Make image dirs data format streamable

src/datumaro/plugins/data_formats/image_dir.py

@@ -73,6 +73,10 @@ def __init__(
             )
         self._ann_types = set()
 
+    @property
+    def is_stream(self) -> bool:
+        return True
+
 
 class ImageDirExporter(Exporter):
     DEFAULT_IMAGE_EXT = ".jpg"

tests/unit/test_image_dir_format.py

@@ -1,20 +1,24 @@
-from unittest import TestCase
-
 import numpy as np
+import pytest
 
+from datumaro.components.dataset import Dataset, StreamDataset
 from datumaro.components.dataset_base import DatasetItem
 from datumaro.components.media import Image
 from datumaro.components.project import Dataset
 from datumaro.plugins.data_formats.image_dir import ImageDirExporter
 
 from ..requirements import Requirements, mark_requirement
 
-from tests.utils.test_utils import TestDir, check_save_and_load
+from tests.utils.assets import get_test_asset_path
+from tests.utils.test_utils import TestDir, check_save_and_load, compare_datasets
+
+DUMMY_DATASET_DIR = get_test_asset_path("image_dir_dataset")
+FORMAT_NAME = "image_dir"
 
 
-class ImageDirFormatTest(TestCase):
+class ImageDirFormatTest:
     @mark_requirement(Requirements.DATUM_GENERAL_REQ)
-    def test_can_load(self):
+    def test_can_load(self, helper_tc):
         dataset = Dataset.from_iterable(
             [
                 DatasetItem(id=1, media=Image.from_numpy(data=np.ones((10, 6, 3)))),
@@ -24,16 +28,31 @@ def test_can_load(self):
 
         with TestDir() as test_dir:
             check_save_and_load(
-                self,
+                helper_tc,
                 dataset,
                 ImageDirExporter.convert,
                 test_dir,
-                importer="image_dir",
+                importer=FORMAT_NAME,
                 require_media=True,
             )
 
     @mark_requirement(Requirements.DATUM_GENERAL_REQ)
-    def test_can_save_dataset_with_cyrillic_and_spaces_in_filename(self):
+    @pytest.mark.parametrize("dataset_cls, is_stream", [(Dataset, False), (StreamDataset, True)])
+    def test_can_import(self, dataset_cls, is_stream, helper_tc):
+        expected_dataset = Dataset.from_iterable(
+            [
+                DatasetItem(id="1", media=Image.from_numpy(data=np.zeros((4, 3, 3)), ext=".JPEG")),
+                DatasetItem(id="2", media=Image.from_numpy(data=np.zeros((3, 4, 3)), ext=".bmp")),
+            ]
+        )
+
+        actual_dataset = dataset_cls.import_from(DUMMY_DATASET_DIR, FORMAT_NAME)
+
+        assert actual_dataset.is_stream == is_stream
+        compare_datasets(helper_tc, expected_dataset, actual_dataset)
+
+    @mark_requirement(Requirements.DATUM_GENERAL_REQ)
+    def test_can_save_dataset_with_cyrillic_and_spaces_in_filename(self, helper_tc):
         dataset = Dataset.from_iterable(
             [
                 DatasetItem(
@@ -44,11 +63,11 @@ def test_can_save_dataset_with_cyrillic_and_spaces_in_filename(self):
 
         with TestDir() as test_dir:
             check_save_and_load(
-                self, dataset, ImageDirExporter.convert, test_dir, importer="image_dir"
+                helper_tc, dataset, ImageDirExporter.convert, test_dir, importer=FORMAT_NAME
             )
 
     @mark_requirement(Requirements.DATUM_GENERAL_REQ)
-    def test_can_save_and_load_image_with_arbitrary_extension(self):
+    def test_can_save_and_load_image_with_arbitrary_extension(self, helper_tc):
         dataset = Dataset.from_iterable(
             [
                 DatasetItem(id="1", media=Image.from_numpy(data=np.zeros((4, 3, 3)), ext=".JPEG")),
@@ -58,10 +77,10 @@ def test_can_save_and_load_image_with_arbitrary_extension(self):
 
         with TestDir() as test_dir:
             check_save_and_load(
-                self,
+                helper_tc,
                 dataset,
                 ImageDirExporter.convert,
                 test_dir,
-                importer="image_dir",
+                importer=FORMAT_NAME,
                 require_media=True,
             )