feat: check that event_id in associated detector_number

src/chexus/validators.py

@@ -284,6 +284,31 @@ def validate(self, node: Dataset | Group) -> Violation | None:
         )
 
 
+class event_id_subset_of_detector_number(Validator):
+    def __init__(self) -> None:
+        super().__init__(
+            "event_id is not subset of associated detector_numbers",
+            "The values in the event_id field in NXevent_data should be "
+            "a subset of the values in the detector_number dataset on the "
+            "associated NXdetector.",
+        )
+
+    def applies_to(self, node: Dataset | Group) -> bool:
+        return (
+            isinstance(node, Group)
+            and node.attrs.get('NX_class') == 'NXevent_data'
+            and 'event_id' in node.children
+            and 'detector_number' in node.parent.children
+        )
+
+    def validate(self, node: Dataset | Group) -> Violation | None:
+        if not np.isin(
+            node.children['event_id'].value,
+            node.parent.children['detector_number'].value,
+        ).all():
+            return Violation(node.name)
+
+
 physical_components = [
     "NXaperture",
     "NXattenuator",
@@ -374,6 +399,7 @@ def base_validators(*, has_scipp=True):
         units_invalid(),
         NXlog_has_value(),
         detector_numbers_unique_in_all_detectors(),
+        event_id_subset_of_detector_number(),
     ]
     if has_scipp:
         validators += [

tests/validators_test.py

@@ -505,3 +505,54 @@ def test_duplicate_detector_number(value):
     assert validator.validate(det2) is None
     # Second time the same detector numbers are seen, we expect a violation
     assert validator.validate(det) is not None
+
+
+def test_event_id_not_in_detector_number():
+    det = chexus.Group(name="detector1", attrs={"NX_class": "NXdetector"})
+    det.children = {
+        'data_good': chexus.Group(
+            name="data_good", attrs={"NX_class": "NXevent_data"}, parent=det
+        ),
+        'data_bad': chexus.Group(
+            name="data_bad", attrs={"NX_class": "NXevent_data"}, parent=det
+        ),
+        'detector_number': chexus.Dataset(
+            name="detector_number", value=[1, 2, 3], shape=(3,), dtype=int, parent=det
+        ),
+    }
+    det.children['data_good'].children = {
+        'event_id': chexus.Dataset(
+            name='event_id',
+            value=[1, 2, 3],
+            shape=(3,),
+            dtype=int,
+            parent=det.children['data_good'],
+        )
+    }
+    det.children['data_bad'].children = {
+        'event_id': chexus.Dataset(
+            name='event_id',
+            value=[1, 4],
+            shape=(2,),
+            dtype=int,
+            parent=det.children['data_bad'],
+        )
+    }
+    assert chexus.validators.event_id_subset_of_detector_number().applies_to(
+        det.children['data_good']
+    )
+    assert chexus.validators.event_id_subset_of_detector_number().applies_to(
+        det.children['data_bad']
+    )
+    assert (
+        chexus.validators.event_id_subset_of_detector_number().validate(
+            det.children['data_good']
+        )
+        is None
+    )
+    assert (
+        chexus.validators.event_id_subset_of_detector_number().validate(
+            det.children['data_bad']
+        )
+        is not None
+    )