test(explorer): add coverage for SPARQL route (#773)

sparql.py handles direct SPARQL query execution against the live graph with no test coverage anywhere in the repo. Adds coverage for the read-only allowlist (the actual security boundary here), row/timeout limits, error handling, and RDF projection fidelity.

CHANGELOG.md

@@ -9,6 +9,14 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+### Added
+
+- **Comprehensive unit and security test suite for the `/api/sparql` Explorer route** (#773) by @Sameer6305
+  - Added `tests/explorer/test_sparql_route.py` (30 tests) covering the SPARQL Explorer route (`semantica/explorer/routes/sparql.py`), which executes arbitrary SPARQL queries against an in-memory rdflib projection of the live graph and previously had zero test coverage
+  - Verified read-only allowlist enforcement against write and mutation queries (`INSERT DATA`, `DELETE DATA`, `DELETE WHERE`, `DROP ALL`, `CLEAR ALL`, `LOAD`, `CREATE GRAPH`, `MODIFY`, comments, and multi-statement injections like `SELECT ... ; DROP ALL`), confirming rejected queries short-circuit before any graph is built or queried
+  - Verified resource-limiting behavior, confirming row capping (`_SPARQL_MAX_ROWS`) truncates results and sets `truncated: true`, query timeout (`_SPARQL_TIMEOUT_S`) returns a clean error message without crashing, and concurrency semaphore (`_SPARQL_MAX_CONCURRENT`) prevents thread starvation under load
+  - Verified RDF projection fidelity for node properties and edge relationships, and error formatting for malformed SPARQL syntax with line and column extraction
+
 ### Fixed
 
 - **`react-hooks/set-state-in-effect` cascading renders across 12 Explorer workspace files** (#769, #796) by @Sameer6305 and @KaifAhmad1

docs/reference/explorer.md

@@ -262,7 +262,7 @@ Full interactive docs at `http://localhost:8000/docs`. All endpoints accept and
 
     | Endpoint | Method | Description |
     | :-------- | :------ | :----------- |
-    | `/api/sparql` | `POST` | Execute a SPARQL SELECT or ASK query |
+    | `/api/sparql` | `POST` | Execute a read-only SPARQL query (`SELECT`, `ASK`, `CONSTRUCT`, or `DESCRIBE`) |
 
