Expand YOLO training AOIs safely

backend/tests/test_sprint130_operator_yolo_tile_dataset.py

@@ -6,6 +6,7 @@ import subprocess
 import sys
 
 import numpy as np
+import pytest
 
 
 ROOT = Path(__file__).resolve().parents[2]
@@ -75,6 +76,30 @@ def test_operator_yolo_tile_dataset_export_help_does_not_require_gis_dependencie
     assert "--blank-range-threshold" in result.stdout
 
 
+def test_default_validation_split_is_explicit_and_rejects_holdout_leakage() -> None:
+    module = load_tile_exporter()
+    samples = [
+        {"sample_slug": "geel", "recommended_split": "train"},
+        {"sample_slug": "turnhout", "recommended_split": "val"},
+        {"sample_slug": "retie", "recommended_split": "val"},
+        {"sample_slug": "westerlo", "recommended_split": "val"},
+        {"sample_slug": "arendonk_heide", "recommended_split": "val"},
+    ]
+
+    assert module.DEFAULT_VALIDATION_SAMPLE_SLUGS == frozenset(
+        {"turnhout", "retie", "westerlo", "arendonk_heide"}
+    )
+    assert module.validate_validation_split(
+        samples,
+        set(module.DEFAULT_VALIDATION_SAMPLE_SLUGS),
+    ) == set(module.DEFAULT_VALIDATION_SAMPLE_SLUGS)
+
+    with pytest.raises(SystemExit, match="recommended validation holdouts"):
+        module.validate_validation_split(samples, {"turnhout"})
+    with pytest.raises(SystemExit, match="unknown samples"):
+        module.validate_validation_split(samples, {"turnhout", "missing"})
+
+
 def test_iter_tile_windows_covers_edges_without_duplicates() -> None:
     module = load_tile_exporter()