#!/usr/bin/env python3
"""Drift guard: the site's generated catalog MUST match the API source.

Run in deploy.sh before the site build. Re-parses the API catalog and compares
it to the committed site/src/lib/service-catalog.generated.ts. Exits 1 on any
difference, so a price edited in the API but not regenerated (or a hand-edited
generated file) is caught before it reaches customers.

Usage: python3 scripts/check-service-catalog-drift.py
"""
import re
import sys
from pathlib import Path

ROOT = Path(__file__).resolve().parent.parent
sys.path.insert(0, str(ROOT / "scripts"))
import importlib.util

spec = importlib.util.spec_from_file_location("gen_catalog", ROOT / "scripts/gen-service-catalog.py")
gen = importlib.util.module_from_spec(spec)
spec.loader.exec_module(gen)

GEN = ROOT / "site/src/lib/service-catalog.generated.ts"


def parse_generated(ts: str) -> dict:
    m = re.search(r"export const SERVICE_META[^=]*=\s*\{(.*)\n\};", ts, re.S)
    if not m:
        raise SystemExit("drift-check: cannot parse generated SERVICE_META")
    body = m.group(1)
    out = {}
    for em in re.finditer(r'"([a-z0-9\-]+)":\s*\{(.*?)\},', body):
        slug = em.group(1)
        inner = em.group(2)
        name_m = re.search(r'name:\s*"((?:[^"\\]|\\.)*)"', inner)
        price_m = re.search(r"price_cents:\s*(\d+)", inner)
        gov_m = re.search(r'gov_fee_label:\s*"((?:[^"\\]|\\.)*)"', inner)
        interval_m = re.search(r'billing_interval:\s*"(month|year)"', inner)
        methods_m = re.search(r"allowed_methods:\s*\[([^\]]*)\]", inner)
        entry = {"name": gen._unescape(name_m.group(1)), "price_cents": int(price_m.group(1))}
        if gov_m:
            entry["gov_fee_label"] = gen._unescape(gov_m.group(1))
        if interval_m:
            entry["billing_interval"] = interval_m.group(1)
        if methods_m:
            entry["allowed_methods"] = re.findall(r'"([a-z]+)"', methods_m.group(1))
        out[slug] = entry
    return out


def main() -> int:
    api = gen.parse_catalog(gen.SRC.read_text())
    have = parse_generated(GEN.read_text())
    problems = []
    for slug, a in api.items():
        g = have.get(slug)
        if not g:
            problems.append(f"{slug}: missing from generated file")
            continue
        if a["price_cents"] != g["price_cents"]:
            problems.append(f"{slug}: price API={a['price_cents']} generated={g['price_cents']}")
        if a["name"] != g["name"]:
            problems.append(f"{slug}: name mismatch")
        if a.get("gov_fee_label") != g.get("gov_fee_label"):
            problems.append(f"{slug}: gov_fee_label mismatch")
        if a.get("billing_interval") != g.get("billing_interval"):
            problems.append(f"{slug}: billing_interval API={a.get('billing_interval')} generated={g.get('billing_interval')}")
        if a.get("allowed_methods") != g.get("allowed_methods"):
            problems.append(f"{slug}: allowed_methods API={a.get('allowed_methods')} generated={g.get('allowed_methods')}")
    for slug in have:
        if slug not in api:
            problems.append(f"{slug}: in generated file but not in API")
    if problems:
        print("SERVICE CATALOG DRIFT DETECTED (run: python3 scripts/gen-service-catalog.py):", file=sys.stderr)
        for p in problems:
            print("  - " + p, file=sys.stderr)
        return 1
    print(f"drift-check: OK -- {len(api)} services, API and generated catalog match.")
    return 0


if __name__ == "__main__":
    sys.exit(main())