   </Accordion>
   <Accordion title="Decisions, Provenance, Annotations & Export">

tests/explorer/test_sparql_route.py

@@ -0,0 +1,266 @@
+"""Tests for the SPARQL Explorer route (semantica/explorer/routes/sparql.py).
+
+Filed as #773: this route executes arbitrary SPARQL against an in-memory
+rdflib projection of the live graph and had zero test coverage anywhere in
+the repo. Coverage here focuses on the two things that matter most for a
+query-execution surface: (1) the read-only allowlist can't be bypassed, and
+(2) the resource-limiting behavior (row cap, timeout) actually engages.
+"""
+
+import asyncio
+from unittest.mock import patch
+
+import pytest
+
+from semantica.context.context_graph import ContextGraph
+from semantica.explorer.app import create_app
+from semantica.explorer.session import GraphSession
+
+try:
+    from starlette.testclient import TestClient
+except ImportError:
+    pytest.skip(
+        "starlette TestClient is required for explorer tests. Install semantica[explorer].",
+        allow_module_level=True,
+    )
+
+import semantica.explorer.routes.sparql as sparql_mod
+
+
+def _build_sample_graph() -> ContextGraph:
+    graph = ContextGraph(advanced_analytics=False)
+    graph.add_node(
+        "python",
+        node_type="language",
+        content="Python programming language",
+        popularity="high",
+    )
+    graph.add_node("javascript", node_type="language", content="JavaScript programming language")
+    graph.add_node("web_dev", node_type="concept", content="Web Development")
+    graph.add_edge("python", "web_dev", edge_type="used_in", weight=0.5)
+    return graph
+
+
+@pytest.fixture(scope="module")
+def client():
+    session = GraphSession(_build_sample_graph())
+    app = create_app(session=session)
+    with TestClient(app) as test_client:
+        yield test_client
+
+
+def _post(client, query):
+    return client.post("/api/sparql", json={"query": query})
+
+
+# ---------------------------------------------------------------------------
+# Happy-path query types
+# ---------------------------------------------------------------------------
+
+def test_select_returns_expected_columns_and_rows(client):
+    resp = _post(client, "SELECT ?s ?label WHERE { ?s <http://www.w3.org/2000/01/rdf-schema#label> ?label }")
+    assert resp.status_code == 200
+    payload = resp.json()
+    assert payload["error"] is None
+    assert set(payload["columns"]) == {"s", "label"}
+    assert payload["total"] == len(payload["rows"])
+    labels = {row["label"] for row in payload["rows"]}
+    assert "Python programming language" in labels
+    assert "JavaScript programming language" in labels
+
+
+def test_ask_query_returns_boolean_like_result(client):
+    resp = _post(
+        client,
+        "ASK { ?s <http://www.w3.org/2000/01/rdf-schema#label> \"Python programming language\" }",
+    )
+    assert resp.status_code == 200
+    payload = resp.json()
+    assert payload["error"] is None
+
+
+def test_construct_query_succeeds(client):
+    resp = _post(
+        client,
+        "CONSTRUCT { ?s <http://www.w3.org/2000/01/rdf-schema#label> ?label } "
+        "WHERE { ?s <http://www.w3.org/2000/01/rdf-schema#label> ?label }",
+    )
+    assert resp.status_code == 200
+    payload = resp.json()
+    assert payload["error"] is None
+
+
+def test_describe_query_succeeds(client):
+    resp = _post(client, "DESCRIBE <http://semantica.local/entity/python>")
+    assert resp.status_code == 200
+    payload = resp.json()
+    assert payload["error"] is None
+
+
+def test_lowercase_query_keyword_is_accepted(client):
+    """The allowlist regex is case-insensitive; confirm lowercase keywords work too."""
+    resp = _post(client, "select ?s where { ?s a <http://semantica.local/entity/language> }")
+    assert resp.status_code == 200
+    payload = resp.json()
+    assert payload["error"] is None
+
+
+def test_select_with_no_results_returns_empty_rows(client):
+    resp = _post(client, "SELECT ?s WHERE { ?s a <http://semantica.local/entity/nonexistent> }")
+    assert resp.status_code == 200
+    payload = resp.json()
+    assert payload["error"] is None
+    assert payload["rows"] == []
+    assert payload["total"] == 0
+
+
+# ---------------------------------------------------------------------------
+# Read-only allowlist: this is the security-relevant surface (#773's core risk)
+# ---------------------------------------------------------------------------
+
+@pytest.mark.parametrize(
+    "query",
+    [
+        "INSERT DATA { <http://semantica.local/entity/x> a <http://semantica.local/entity/hacked> }",
+        "DELETE DATA { <http://semantica.local/entity/python> a <http://semantica.local/entity/language> }",
+        "DELETE WHERE { ?s ?p ?o }",
+        "DROP ALL",
+        "DROP GRAPH <http://example.org/g>",
+        "CLEAR ALL",
+        "CLEAR GRAPH <http://example.org/g>",
+        "LOAD <http://example.org/data.ttl>",
+        "CREATE GRAPH <http://example.org/g>",
+        "MODIFY <http://example.org/g> DELETE { ?s ?p ?o } WHERE { ?s ?p ?o }",
+        "",
+        "   ",
+        "not a sparql query at all",
+        # A write statement smuggled after a comment/whitespace prefix, or
+        # appended after a valid-looking read query, must still be rejected
