from __future__ import annotations

import argparse
import json
from pathlib import Path
import sys

REPO_ROOT = Path(__file__).resolve().parents[1]
sys.path.insert(0, str(REPO_ROOT))

from ingestion.common.cloud_lake import client
from ingestion.common.load_config import LoadConfigBase, load_config_file
from ingestion.common.sql import split_sql


def main() -> None:
    parser = argparse.ArgumentParser(description="Run idempotent warehouse setup from a load source config.")
    parser.add_argument("--source", required=True, help="Path to the load source.yaml.")
    args = parser.parse_args()

    # Loader-neutral: setup only needs the storage-agnostic base (name, DSN, setup files),
    # so it loads any load source — s3_bulk_load, gcs_bulk_load, or a future loader — the same way.
    config = load_config_file(resolve_repo_path(args.source), LoadConfigBase)
    executed = run_databend_setup(config)
    print(json.dumps({"source": config.name, "databend": {"executed": executed}}, indent=2))


def run_databend_setup(config: LoadConfigBase) -> list[str]:
    files = config.setup_files
    if not files:
        return []

    cursor = client(config).cursor()
    executed = []
    for file_name in files:
        path = resolve_repo_path(file_name)
        for statement in split_sql(path.read_text()):
            cursor.execute(statement)
        executed.append(str(path.relative_to(REPO_ROOT)))
    return executed


def resolve_repo_path(file_name: str) -> Path:
    path = Path(file_name)
    if path.is_absolute():
        return path
    return REPO_ROOT / path


if __name__ == "__main__":
    main()