

import json
import re
from collections import Counter
from datetime import datetime, timezone
from pathlib import Path

ROOT = Path('.')
OUTPUT = ROOT / 'video_index.json'
VIDEO_EXTS = {'.mov', '.mp4', '.m4v', '.webm'}
DATE_RE = re.compile(r'(?<!\d)(\d{8})(?!\d)')


def camera_from_name(filename: str) -> str:
    stem = Path(filename).stem
    parts = stem.split('_')
    if len(parts) <= 1:
        return 'video'
    return ' '.join(parts[1:]).replace('-', ' ').strip() or 'video'


def main() -> None:
    rows = []
    for path in ROOT.rglob('*'):
        if not path.is_file() or path.name == OUTPUT.name:
            continue
        if path.suffix.lower() not in VIDEO_EXTS:
            continue

        match = DATE_RE.search(path.name)
        if not match:
            continue

        rel = path.relative_to(ROOT).as_posix()
        archive = rel.split('/')[0] if '/' in rel else 'root'
        rows.append(
            {
                'date': match.group(1),
                'filename': path.name,
                'path': rel,
                'archive': archive,
                'camera': camera_from_name(path.name),
                'ext': path.suffix.lower().lstrip('.'),
                'url': rel,
            }
        )

    rows.sort(key=lambda x: (x['date'], x['archive'], x['filename']))

    date_values = [row['date'] for row in rows]
    payload = {
        'meta': {
            'generated_at': datetime.now(timezone.utc).isoformat(),
            'record_count': len(rows),
            'archive_count': len({row['archive'] for row in rows}),
            'camera_count': len({row['camera'] for row in rows}),
            'min_date': min(date_values) if date_values else None,
            'max_date': max(date_values) if date_values else None,
            'archives': dict(sorted(Counter(row['archive'] for row in rows).items())),
        },
        'records': rows,
    }

    OUTPUT.write_text(json.dumps(payload, indent=2), encoding='utf-8')
    print(f'Wrote {len(rows)} records to {OUTPUT}')


if __name__ == '__main__':
    main()