tests(annotations): guard on the 'dataset' extra so base fast-test tier skips cleanly

Fast Pytest Tests failed at COLLECTION in the base '--extra test' tier
with 'ModuleNotFoundError: No module named datasets': tests/annotations/
conftest.py imported the fixture dataset builder (-> lerobot.datasets ->
the HF 'datasets' lib + pandas/pyarrow), which only ship under the
'dataset' extra, so the whole annotations package crashed.

Fix uses the repo's proven module-level guard pattern (see
tests/datasets/test_language.py), NOT a conftest-level importorskip —
verified empirically that pytest.importorskip raised during conftest
*import* is treated as a collection ERROR (exit 1), while module-level
importorskip is a clean SKIP.

  * conftest.py: import build_annotation_dataset LAZILY inside the
    fixtures so the conftest itself imports cleanly in every tier.
  * test_modules / test_validator / test_writer / test_pipeline_recipe_
    render: add module-level pytest.importorskip('datasets') +
    ('pandas') before the pyarrow / lerobot.* imports (# noqa: E402 to
    match the existing convention). pyarrow-importing modules place the
    guard before the pyarrow import.
  * tests/scripts/test_lerobot_annotate.py: same guard (its _push_to_hub
    path imports lerobot.datasets).

Result:
  - base / hardware / viz tiers (no dataset extra): annotation tests
    skip cleanly; the rest of the suite runs -> exit 0.
  - dataset tier: datasets present -> guards pass through -> annotation
    tests run with the stub VLM. The pipeline modules import only
    stdlib + relative + lerobot.datasets (no module-level datatrove /
    vllm / openai), so they import fine there.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

tests/annotations/conftest.py

@@ -26,12 +26,20 @@ from pathlib import Path
 
 import pytest
 
-from tests.fixtures.dataset_factories import build_annotation_dataset
+# NOTE: ``build_annotation_dataset`` pulls in ``lerobot.datasets`` (-> the HF
+# ``datasets`` library + ``pandas``), which only ship under the ``dataset``
+# extra. It is imported LAZILY inside the fixtures below so this conftest
+# imports cleanly in dependency tiers without that extra (e.g. the base
+# ``--extra test`` fast-test tier). The annotation test modules guard
+# themselves with a module-level ``pytest.importorskip("datasets")`` so
+# their collection is skipped — never erroring — when the extra is absent.
 
 
 @pytest.fixture
 def fixture_dataset_root(tmp_path: Path) -> Path:
     """A tiny dataset with two episodes, 12 frames each at 10 fps."""
+    from tests.fixtures.dataset_factories import build_annotation_dataset
+
     return build_annotation_dataset(
         tmp_path / "ds",
         episode_specs=[
@@ -44,6 +52,8 @@ def fixture_dataset_root(tmp_path: Path) -> Path:
 
 @pytest.fixture
 def single_episode_root(tmp_path: Path) -> Path:
+    from tests.fixtures.dataset_factories import build_annotation_dataset
+
     return build_annotation_dataset(
         tmp_path / "ds_one",
         episode_specs=[(0, 30, "Pour water from the bottle into the cup.")],

tests/annotations/test_modules.py

@@ -22,21 +22,29 @@ from dataclasses import dataclass, field
 from pathlib import Path
 from typing import Any
 
-from lerobot.annotations.steerable_pipeline.config import (
+import pytest
+
+# ``lerobot.annotations`` imports pull in ``lerobot.datasets`` (-> the HF
+# ``datasets`` library), which only ships under the ``dataset`` extra. Skip
+# this module in tiers without it instead of erroring at import.
+pytest.importorskip("datasets", reason="datasets is required (install lerobot[dataset])")
+pytest.importorskip("pandas", reason="pandas is required (install lerobot[dataset])")
+
+from lerobot.annotations.steerable_pipeline.config import (  # noqa: E402
     InterjectionsConfig,
     PlanConfig,
     VqaConfig,
 )
-from lerobot.annotations.steerable_pipeline.modules import (
+from lerobot.annotations.steerable_pipeline.modules import (  # noqa: E402
     GeneralVqaModule,
     InterjectionsAndSpeechModule,
     PlanSubtasksMemoryModule,
 )
-from lerobot.annotations.steerable_pipeline.reader import iter_episodes
-from lerobot.annotations.steerable_pipeline.staging import EpisodeStaging
-from lerobot.annotations.steerable_pipeline.vlm_client import StubVlmClient
+from lerobot.annotations.steerable_pipeline.reader import iter_episodes  # noqa: E402
+from lerobot.annotations.steerable_pipeline.staging import EpisodeStaging  # noqa: E402
+from lerobot.annotations.steerable_pipeline.vlm_client import StubVlmClient  # noqa: E402
 
-from ._helpers import make_canned_responder
+from ._helpers import make_canned_responder  # noqa: E402
 
