bot suggestions: improve tests, empty list check

Author: saathviksheerla

src/routers/openml/tasks.py

@@ -263,11 +263,17 @@ async def list_tasks(  # noqa: PLR0913, PLR0912, C901, PLR0915
         clauses.append("AND d.`name` = :data_name")
         parameters["data_name"] = data_name
 
-    if task_id:
+    if task_id is not None:
+        if not task_id:
+            msg = "No tasks match the search criteria."
+            raise NoResultsError(msg)
         clauses.append("AND t.`task_id` IN :task_ids")
         parameters["task_ids"] = task_id
 
-    if data_id:
+    if data_id is not None:
+        if not data_id:
+            msg = "No tasks match the search criteria."
+            raise NoResultsError(msg)
         clauses.append("AND d.`did` IN :data_ids")
         parameters["data_ids"] = data_id
 
@@ -318,9 +324,9 @@ async def list_tasks(  # noqa: PLR0913, PLR0912, C901, PLR0915
         """,  # noqa: S608
     )
 
-    if task_id:
+    if task_id is not None:
         main_query = main_query.bindparams(bindparam("task_ids", expanding=True))
-    if data_id:
+    if data_id is not None:
         main_query = main_query.bindparams(bindparam("data_ids", expanding=True))
 
     result = await expdb.execute(main_query, parameters=parameters)

tests/routers/openml/migration/tasks_migration_test.py

@@ -112,9 +112,26 @@ def _normalize_py_task(task: dict[str, Any]) -> dict[str, Any]:
     ({"type": 1}, {"task_type_id": 1}),  # by task type
     ({"tag": "OpenML100"}, {"tag": "OpenML100"}),  # by tag
     ({"type": 1, "tag": "OpenML100"}, {"task_type_id": 1, "tag": "OpenML100"}),  # combined
+    ({"data_name": "iris"}, {"data_name": "iris"}),  # by dataset name
+    ({"data_id": 61}, {"data_id": [61]}),  # by dataset id
+    ({"data_tag": "study_14"}, {"data_tag": "study_14"}),  # by dataset tag
+    ({"number_instances": "150"}, {"number_instances": "150"}),  # quality filter
+    (
+        {"data_id": 61, "number_instances": "150"},
+        {"data_id": [61], "number_instances": "150"},
+    ),
 ]
 
-_FILTER_IDS = ["type", "tag", "type_and_tag"]
+_FILTER_IDS = [
+    "type",
+    "tag",
+    "type_and_tag",
+    "data_name",
+    "data_id",
+    "data_tag",
+    "number_instances",
+    "data_and_quality",
+]
 
 
 @pytest.mark.parametrize(
@@ -156,7 +173,10 @@ async def test_list_tasks_equal(
     assert php_response.status_code == HTTPStatus.OK
     assert py_response.status_code == HTTPStatus.OK
 
-    php_tasks: list[dict[str, Any]] = php_response.json()["tasks"]["task"]
+    php_tasks_raw = php_response.json()["tasks"]["task"]
+    php_tasks: list[dict[str, Any]] = (
+        php_tasks_raw if isinstance(php_tasks_raw, list) else [php_tasks_raw]
+    )
     py_tasks: list[dict[str, Any]] = [_normalize_py_task(t) for t in py_response.json()]
 
     php_ids = {int(t["task_id"]) for t in php_tasks}

tests/routers/openml/task_list_test.py

@@ -114,7 +114,10 @@ async def test_list_tasks_pagination_order_stable(py_api: httpx.AsyncClient) ->
     r2 = await py_api.post("/tasks/list", json={"pagination": {"limit": 5, "offset": 5}})
     ids1 = [t["task_id"] for t in r1.json()]
     ids2 = [t["task_id"] for t in r2.json()]
-    assert max(ids1) < min(ids2)
+    assert ids1 == sorted(ids1)
+    assert ids2 == sorted(ids2)
+    if ids1 and ids2:
+        assert max(ids1) < min(ids2)
 
 
 async def test_list_tasks_number_instances_range(py_api: httpx.AsyncClient) -> None: