#!/usr/bin/env python3
"""Check the content-addressed Phase 3A conformance corpus."""

from __future__ import annotations

import hashlib
import json
import pathlib
import sys

ROOT = pathlib.Path(__file__).resolve().parents[1]
MANIFEST = ROOT / "fixtures/manifest.json"


def digest(paths: list[str]) -> str:
    value = hashlib.sha256()
    for relative in paths:
        path = ROOT / relative
        value.update(relative.encode("utf-8"))
        value.update(b"\0")
        value.update(path.read_bytes())
        value.update(b"\0")
    return value.hexdigest()


def main() -> int:
    manifest = json.loads(MANIFEST.read_text(encoding="utf-8"))
    paths = sorted(path.relative_to(ROOT).as_posix() for path in (ROOT / "fixtures/conformance").glob("*.tsv"))
    if paths != manifest.get("files"):
        raise ValueError("fixture manifest file list is stale")
    actual = digest(paths)
    expected = manifest.get("corpus_sha256")
    if not expected or actual != expected:
        raise ValueError(f"fixture corpus hash mismatch: {actual}")
    json_paths = sorted(
        path.relative_to(ROOT).as_posix()
        for directory in (ROOT / "fixtures/valid", ROOT / "fixtures/invalid")
        for path in directory.glob("*.json")
    )
    if json_paths != manifest.get("json_files"):
        raise ValueError("JSON fixture manifest file list is stale")
    json_actual = digest(json_paths)
    if json_actual != manifest.get("json_corpus_sha256"):
        raise ValueError(f"JSON fixture corpus hash mismatch: {json_actual}")
    print(f"Fixture corpus SHA256 {actual}; JSON SHA256 {json_actual}")
    return 0


if __name__ == "__main__":
    try:
        raise SystemExit(main())
    except (OSError, ValueError, json.JSONDecodeError) as exc:
        print(f"fixture_digest: {exc}", file=sys.stderr)
        raise SystemExit(1)