from __future__ import annotations

import argparse
import json
import re
from pathlib import Path

import yaml


ROOT = Path(__file__).resolve().parents[1]
REGISTRY = ROOT / "spec" / "13_formula_registry.yaml"
GOLDEN = ROOT / "spec" / "formula_golden_cases_v2.yaml"
GOLDEN_FALLBACK = ROOT / "Temp" / "formula_golden_cases.yaml"


def _registry_ids() -> set[str]:
    if not REGISTRY.exists():
        return set()
    payload = yaml.safe_load(REGISTRY.read_text(encoding="utf-8"))
    text = json.dumps(payload, ensure_ascii=False)
    ids = set(re.findall(r'\b([A-Z][A-Z0-9_]+_V[0-9]+)\b', text))
    return ids


def main() -> int:
    ap = argparse.ArgumentParser()
    ap.parse_args()

    golden_path = GOLDEN if GOLDEN.exists() else GOLDEN_FALLBACK
    if not golden_path.exists():
        print("FORMULA_GOLDEN_CASES_FAIL")
        print("- missing file: spec/formula_golden_cases_v2.yaml")
        return 1
    payload = yaml.safe_load(golden_path.read_text(encoding="utf-8"))
    rows = payload.get("golden_cases_v2") if isinstance(payload, dict) else None
    if rows is None and isinstance(payload, dict):
        rows = payload.get("golden_cases")
    if not isinstance(rows, list):
        print("FORMULA_GOLDEN_CASES_FAIL")
        print("- golden_cases must be list")
        return 1

    reg = _registry_ids()
    errors: list[str] = []
    warnings: list[str] = []
    covered = 0
    for i, r in enumerate(rows):
        if not isinstance(r, dict):
            errors.append(f"golden_cases[{i}] must be object")
            continue
        fid = str(r.get("formula_id") or "")
        if not fid:
            errors.append(f"golden_cases[{i}] missing formula_id")
            continue
        if reg and fid not in reg:
            warnings.append(f"golden_cases[{i}] formula_id not in registry: {fid}")
        if not isinstance(r.get("cases"), list) or len(r["cases"]) == 0:
            errors.append(f"golden_cases[{i}] must include non-empty cases")
        else:
            covered += 1

    if errors:
        print("FORMULA_GOLDEN_CASES_FAIL")
        for e in errors:
            print(f"- {e}")
        return 1

    if warnings:
        print("FORMULA_GOLDEN_CASES_WARN")
        for w in warnings:
            print(f"- {w}")
    else:
        print("FORMULA_GOLDEN_CASES_OK")
    print(f"rows={len(rows)} covered={covered} warnings={len(warnings)} source={golden_path.name}")
    return 0


if __name__ == "__main__":
    raise SystemExit(main())