Spaces:
Runtime error
Runtime error
| """ | |
| tests/test_duplicates.py | |
| -------------------------- | |
| Tests for exact (SHA256) and visual (perceptual hash) duplicate detection. | |
| Run with: pytest tests/test_duplicates.py -v | |
| """ | |
| import sys | |
| from pathlib import Path | |
| import numpy as np | |
| sys.path.insert(0, str(Path(__file__).resolve().parent.parent)) | |
| from duplicate_detector import ( | |
| TOTAL_HASH_BITS, | |
| compute_dhash, | |
| compute_sha256, | |
| find_duplicates, | |
| hamming_distance, | |
| ) | |
| from image_utils import image_to_png_bytes, to_pil | |
| def _sprite(color, size=16): | |
| arr = np.zeros((size, size, 4), dtype=np.uint8) | |
| arr[:, :, 3] = 255 | |
| arr[4:size - 4, 4:size - 4] = [*color, 255] | |
| return arr | |
| def test_identical_sprites_have_matching_sha256(): | |
| a = _sprite((200, 50, 50)) | |
| b = a.copy() | |
| sha_a = compute_sha256(image_to_png_bytes(to_pil(a))) | |
| sha_b = compute_sha256(image_to_png_bytes(to_pil(b))) | |
| assert sha_a == sha_b | |
| def test_different_sprites_have_different_sha256(): | |
| a = _sprite((200, 50, 50)) | |
| b = _sprite((50, 50, 200)) | |
| sha_a = compute_sha256(image_to_png_bytes(to_pil(a))) | |
| sha_b = compute_sha256(image_to_png_bytes(to_pil(b))) | |
| assert sha_a != sha_b | |
| def test_identical_sprites_have_zero_hamming_distance(): | |
| a = _sprite((200, 50, 50)) | |
| b = a.copy() | |
| assert hamming_distance(compute_dhash(a), compute_dhash(b)) == 0 | |
| def test_same_shape_different_color_not_flagged_as_near_duplicate(): | |
| """ | |
| CRITICAL: two sprites with the identical silhouette but different | |
| colors (e.g. a green plant vs. a purple plant) must NOT be flagged as | |
| visual duplicates -- color is part of the identity of a sprite. | |
| """ | |
| a = _sprite((50, 180, 50)) | |
| b = _sprite((180, 50, 180)) | |
| dist = hamming_distance(compute_dhash(a), compute_dhash(b)) | |
| threshold = round(0.06 * TOTAL_HASH_BITS) | |
| assert dist > threshold | |
| def test_near_identical_sprite_has_small_distance(): | |
| """A single differing pixel should produce only a small hash distance.""" | |
| a = _sprite((50, 180, 50)) | |
| b = a.copy() | |
| b[5, 5] = [45, 175, 45, 255] | |
| dist = hamming_distance(compute_dhash(a), compute_dhash(b)) | |
| assert dist <= 3 | |
| def test_find_duplicates_flags_exact_match(): | |
| a = _sprite((200, 50, 50)) | |
| b = a.copy() | |
| sha_a = compute_sha256(image_to_png_bytes(to_pil(a))) | |
| sha_b = compute_sha256(image_to_png_bytes(to_pil(b))) | |
| dhash_a = compute_dhash(a) | |
| dhash_b = compute_dhash(b) | |
| existing = [{"id": "sprite_0001", "sha256": sha_a, "phash": dhash_a}] | |
| matches = find_duplicates(sha_b, dhash_b, existing) | |
| assert len(matches) == 1 | |
| assert matches[0].exact is True | |
| assert matches[0].sprite_id == "sprite_0001" | |
| assert matches[0].similarity_percent == 100.0 | |
| def test_find_duplicates_returns_empty_for_distinct_sprite(): | |
| a = _sprite((200, 50, 50)) | |
| c = _sprite((30, 200, 30)) | |
| sha_a = compute_sha256(image_to_png_bytes(to_pil(a))) | |
| dhash_a = compute_dhash(a) | |
| sha_c = compute_sha256(image_to_png_bytes(to_pil(c))) | |
| dhash_c = compute_dhash(c) | |
| existing = [{"id": "sprite_0001", "sha256": sha_a, "phash": dhash_a}] | |
| matches = find_duplicates(sha_c, dhash_c, existing) | |
| assert matches == [] | |
| def test_empty_sprite_does_not_crash_hashing(): | |
| empty = np.zeros((0, 0, 4), dtype=np.uint8) | |
| h = compute_dhash(empty) | |
| assert isinstance(h, str) | |
| assert len(h) > 0 | |
| def test_tiny_sprite_hashing_does_not_crash(): | |
| tiny = _sprite((100, 150, 200), size=4) | |
| h = compute_dhash(tiny) | |
| assert isinstance(h, str) | |
| def test_duplicate_of_accepted_sprite_flagged_across_different_sheets(): | |
| """Duplicate detection must work even when sprites came from different sheets.""" | |
| a = _sprite((90, 90, 220)) | |
| b = a.copy() | |
| sha_a = compute_sha256(image_to_png_bytes(to_pil(a))) | |
| dhash_a = compute_dhash(a) | |
| sha_b = compute_sha256(image_to_png_bytes(to_pil(b))) | |
| dhash_b = compute_dhash(b) | |
| existing = [{"id": "sprite_from_sheet_1", "sha256": sha_a, "phash": dhash_a}] | |
| matches = find_duplicates(sha_b, dhash_b, existing) | |
| assert len(matches) == 1 | |
| assert matches[0].sprite_id == "sprite_from_sheet_1" | |
| if __name__ == "__main__": | |
| import traceback | |
| tests = [(name, fn) for name, fn in list(globals().items()) if name.startswith("test_") and callable(fn)] | |
| passed, failed = 0, 0 | |
| for name, fn in tests: | |
| try: | |
| fn() | |
| print(f"PASS {name}") | |
| passed += 1 | |
| except Exception: | |
| print(f"FAIL {name}") | |
| traceback.print_exc() | |
| failed += 1 | |
| print(f"\n{passed} passed, {failed} failed") | |
| sys.exit(1 if failed else 0) | |