Milestone 2’s core deterministic scanner path

tests/test_registry_service.py

@@ -96,3 +96,51 @@ def test_capability_must_belong_to_repository(tmp_path):
         assert "ability" in str(exc)
     else:
         raise AssertionError("expected a NotFoundError")
+
+
+def test_analyze_repository_records_snapshot_and_observed_facts(tmp_path):
+    source = tmp_path / "repo"
+    source.mkdir()
+    (source / "README.md").write_text("# Example\n", encoding="utf-8")
+    (source / "requirements.txt").write_text("fastapi\n", encoding="utf-8")
+    (source / "app.py").write_text(
+        "from fastapi import FastAPI\n"
+        "app = FastAPI()\n"
+        '@app.get("/health")\n'
+        "def health():\n"
+        "    return {}\n",
+        encoding="utf-8",
+    )
+
+    service = make_service(tmp_path)
+    repository = service.register_repository(
+        name="Example",
+        url=str(source),
+        description="A local fixture repository",
+    )
+
+    summary = service.analyze_repository(repository.id)
+
+    assert summary.analysis_run.status == "completed"
+    assert summary.snapshot is not None
+    assert summary.snapshot.file_count == 3
+    assert service.get_repository(repository.id).status == "analyzed"
+    fact_names = {(fact.kind, fact.name, fact.path) for fact in summary.facts}
+    assert ("documentation", "README", "README.md") in fact_names
+    assert ("framework", "FastAPI", "requirements.txt") in fact_names
+    assert ("interface", "python route decorator", "app.py") in fact_names
+
+
+def test_analyze_repository_failure_is_recorded(tmp_path):
+    service = make_service(tmp_path)
+    repository = service.register_repository(
+        name="Missing",
+        url=str(tmp_path / "does-not-exist"),
+    )
+
+    summary = service.analyze_repository(repository.id)
+
+    assert summary.analysis_run.status == "failed"
+    assert summary.snapshot is None
+    assert "does not exist" in (summary.analysis_run.error_message or "")
+    assert service.get_repository(repository.id).status == "analysis_failed"

tests/test_repository_scanner.py

@@ -0,0 +1,36 @@
+from repo_registry.repo_scanning.scanner import DeterministicScanner
+
+
+def test_deterministic_scanner_extracts_structural_facts(tmp_path):
+    repo = tmp_path / "sample"
+    repo.mkdir()
+    (repo / "README.md").write_text("# MailRouter\n", encoding="utf-8")
+    (repo / "pyproject.toml").write_text(
+        '[project]\ndependencies = ["fastapi", "pytest"]\n',
+        encoding="utf-8",
+    )
+    (repo / "src").mkdir()
+    (repo / "src" / "routes.py").write_text(
+        "from fastapi import APIRouter\n"
+        "router = APIRouter()\n"
+        '@router.post("/classify-email")\n'
+        "def classify_email():\n"
+        "    return {}\n",
+        encoding="utf-8",
+    )
+    (repo / "tests").mkdir()
+    (repo / "tests" / "test_routes.py").write_text("def test_ok(): pass\n", encoding="utf-8")
+
+    result = DeterministicScanner().scan(repo)
+
+    facts = {(fact.kind, fact.name, fact.path) for fact in result.facts}
+    assert result.file_count == 4
+    assert ("documentation", "README", "README.md") in facts
+    assert ("manifest", "pyproject.toml", "pyproject.toml") in facts
+    assert ("test", "test_routes.py", "tests/test_routes.py") in facts
+    assert ("framework", "FastAPI", "pyproject.toml") in facts
+    assert ("framework", "pytest", "pyproject.toml") in facts
+    assert ("interface", "python route decorator", "src/routes.py") in facts
+
+    languages = {fact.name: fact.metadata["file_count"] for fact in result.facts if fact.kind == "language"}
+    assert languages == {"Python": 2}

tests/test_web_api.py

@@ -67,3 +67,43 @@ def test_api_manual_registry_loop(tmp_path):
         assert search_response.json()
     finally:
         app.dependency_overrides.clear()
+
+
+def test_api_analysis_run_loop(tmp_path):
+    source = tmp_path / "repo"
+    source.mkdir()
+    (source / "README.md").write_text("# Searchable\n", encoding="utf-8")
+    (source / "package.json").write_text(
+        '{"dependencies":{"react":"latest","vite":"latest"}}',
+        encoding="utf-8",
+    )
+
+    def override_settings():
+        return Settings(database_path=str(tmp_path / "api-analysis.sqlite3"))
+
+    app.dependency_overrides[get_settings] = override_settings
+    client = TestClient(app)
+    try:
+        repository_response = client.post(
+            "/repos",
+            json={"name": "Frontend", "url": str(source)},
+        )
+        repository_id = repository_response.json()["id"]
+
+        run_response = client.post(f"/repos/{repository_id}/analysis-runs", json={})
+        assert run_response.status_code == 201
+        run = run_response.json()
+        assert run["analysis_run"]["status"] == "completed"
+        assert run["snapshot"]["file_count"] == 2
+
+        facts_response = client.get(f"/repos/{repository_id}/observed-facts")
+        assert facts_response.status_code == 200
+        fact_names = {
+            (fact["kind"], fact["name"], fact["path"])
+            for fact in facts_response.json()
+        }
+        assert ("documentation", "README", "README.md") in fact_names
+        assert ("framework", "React", "package.json") in fact_names
+        assert ("framework", "Vite", "package.json") in fact_names
+    finally:
+        app.dependency_overrides.clear()