 
 @dataclass

tests/annotations/test_pipeline_recipe_render.py

@@ -19,26 +19,34 @@ from __future__ import annotations
 
 from pathlib import Path
 
-import pyarrow.parquet as pq
+import pytest
 
-from lerobot.annotations.steerable_pipeline.config import (
+# ``pyarrow`` and the ``lerobot.datasets`` chain (-> the HF ``datasets``
+# library) only ship under the ``dataset`` extra. Skip this module in
+# tiers without it instead of erroring at import.
+pytest.importorskip("datasets", reason="datasets is required (install lerobot[dataset])")
+pytest.importorskip("pandas", reason="pandas is required (install lerobot[dataset])")
+
+import pyarrow.parquet as pq  # noqa: E402
+
+from lerobot.annotations.steerable_pipeline.config import (  # noqa: E402
     AnnotationPipelineConfig,
     InterjectionsConfig,
     PlanConfig,
     VqaConfig,
 )
-from lerobot.annotations.steerable_pipeline.executor import Executor
-from lerobot.annotations.steerable_pipeline.modules import (
+from lerobot.annotations.steerable_pipeline.executor import Executor  # noqa: E402
+from lerobot.annotations.steerable_pipeline.modules import (  # noqa: E402
     GeneralVqaModule,
     InterjectionsAndSpeechModule,
     PlanSubtasksMemoryModule,
 )
-from lerobot.annotations.steerable_pipeline.validator import StagingValidator
-from lerobot.annotations.steerable_pipeline.writer import LanguageColumnsWriter
-from lerobot.configs.recipe import MessageTurn, TrainingRecipe
-from lerobot.datasets.language_render import render_sample
+from lerobot.annotations.steerable_pipeline.validator import StagingValidator  # noqa: E402
+from lerobot.annotations.steerable_pipeline.writer import LanguageColumnsWriter  # noqa: E402
+from lerobot.configs.recipe import MessageTurn, TrainingRecipe  # noqa: E402
+from lerobot.datasets.language_render import render_sample  # noqa: E402
 
-from ._helpers import make_canned_responder
+from ._helpers import make_canned_responder  # noqa: E402
 
 
 def _build_pr1_style_blend_recipe() -> TrainingRecipe:

tests/annotations/test_validator.py

@@ -20,10 +20,18 @@ from __future__ import annotations
 import json
 from pathlib import Path
 
-from lerobot.annotations.steerable_pipeline.reader import iter_episodes
-from lerobot.annotations.steerable_pipeline.staging import EpisodeStaging
-from lerobot.annotations.steerable_pipeline.validator import StagingValidator
-from lerobot.annotations.steerable_pipeline.writer import speech_atom
+import pytest
+
+# ``lerobot.annotations`` imports pull in ``lerobot.datasets`` (-> the HF
+# ``datasets`` library), which only ships under the ``dataset`` extra. Skip
+# this module in tiers without it instead of erroring at import.
+pytest.importorskip("datasets", reason="datasets is required (install lerobot[dataset])")
+pytest.importorskip("pandas", reason="pandas is required (install lerobot[dataset])")
+
+from lerobot.annotations.steerable_pipeline.reader import iter_episodes  # noqa: E402
+from lerobot.annotations.steerable_pipeline.staging import EpisodeStaging  # noqa: E402
+from lerobot.annotations.steerable_pipeline.validator import StagingValidator  # noqa: E402
+from lerobot.annotations.steerable_pipeline.writer import speech_atom  # noqa: E402
 
 
 def _validate(root: Path, staging_dir: Path):

tests/annotations/test_writer.py

@@ -20,12 +20,19 @@ from __future__ import annotations
 import json
 from pathlib import Path
 
-import pyarrow.parquet as pq
 import pytest
 
-from lerobot.annotations.steerable_pipeline.reader import iter_episodes
-from lerobot.annotations.steerable_pipeline.staging import EpisodeStaging
-from lerobot.annotations.steerable_pipeline.writer import (
+# ``pyarrow`` and the ``lerobot.annotations`` -> ``lerobot.datasets`` chain
+# (-> the HF ``datasets`` library) only ship under the ``dataset`` extra.
+# Skip this module in tiers without it instead of erroring at import.
+pytest.importorskip("datasets", reason="datasets is required (install lerobot[dataset])")
+pytest.importorskip("pandas", reason="pandas is required (install lerobot[dataset])")
+
+import pyarrow.parquet as pq  # noqa: E402
+
+from lerobot.annotations.steerable_pipeline.reader import iter_episodes  # noqa: E402
+from lerobot.annotations.steerable_pipeline.staging import EpisodeStaging  # noqa: E402
+from lerobot.annotations.steerable_pipeline.writer import (  # noqa: E402
     LanguageColumnsWriter,
     speech_atom,
 )

tests/scripts/test_lerobot_annotate.py

@@ -3,6 +3,13 @@
 import json
 from types import SimpleNamespace
 
+import pytest
+
+# ``lerobot.scripts.lerobot_annotate`` (and the ``_push_to_hub`` path it
+# exercises) imports ``lerobot.datasets``, which only ships under the
+# ``dataset`` extra. Skip in tiers without it instead of erroring.
+pytest.importorskip("datasets", reason="datasets is required (install lerobot[dataset])")
+
 
 def test_push_to_hub_tags_uploaded_dataset_revision(tmp_path, monkeypatch):
     from lerobot.scripts.lerobot_annotate import _push_to_hub