#!/usr/bin/env python3
"""Detect corrupted ICU placeholders in ARB localization files.

Guards against the class of bug where a translation string value gets a
literal '?' glued next to an ICU placeholder brace (e.g. "{hours}??
{minutes}?"), silently destroying the translated text while staying valid
JSON/ICU. This does not require Flutter tooling, so it can run early and
fast in CI.
"""

from __future__ import annotations

import glob
import json
import re
import sys
from pathlib import Path

# A literal '?' sitting immediately adjacent (0-1 whitespace chars) to a
# placeholder brace. Narrow on purpose: a normal sentence ending in a real
# '?' does not sit glued to a '{' or '}', so it won't match.
CORRUPTION_PATTERN = re.compile(r"\}\s?\?|\?\s?\{")

ARB_GLOB = "lib/l10n/app_*.arb"


def find_repo_root(start: Path) -> Path:
    """Walk up from `start` until a directory containing lib/l10n is found."""
    current = start.resolve()
    for candidate in [current, *current.parents]:
        if (candidate / "lib" / "l10n").is_dir():
            return candidate
    return current


def check_file(path: Path) -> list[str]:
    """Return a list of human-readable findings for a single ARB file."""
    findings: list[str] = []

    try:
        raw = path.read_text(encoding="utf-8")
    except UnicodeDecodeError as exc:
        return [f"{path}: FAILED TO DECODE AS UTF-8 -> {exc}"]

    try:
        data = json.loads(raw)
    except json.JSONDecodeError as exc:
        return [f"{path}: FAILED TO PARSE JSON -> {exc}"]

    for key, value in data.items():
        if key.startswith("@"):
            continue

        meta = data.get(f"@{key}")
        if not isinstance(meta, dict) or "placeholders" not in meta:
            continue

        if not isinstance(value, str):
            continue

        if CORRUPTION_PATTERN.search(value):
            findings.append(f'{path}:{key} -> "{value}"')

    return findings


def main() -> int:
    repo_root = find_repo_root(Path(__file__).parent)
    arb_files = sorted(glob.glob(str(repo_root / ARB_GLOB)))

    if not arb_files:
        print(f"No ARB files found matching {ARB_GLOB} under {repo_root}")
        return 1

    all_findings: list[str] = []
    for arb_path in arb_files:
        all_findings.extend(check_file(Path(arb_path)))

    if all_findings:
        print("Placeholder corruption detected in ARB files:")
        for finding in all_findings:
            print(f"  {finding}")
        print(f"\n{len(all_findings)} finding(s) across {len(arb_files)} file(s).")
        return 1

    print(f"OK: checked {len(arb_files)} ARB file(s), no placeholder corruption found.")
    return 0


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