Add unit tests (#1048)

Co-authored-by: Mark Merling <[email protected]>
Co-authored-by: fatih c. akyon <[email protected]>

tests/test_cvutils.py

@@ -0,0 +1,91 @@
+import unittest
+from unittest.mock import patch
+
+import numpy as np
+from PIL import Image
+
+from sahi.utils.cv import (
+    Colors,
+    apply_color_mask,
+    exif_transpose,
+    get_bbox_from_bool_mask,
+    get_coco_segmentation_from_bool_mask,
+    read_image,
+)
+
+
+class TestCvUtils(unittest.TestCase):
+    def test_hex_to_rgb(self):
+        colors = Colors()
+        self.assertEqual(colors.hex_to_rgb("#FF3838"), (255, 56, 56))
+
+    def test_hex_to_rgb_retrieve(self):
+        colors = Colors()
+        self.assertEqual(colors(0), (255, 56, 56))
+
+    @patch("sahi.utils.cv.cv2.cvtColor")
+    @patch("sahi.utils.cv.cv2.imread")
+    def test_read_image(self, mock_imread, mock_cvtColor):
+        fake_image = "test.jpg"
+        fake_image_val = np.array([[[10, 20, 30]]], dtype=np.uint8)
+        fake_image_rbg_val = np.array([[[10, 20, 30]]], dtype=np.uint8)
+        mock_imread.return_value = fake_image_val
+        mock_cvtColor.return_value = fake_image_rbg_val
+
+        result = read_image(fake_image)
+
+        # mock_cv2.assert_called_once_with(fake_image)
+        mock_imread.assert_called_once_with(fake_image)
+        np.testing.assert_array_equal(result, fake_image_rbg_val)
+
+    def test_apply_color_mask(self):
+        image = np.array([[0, 1]], dtype=np.uint8)
+        color = (255, 0, 0)
+
+        expected_output = np.array([[[0, 0, 0], [255, 0, 0]]], dtype=np.uint8)
+
+        result = apply_color_mask(image, color)
+
+        np.testing.assert_array_equal(result, expected_output)
+
+    def test_get_coco_segmentation_from_bool_mask_simple(self):
+        mask = np.zeros((10, 10), dtype=bool)
+        result = get_coco_segmentation_from_bool_mask(mask)
+        self.assertEqual(result, [])
+
+    def test_get_coco_segmentation_from_bool_mask_polygon(self):
+        mask = np.zeros((10, 20), dtype=bool)
+        mask[1:4, 1:4] = True
+        mask[5:8, 5:8] = True
+        result = get_coco_segmentation_from_bool_mask(mask)
+        self.assertEqual(len(result), 2)
+
+    def test_get_bbox_from_bool_mask(self):
+        mask = np.array(
+            [
+                [False, False, False],
+                [False, True, True],
+                [False, True, True],
+                [False, False, False],
+            ]
+        )
+        expected_result = [1, 1, 2, 2]
+        result = get_bbox_from_bool_mask(mask)
+        self.assertEqual(result, expected_result)
+
+    def test_exif_transpose_simple(self):
+        test_image = Image.new("RGB", (100, 100), color="red")
+        transposed = exif_transpose(test_image)
+        self.assertEqual(transposed, test_image)
+
+    def test_exif_transpose_non_standard(self):
+        test_image = Image.new("RGB", (100, 100), color="red")
+        exif = test_image.getexif()
+        exif[0x0112] = 9
+        test_image.info["exif"] = exif.tobytes()
+        transposed = exif_transpose(test_image)
+        self.assertEqual(transposed, test_image)
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_fileutils.py

@@ -2,6 +2,7 @@
 # Code written by Fatih C Akyon, 2020.
 
 import unittest
+from unittest.mock import patch
 
 
 class TestFileUtils(unittest.TestCase):
@@ -22,6 +23,16 @@ def test_list_files_recursively(self):
         self.assertEqual(len(relative_filepath_list), 7)
         self.assertEqual(len(abs_filepath_list), 7)
 
+    def test_increment_path(self):
+        from sahi.utils.file import increment_path
+
+        with patch("sahi.utils.file.Path.exists", return_value=False):
+            path = increment_path("test.txt")
+            self.assertEqual(path, "test.txt")
+        with patch("sahi.utils.file.Path.exists", return_value=True):
+            path = increment_path("test.txt", exist_ok=False)
+            self.assertEqual(path, "test.txt2")
+
 
 if __name__ == "__main__":
     unittest.main()

tests/test_postprocessutils.py

@@ -0,0 +1,31 @@
+import unittest
+
+import torch
+
+from sahi.postprocess.utils import ObjectPredictionList
+
+
+class TestPostprocessUtils(unittest.TestCase):
+    def setUp(self):
+        self.test_input = [ObjectPredictionList([1, 2, 3, 4])]
+
+    def test_get_item_int(self):
+        obj = self.test_input[0]
+        self.assertEqual(obj[0].tolist(), 1)
+
+    def test_len(self):
+        obj = self.test_input[0]
+        self.assertEqual(len(obj), 4)
+
+    def test_extend(self):
+        obj = self.test_input[0]
+        obj.extend(ObjectPredictionList([torch.randn(1, 2, 3, 4)]))
+        self.assertEqual(len(obj), 5)
+
+    def test_tostring(self):
+        obj = self.test_input[0]
+        self.assertEqual(str(obj), str([1, 2, 3, 4]))
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_prediction.py

@@ -0,0 +1,18 @@
+import unittest
+
+import numpy as np
+
+from sahi.prediction import PredictionScore
+
+
+class TestPrediction(unittest.TestCase):
+    def test_prediction_score(self):
+
+        prediction_score = PredictionScore(np.array(0.6))
+        self.assertEqual(type(prediction_score.value), float)
+        self.assertEqual(prediction_score.is_greater_than_threshold(0.5), True)
+        self.assertEqual(prediction_score.is_greater_than_threshold(0.7), False)
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_torchutils.py

@@ -0,0 +1,26 @@
+import unittest
+
+import numpy as np
+import torch
+
+from sahi.utils.torch import empty_cuda_cache, to_float_tensor, torch_to_numpy
+
+
+class TestTorchUtils(unittest.TestCase):
+    def test_empty_cuda_cache(self):
+        if torch.cuda.is_available():
+            self.assertIsNone(empty_cuda_cache())
+
+    def test_to_float_tensor(self):
+
+        img = to_float_tensor(np.random.randint(0, 256, (10, 10, 3), dtype=np.uint8))
+        self.assertEqual(img.shape, (3, 10, 10))
+
+    def test_torch_to_numpy(self):
+        img_t = torch.tensor(np.random.rand(3, 10, 10))
+        img = torch_to_numpy(img_t)
+        self.assertEqual(img.shape, (10, 10, 3))
+
+
+if __name__ == "__main__":
+    unittest.main()