from __future__ import annotations import hashlib import json import subprocess import sys from pathlib import Path import pytest ROOT = Path(__file__).resolve().parents[1] SCRIPTS = ROOT / "scripts" if str(SCRIPTS) not in sys.path: sys.path.insert(0, str(SCRIPTS)) import import_yinduzhanxing as importer # noqa: E402 POLICY = ROOT / "references/cross_project_contract/sync_policy.v2.json" def _git(root: Path, *args: str) -> str: return subprocess.check_output(["git", "-C", str(root), *args], text=True).strip() def _target(tmp_path: Path) -> Path: target = tmp_path / "commercial" target.mkdir() subprocess.run(["git", "init", "-q", str(target)], check=True) subprocess.run(["git", "-C", str(target), "config", "user.email", "tests@example.invalid"], check=True) subprocess.run(["git", "-C", str(target), "config", "user.name", "Tests"], check=True) (target / "README.md").write_text("commercial\n", encoding="utf-8") subprocess.run(["git", "-C", str(target), "add", "README.md"], check=True) subprocess.run(["git", "-C", str(target), "commit", "-qm", "base"], check=True) return target def _source(tmp_path: Path) -> Path: source = tmp_path / "snapshot" (source / "scripts").mkdir(parents=True) (source / "references").mkdir() (source / "SKILL.md").write_text("# Research skill\n", encoding="utf-8") (source / "AGENTS.md").write_text("research agent\n", encoding="utf-8") (source / "references/strict-workflow-router.md").write_text("router\n", encoding="utf-8") (source / "scripts/unified_consultation_orchestrator.py").write_text("# source\n", encoding="utf-8") (source / "scripts/report_orchestrator.py").write_text("# report\n", encoding="utf-8") (source / "scripts/jyotish_api_server.py").write_text("# api\n", encoding="utf-8") return source def _policy(tmp_path: Path, *, mirror: list[dict[str, str]] | None = None) -> Path: value = json.loads(POLICY.read_text(encoding="utf-8")) if mirror is not None: value["modes"]["mirror"] = mirror path = tmp_path / "policy.json" path.write_text(json.dumps(value), encoding="utf-8") return path def test_v2_policy_is_strictly_one_way_and_has_no_reverse_gate() -> None: policy = importer.load_policy(POLICY) assert policy["direction"] == "research_to_commercial_only" assert policy["reverse_sync"] == "forbidden" assert "commercial_to_research" not in json.dumps(policy, sort_keys=True) assert policy["modes"]["protected"] def test_dry_run_writes_no_target_file_and_snapshot_commit_is_unknown(tmp_path: Path) -> None: source, target = _source(tmp_path), _target(tmp_path) before = _git(target, "status", "--porcelain=v1") manifest = importer.build_manifest(source, policy_path=POLICY, target=target) after = _git(target, "status", "--porcelain=v1") assert before == after == "" assert manifest["source_mode"] == "snapshot" assert manifest["source_commit"] == "unknown" assert manifest["mirror_files"][0]["status"] == "new" assert not (target / "references/upstream/yinduzhanxing/SKILL.md").exists() def test_apply_handles_new_update_and_unchanged_with_recomputable_hashes(tmp_path: Path) -> None: source, target = _source(tmp_path), _target(tmp_path) destination = target / "references/upstream/yinduzhanxing/SKILL.md" first = importer.build_manifest(source, policy_path=POLICY, target=target, apply=True) assert first["mirror_files"][0]["status"] == "applied" assert destination.read_bytes() == (source / "SKILL.md").read_bytes() unchanged = importer.build_manifest(source, policy_path=POLICY, target=target) assert unchanged["mirror_files"][0]["status"] == "unchanged" assert unchanged["mirror_files"][0]["source_sha256"] == hashlib.sha256(destination.read_bytes()).hexdigest() (source / "SKILL.md").write_text("# Updated research skill\n", encoding="utf-8") update = importer.build_manifest(source, policy_path=POLICY, target=target) assert update["mirror_files"][0]["status"] == "update" assert