"""
Extract unique video URLs from all NitroGen shards.
Reads metadata.json from each tar.gz without fully extracting.
Outputs a JSON file mapping video_id -> {url, source, game, ...}.
"""

import tarfile
import json
import os
import sys
from pathlib import Path
from collections import defaultdict

ACTIONS_DIR = Path(__file__).parent / "actions"
OUTPUT_FILE = Path(__file__).parent / "video_urls.json"


def extract_urls_from_shard(shard_path: Path) -> dict:
    videos = {}
    try:
        with tarfile.open(shard_path, "r:gz") as tar:
            for member in tar:
                if member.name.endswith("metadata.json"):
                    f = tar.extractfile(member)
                    if f is None:
                        continue
                    meta = json.loads(f.read().decode("utf-8"))
                    ov = meta.get("original_video", {})
                    vid = ov.get("video_id")
                    if vid and vid not in videos:
                        videos[vid] = {
                            "url": ov.get("url", ""),
                            "source": ov.get("source", ""),
                            "video_id": vid,
                            "resolution": ov.get("resolution"),
                            "game": meta.get("game", ""),
                        }
    except Exception as e:
        print(f"  [ERROR] {shard_path.name}: {e}", file=sys.stderr)
    return videos


def main():
    shards = sorted(ACTIONS_DIR.glob("SHARD_*.tar.gz"))
    print(f"Found {len(shards)} shards in {ACTIONS_DIR}")

    all_videos = {}
    for i, shard in enumerate(shards):
        print(f"[{i+1:3d}/{len(shards)}] Processing {shard.name} ...")
        videos = extract_urls_from_shard(shard)
        all_videos.update(videos)
        print(f"          Found {len(videos)} videos (total unique: {len(all_videos)})")

    with open(OUTPUT_FILE, "w") as f:
        json.dump(all_videos, f, indent=2)
    print(f"\nDone! {len(all_videos)} unique videos saved to {OUTPUT_FILE}")

    sources = defaultdict(int)
    for v in all_videos.values():
        sources[v["source"]] += 1
    print("Sources breakdown:")
    for src, cnt in sorted(sources.items(), key=lambda x: -x[1]):
        print(f"  {src}: {cnt}")


if __name__ == "__main__":
    main()