ITWorx_MCP_Hub-Public/tests/test_wp220_export.py

"""WP-220B regression tests: deterministic, clean-room release export."""

from __future__ import annotations

import json
import subprocess
import sys
import tarfile
from pathlib import Path

import pytest

from scripts import create_manifest, export_release


def _git(repo: Path, *args: str) -> str:
    result = subprocess.run(  # noqa: S603
        ["git", *args],  # noqa: S607
        cwd=repo,
        capture_output=True,
        text=True,
        check=True,
    )
    return result.stdout


def _build_repo(tmp_path: Path) -> Path:
    repo = tmp_path / "repo"
    repo.mkdir()
    _git(repo, "init", "-q", "-b", "main")
    _git(repo, "config", "user.email", "test@example.invalid")
    _git(repo, "config", "user.name", "WP-220 Test")
    _git(repo, "config", "core.autocrlf", "false")
    (repo / ".gitattributes").write_text("* text=auto eol=lf\n", encoding="utf-8")
    (repo / ".gitignore").write_text("secret.local\n", encoding="utf-8")
    (repo / "README.md").write_text("hello\n", encoding="utf-8")
    (repo / "src").mkdir()
    (repo / "src" / "app.py").write_text("print('hi')\n", encoding="utf-8")
    (repo / "pyproject.toml").write_text(
        '[project]\nname = "x"\nversion = "0.0.0"\ndependencies = []\n\n'
        "[dependency-groups]\ndev = []\n",
        encoding="utf-8",
    )
    (repo / "evidence").mkdir()
    (repo / "evidence" / "NOTE.md").write_text("evidence\n", encoding="utf-8")
    (repo / "secret.local").write_text("not tracked, must never be exported\n", encoding="utf-8")
    _git(repo, "add", "-A")
    _git(repo, "commit", "-q", "-m", "initial commit")
    manifest = create_manifest.build_manifest("HEAD", repo)
    (repo / "PACK_MANIFEST.json").write_text(
        json.dumps(manifest, indent=2) + "\n", encoding="utf-8"
    )
    _git(repo, "add", "PACK_MANIFEST.json")
    _git(repo, "commit", "-q", "-m", "record manifest")
    return repo


@pytest.fixture
def repo(tmp_path: Path) -> Path:
    return _build_repo(tmp_path)


def _run_export(repo: Path, output_dir: Path, *extra: str) -> subprocess.CompletedProcess[str]:
    script = Path(export_release.__file__)
    return subprocess.run(  # noqa: S603
        [sys.executable, str(script), "--root", str(repo), "--output-dir", str(output_dir), *extra],
        cwd=repo,
        capture_output=True,
        text=True,
        check=False,
    )


def test_source_archive_contains_only_tracked_content(repo: Path, tmp_path: Path) -> None:
    output_dir = tmp_path / "export"
    result = _run_export(repo, output_dir)
    assert result.returncode == 0, result.stdout + result.stderr

    archive_path = output_dir / "itworx-mcp-hub-source.tar.gz"
    assert archive_path.is_file()
    with tarfile.open(archive_path) as archive:
        names = set(archive.getnames())
    assert "secret.local" not in names
    assert "README.md" in names
    assert "src/app.py" in names


def test_export_verifies_clean_against_manifest(repo: Path, tmp_path: Path) -> None:
    output_dir = tmp_path / "export"
    result = _run_export(repo, output_dir, "--verify")
    assert result.returncode == 0, result.stdout + result.stderr

    export_manifest = json.loads((output_dir / "EXPORT_MANIFEST.json").read_text(encoding="utf-8"))
    assert export_manifest["sourceArchiveVerifiedAgainstManifest"] is True
    assert export_manifest["sourceArchiveVerificationErrors"] == []


def test_export_detects_tampering_via_verify(repo: Path, tmp_path: Path) -> None:
    (repo / "src" / "app.py").write_text("print('tampered')\n", encoding="utf-8")
    _git(repo, "commit", "-am", "tamper without regenerating manifest")

    output_dir = tmp_path / "export"
    result = _run_export(repo, output_dir, "--verify")
    assert result.returncode == 1
    export_manifest = json.loads((output_dir / "EXPORT_MANIFEST.json").read_text(encoding="utf-8"))
    assert export_manifest["sourceArchiveVerifiedAgainstManifest"] is False
    assert export_manifest["sourceArchiveVerificationErrors"]


def test_checksums_file_matches_produced_artifacts(repo: Path, tmp_path: Path) -> None:
    output_dir = tmp_path / "export"
    _run_export(repo, output_dir, "--include-evidence")

    checksums_text = (output_dir / "CHECKSUMS.sha256").read_text(encoding="utf-8")
    lines = [line for line in checksums_text.splitlines() if line.strip()]
    assert any(line.endswith("itworx-mcp-hub-source.tar.gz") for line in lines)
    assert any(line.endswith("sbom.cdx.json") for line in lines)

    import hashlib

    for line in lines:
        digest, name = line.split("  ", 1)
        data = (output_dir / name).read_bytes()
        assert hashlib.sha256(data).hexdigest() == digest


def test_history_bundle_reproduces_identical_tree(repo: Path, tmp_path: Path) -> None:
    output_dir = tmp_path / "export"
    _run_export(repo, output_dir, "--include-history")

    bundle_path = output_dir / "itworx-mcp-hub-history.bundle"
    assert bundle_path.is_file()

    clone_dir = tmp_path / "clone-from-bundle"
    _git(tmp_path, "clone", "-q", str(bundle_path), str(clone_dir))
    original_tree = create_manifest.resolve_tree("HEAD", repo)
    cloned_tree = create_manifest.resolve_tree("HEAD", clone_dir)
    assert original_tree == cloned_tree


def test_sbom_is_generated_and_records_source_commit(repo: Path, tmp_path: Path) -> None:
    output_dir = tmp_path / "export"
    _run_export(repo, output_dir)

    sbom = json.loads((output_dir / "sbom.cdx.json").read_text(encoding="utf-8"))
    head_commit = create_manifest.resolve_commit("HEAD", repo)
    assert sbom["metadata"]["sourceCommit"] == head_commit


def test_evidence_archive_contains_only_evidence_subtree(repo: Path, tmp_path: Path) -> None:
    output_dir = tmp_path / "export"
    _run_export(repo, output_dir, "--include-evidence")

    evidence_archive = output_dir / "itworx-mcp-hub-evidence.tar.gz"
    assert evidence_archive.is_file()
    with tarfile.open(evidence_archive) as archive:
        names = set(archive.getnames())
    assert names == {"evidence", "evidence/NOTE.md"}