update["mirror_files"][0]["target_sha256_before"] != update["mirror_files"][0]["target_sha256_after"] def test_semantic_files_never_overwrite_target(tmp_path: Path) -> None: source, target = _source(tmp_path), _target(tmp_path) (target / "SKILL.md").write_text("commercial skill\n", encoding="utf-8") before = (target / "SKILL.md").read_bytes() manifest = importer.build_manifest(source, policy_path=POLICY, target=target, apply=True) assert (target / "SKILL.md").read_bytes() == before row = next(item for item in manifest["semantic_merge_files"] if item["path"] == "SKILL.md") assert row["status"] == "review_required" assert manifest["operator_review_required"] is True def test_protected_mapping_is_rejected_instead_of_silently_skipped(tmp_path: Path) -> None: source, target = _source(tmp_path), _target(tmp_path) policy = _policy(tmp_path, mirror=[{"source": "SKILL.md", "target": "frontend/src/stolen.ts", "license": "MIT"}]) with pytest.raises(importer.ImportRejected, match="protected_rejected"): importer.build_manifest(source, policy_path=policy, target=target, apply=True) assert not (target / "frontend/src/stolen.ts").exists() @pytest.mark.parametrize("path", ["../escape", "/absolute/path", "safe/../../escape"]) def test_path_traversal_and_absolute_paths_are_rejected(tmp_path: Path, path: str) -> None: source, target = _source(tmp_path), _target(tmp_path) policy = _policy(tmp_path, mirror=[{"source": "SKILL.md", "target": path, "license": "MIT"}]) with pytest.raises(importer.ImportRejected, match="unsafe_path"): importer.build_manifest(source, policy_path=policy, target=target) def test_symlink_escape_is_rejected(tmp_path: Path) -> None: source, target = _source(tmp_path), _target(tmp_path) outside = tmp_path / "outside.md" outside.write_text("outside\n", encoding="utf-8") (source / "linked.md").symlink_to(outside) policy = _policy(tmp_path, mirror=[{"source": "linked.md", "target": "references/upstream/linked.md", "license": "MIT"}]) with pytest.raises(importer.ImportRejected, match="symlink_escape|non_regular_or_symlink"): importer.build_manifest(source, policy_path=policy, target=target) def test_sensitive_source_is_rejected(tmp_path: Path) -> None: source, target = _source(tmp_path), _target(tmp_path) (source / ".env").write_text("OPENAI_API_KEY=not-a-real-key\n", encoding="utf-8") policy = _policy(tmp_path, mirror=[{"source": ".env", "target": "references/upstream/env.txt", "license": "MIT"}]) with pytest.raises(importer.ImportRejected, match="sensitive_filename|sensitive_content"): importer.build_manifest(source, policy_path=policy, target=target) def test_overlapping_source_and_target_are_rejected(tmp_path: Path) -> None: target = _target(tmp_path) source = target / "source" source.mkdir() with pytest.raises(importer.ImportRejected, match="must_not_overlap"): importer.build_manifest(source, policy_path=POLICY, target=target) def test_apply_rolls_back_earlier_files_when_later_replace_fails(tmp_path: Path, monkeypatch: pytest.MonkeyPatch) -> None: source, target = _source(tmp_path), _target(tmp_path) (source / "second.md").write_text("second\n", encoding="utf-8") policy = _policy(tmp_path, mirror=[ {"source": "SKILL.md", "target": "references/upstream/one.md", "license": "MIT"}, {"source": "second.md", "target": "references/upstream/two.md", "license": "MIT"}, ]) original_replace = importer.os.replace calls = 0 def fail_second(source_path: str, target_path: Path) -> None: nonlocal calls calls += 1 if calls == 2: raise OSError("simulated replace failure") original_replace(source_path, target_path) monkeypatch.setattr(importer.os, "replace", fail_second) with pytest.raises(OSError, match="simulated"): importer.build_manifest(source, policy_path=policy, target=target, apply=True) assert not (target / "references/upstream/one.md").exists() assert not (target / "references/upstream/two.md").exists()