#!/usr/bin/env python3
# file: scripts/audit_rust_test_layout.py
# version: 1

"""Audit the Knowledge Rust unit/integration test layout."""

from __future__ import annotations

import argparse
import pathlib
import re
import sys

_TEST_ATTRIBUTE = re.compile(r"(?m)^\s*#\s*\[\s*test\s*\]")
_PATH_ATTRIBUTE = re.compile(r'#\s*\[\s*path\s*=\s*"([^"]+)"\s*\]')
_FORBIDDEN_MODULE = re.compile(r"(?m)^\s*mod\s+unit_tests\s*;")


def audit_crate(crate_root: pathlib.Path) -> list[str]:
    """Return layout violations for one crate."""
    errors: list[str] = []
    src_root = crate_root / "src"
    referenced_unit_tests: set[pathlib.Path] = set()

    if src_root.is_dir():
        for source in sorted(src_root.rglob("*.rs")):
            text = source.read_text(encoding="utf-8")
            relative = source.relative_to(crate_root)

            if _TEST_ATTRIBUTE.search(text):
                errors.append(f"{crate_root.name}/{relative}: #[test] must live under unit_tests/")

            if source.name in {"tests.rs", "unit_tests.rs"}:
                errors.append(f"{crate_root.name}/{relative}: test implementation file must not live under src/")

            if _FORBIDDEN_MODULE.search(text):
                errors.append(f"{crate_root.name}/{relative}: use #[path = \"../unit_tests/...\"] instead of mod unit_tests;")

            for match in _PATH_ATTRIBUTE.finditer(text):
                candidate = (source.parent / match.group(1)).resolve()
                try:
                    candidate.relative_to((crate_root / "unit_tests").resolve())
                except ValueError:
                    continue
                referenced_unit_tests.add(candidate)

    unit_root = crate_root / "unit_tests"
    if unit_root.is_dir():
        for test_file in sorted(unit_root.rglob("*.rs")):
            resolved = test_file.resolve()
            if resolved not in referenced_unit_tests:
                relative = test_file.relative_to(crate_root)
                errors.append(f"{crate_root.name}/{relative}: unit test file is not attached by any #[path] module")

    return errors


def main() -> int:
    """Run the audit."""
    parser = argparse.ArgumentParser()
    parser.add_argument("root", nargs="?", default="crates")
    arguments = parser.parse_args()

    root = pathlib.Path(arguments.root)
    errors: list[str] = []
    crate_count = 0

    for manifest in sorted(root.glob("*/Cargo.toml")):
        crate_count += 1
        errors.extend(audit_crate(manifest.parent))

    if errors:
        for error in errors:
            print(error, file=sys.stderr)
        return 1

    print(f"Rust test-layout audit: clean ({crate_count} crate(s))")
    return 0


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