+        # since the whole string doesn't start with an allowed keyword.
+        "# comment\nDROP ALL",
+        "SELECT ?s WHERE { ?s ?p ?o } ; DROP ALL",
+    ],
+)
+def test_write_and_non_read_queries_are_rejected(client, query):
+    resp = _post(client, query)
+    assert resp.status_code == 200, "rejection is a normal 200 response with an error field, not an HTTP error"
+    payload = resp.json()
+    assert payload["error"] is not None
+    assert payload["rows"] == []
+    assert payload["columns"] == []
+    assert payload["total"] == 0
+
+
+def test_rejected_query_never_touches_the_graph(client):
+    """A rejected query must short-circuit before any graph is built/queried."""
+    with patch.object(sparql_mod, "_build_rdflib_graph") as mock_build:
+        resp = _post(client, "DROP ALL")
+    assert resp.status_code == 200
+    assert resp.json()["error"] is not None
+    mock_build.assert_not_called()
+
+
+# ---------------------------------------------------------------------------
+# Error handling for malformed queries
+# ---------------------------------------------------------------------------
+
+def test_malformed_query_returns_error_without_crashing(client):
+    resp = _post(client, "SELECT ?s WHERE { this is not valid sparql syntax")
+    assert resp.status_code == 200
+    payload = resp.json()
+    assert payload["error"] is not None
+    assert payload["rows"] == []
+    assert payload["total"] == 0
+
+
+def test_malformed_query_error_includes_line_and_column_when_present(client):
+    resp = _post(client, "SELECT ?s WHERE { $$$ invalid $$$ }")
+    payload = resp.json()
+    assert payload["error"] is not None
+    # error_line/error_column are best-effort extraction from the pyparsing
+    # error message; they may be None depending on rdflib's error text, but
+    # the fields must always be present and of the right type when set.
+    assert payload["error_line"] is None or isinstance(payload["error_line"], int)
+    assert payload["error_column"] is None or isinstance(payload["error_column"], int)
+
+
+# ---------------------------------------------------------------------------
+# Resource limits: row cap and timeout
+# ---------------------------------------------------------------------------
+
+def test_row_cap_truncates_results(client):
+    with patch.object(sparql_mod, "_SPARQL_MAX_ROWS", 1):
+        resp = _post(
+            client,
+            "SELECT ?s WHERE { ?s a <http://semantica.local/entity/language> }",
+        )
+    assert resp.status_code == 200
+    payload = resp.json()
+    assert payload["error"] is None
+    assert len(payload["rows"]) == 1
+    assert payload["truncated"] is True
+
+
+def test_result_below_cap_is_not_marked_truncated(client):
+    resp = _post(client, "SELECT ?s WHERE { ?s a <http://semantica.local/entity/language> }")
+    payload = resp.json()
+    assert payload["truncated"] is False
+
+
+def test_query_timeout_returns_clean_error_not_a_crash(client):
+    async def _raise_timeout(coro, timeout=None):
+        coro.close()  # avoid a 'coroutine was never awaited' warning from the mock
+        raise asyncio.TimeoutError()
+
+    with patch.object(sparql_mod.asyncio, "wait_for", side_effect=_raise_timeout):
+        resp = _post(client, "SELECT ?s WHERE { ?s a <http://semantica.local/entity/language> }")
+    assert resp.status_code == 200
+    payload = resp.json()
+    assert payload["error"] is not None
+    assert "timed out" in payload["error"].lower()
+    assert payload["rows"] == []
+
+
+# ---------------------------------------------------------------------------
+# Data-mapping fidelity: does the graph->RDF projection reflect session state?
+# ---------------------------------------------------------------------------
+
+def test_node_properties_are_projected_as_literals_excluding_reserved_keys(client):
+    resp = _post(
+        client,
+        "SELECT ?p ?v WHERE { <http://semantica.local/entity/python> ?p ?v }",
+    )
+    payload = resp.json()
+    predicates = {row["p"] for row in payload["rows"]}
+    # 'popularity' is a real property and should be projected.
+    assert any("popularity" in p for p in predicates)
+    # content/valid_from/valid_until are excluded from prop: projection
+    # (content instead becomes rdfs:label, handled separately).
+    assert not any(p.endswith("/prop/content") for p in predicates)
+
+
+def test_edges_are_projected_with_their_relationship_type(client):
+    resp = _post(
+        client,
+        "SELECT ?o WHERE { <http://semantica.local/entity/python> "
+        "<http://semantica.local/prop/used_in> ?o }",
+    )
+    payload = resp.json()
+    assert payload["error"] is None
+    assert any("web_dev" in row["o"] for row in payload["rows"])
+
+
+def test_concurrent_requests_all_complete_successfully(client):
+    """Basic smoke test that the concurrency semaphore doesn't deadlock or
+    drop requests under light concurrent load."""
+    import concurrent.futures
+
+    def _run():
+        return _post(client, "SELECT ?s WHERE { ?s a <http://semantica.local/entity/language> }")
+
+    with concurrent.futures.ThreadPoolExecutor(max_workers=6) as pool:
+        results = list(pool.map(lambda _: _run(), range(6)))
+
+    for resp in results:
+        assert resp.status_code == 200
+        assert resp.json()["error"] is None