#!/usr/bin/env python3
"""Extract frames from real drone/robot footage at a fixed sample rate.

Writes {out}/frames/{video}_{idx:05d}.jpg + a manifest.jsonl (one entry per
frame) matching the iNat downloader schema so the SAME Gemini labeler can box
them as ground truth.
"""
import argparse
import json
from pathlib import Path

import cv2


def main():
    ap = argparse.ArgumentParser()
    ap.add_argument("--video", required=True, help="Path to a video file")
    ap.add_argument("--out", required=True)
    ap.add_argument("--fps", type=float, default=2.5, help="frames per second to sample")
    ap.add_argument("--max-frames", type=int, default=0, help="0 = no cap")
    args = ap.parse_args()

    out = Path(args.out)
    (out / "frames").mkdir(parents=True, exist_ok=True)
    cap = cv2.VideoCapture(args.video)
    if not cap.isOpened():
        raise SystemExit(f"cannot open {args.video}")
    src_fps = cap.get(cv2.CAP_PROP_FPS) or 30.0
    step = max(1, int(round(src_fps / args.fps)))
    stem = Path(args.video).stem.replace(" ", "_")
    manifest = (out / "manifest.jsonl").open("a")

    fi = kept = 0
    while True:
        ok, frame = cap.read()
        if not ok:
            break
        if fi % step == 0:
            name = f"{stem}_{kept:05d}.jpg"
            dst = out / "frames" / name
            cv2.imwrite(str(dst), frame)
            manifest.write(json.dumps({
                "common": "weed", "sci": "field footage",
                "obs_id": stem, "photo_id": kept,
                "src_frame": fi, "file": str(dst),
            }) + "\n")
            kept += 1
            if args.max_frames and kept >= args.max_frames:
                break
        fi += 1
    cap.release()
    manifest.close()
    print(f"src_fps={src_fps:.1f} step={step} → {kept} frames from {fi} "
          f"(sampled ~{args.fps} fps) → {out/'frames'}")


if __name__ == "__main__":
    main()