scripts/download_ims_data.py

"""
Download IMS weather data for station 43 (Sde Boker) and cache to Data/ims/.
Resamples 10min to 15min. Use --list-channels to discover channel IDs for RH, Rain, WS, BP.

Usage:
  python -m scripts.download_ims_data --list-channels
  python -m scripts.download_ims_data --from 2024-01-01 --to 2024-12-31
  python -m scripts.download_ims_data --years 2
"""

from __future__ import annotations

import argparse
import sys
from pathlib import Path

# Project root on path for config and src
PROJECT_ROOT = Path(__file__).resolve().parent.parent
if str(PROJECT_ROOT) not in sys.path:
    sys.path.insert(0, str(PROJECT_ROOT))

try:
    from dotenv import load_dotenv
    load_dotenv(PROJECT_ROOT / ".env")
except ImportError:
    pass


def main() -> None:
    parser = argparse.ArgumentParser(
        description="Fetch IMS station 43 data and cache to Data/ims/ (15min resolution)."
    )
    parser.add_argument(
        "--list-channels",
        action="store_true",
        help="List channel IDs and names for station 43 (to find RH, Rain, WS, BP).",
    )
    parser.add_argument(
        "--from",
        dest="from_date",
        metavar="YYYY-MM-DD",
        help="Start date (inclusive).",
    )
    parser.add_argument(
        "--to",
        dest="to_date",
        metavar="YYYY-MM-DD",
        help="End date (inclusive).",
    )
    parser.add_argument(
        "--years",
        type=float,
        default=2,
        metavar="N",
        help="If --from/--to not set, fetch last N years (default: 2).",
    )
    parser.add_argument(
        "--chunk-days",
        type=int,
        default=7,
        metavar="D",
        help="Split range into D-day chunks (default: 7 for better API success). Use 0 to disable chunking.",
    )
    args = parser.parse_args()

    from src.ims_client import IMSClient
    from config import settings

    client = IMSClient()

    if args.list_channels:
        try:
            channels = client.list_channels(settings.IMS_STATION_ID)
        except Exception as e:
            print(f"Error fetching station metadata: {e}", file=sys.stderr)
            sys.exit(1)
        if not channels:
            print("No channels returned. Check API response structure.")
            sys.exit(0)
        print(f"Station {settings.IMS_STATION_ID} channels:")
        print("-" * 60)
        for ch in sorted(channels, key=lambda x: (x.get("channelId") or 0)):
            cid = ch.get("channelId", "?")
            name = ch.get("name", "?")
            units = ch.get("units", "")
            active = ch.get("active", True)
            print(f"  {cid:>4}  {name:<12}  {units:<8}  active={active}")
        return

    # Determine date range
    from datetime import datetime, timedelta, timezone

    end = datetime.now(timezone.utc).date()
    if args.from_date and args.to_date:
        start = datetime.strptime(args.from_date, "%Y-%m-%d").date()
        end = datetime.strptime(args.to_date, "%Y-%m-%d").date()
    elif args.from_date:
        start = datetime.strptime(args.from_date, "%Y-%m-%d").date()
    else:
        start = end - timedelta(days=int(args.years * 365.25))

    from_s = start.strftime("%Y-%m-%d")
    to_s = end.strftime("%Y-%m-%d")
    chunk_days = args.chunk_days if args.chunk_days else None

    print(f"Fetching IMS station {settings.IMS_STATION_ID} from {from_s} to {to_s} (chunk_days={chunk_days})...")
    try:
        df = client.fetch_and_cache(from_s, to_s, chunk_days=chunk_days)
    except Exception as e:
        print(f"Error: {e}", file=sys.stderr)
        sys.exit(1)

    if df.empty:
        print("No data returned. Check token and date range.")
        sys.exit(1)

    out_path = settings.IMS_CACHE_DIR / "ims_merged_15min.csv"
    print(f"Saved {len(df)} rows (15min) to {out_path}")
    print(f"Columns: {list(df.columns)}")


if __name__ == "__main__":
    main()