#!/usr/bin/env python3
"""Enforce a ratcheted source file line-budget contract.

The check fails when:
- a file exceeds the configured line budget and is not allowlisted; or
- an allowlist entry is stale (file removed or now within budget).
"""

from __future__ import annotations

import argparse
from pathlib import Path


def _iter_source_files(roots: list[str], exts: set[str]) -> list[Path]:
    files: list[Path] = []
    for root_text in roots:
        root = Path(root_text)
        if not root.exists():
            continue
        for path in root.rglob("*"):
            if not path.is_file():
                continue
            if path.suffix not in exts:
                continue
            if "__pycache__" in path.parts or ".venv" in path.parts:
                continue
            files.append(path.resolve())
    return sorted(files)


def _load_waivers(path: Path) -> dict[str, str]:
    waivers: dict[str, str] = {}
    if not path.exists():
        return waivers
    for raw_line in path.read_text(encoding="utf-8").splitlines():
        line = raw_line.strip()
        if not line or line.startswith("#"):
            continue
        parts = line.split("\t")
        rel_path = parts[0].strip()
        reason = parts[1].strip() if len(parts) > 1 else ""
        if rel_path:
            waivers[rel_path] = reason
    return waivers


def main() -> int:
    parser = argparse.ArgumentParser()
    parser.add_argument("--roots", nargs="+", default=["ariadne", "scripts", "tests"])
    parser.add_argument("--max-lines", type=int, default=500)
    parser.add_argument("--waivers", default="scripts/loc_hygiene_waivers.tsv")
    args = parser.parse_args()

    repo_root = Path.cwd().resolve()
    waivers = _load_waivers(repo_root / args.waivers)
    source_files = _iter_source_files(args.roots, {".py", ".sh"})

    violations: dict[str, int] = {}
    for path in source_files:
        rel = path.relative_to(repo_root).as_posix()
        lines = len(path.read_text(encoding="utf-8", errors="ignore").splitlines())
        if lines > args.max_lines:
            violations[rel] = lines

    unexpected = sorted(rel for rel in violations if rel not in waivers)
    stale = sorted(rel for rel in waivers if rel not in violations)
    if not unexpected and not stale:
        print(
            f"[hygiene] source line budget check passed (limit={args.max_lines}, over_limit={len(violations)}, waivers={len(waivers)})"
        )
        return 0

    if unexpected:
        print("[hygiene] files over budget missing from waiver list:")
        for rel in unexpected:
            print(f"- {rel}: {violations[rel]} lines (limit {args.max_lines})")
    if stale:
        print("[hygiene] stale waiver entries (remove from waiver list):")
        for rel in stale:
            print(f"- {rel}")
    return 1


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