raising error when adding alias for unknown entity + unit test

spacy/kb.pyx

@@ -39,13 +39,15 @@ cdef class KnowledgeBase:
         cdef vector[float] probs
 
         for entity, prob in zip(entities, probabilities):
-            entity_hash = self.strings.add(entity)
+            entity_hash = self.strings[entity]
+            if not entity_hash in self._entry_index:
+                raise ValueError("Alias '" + alias + "' defined for unknown entity '" + entity + "'")
+
             entry_index = <int64_t>self._entry_index.get(entity_hash)
             entry_indices.push_back(int(entry_index))
             probs.push_back(float(prob))
 
         # TODO: check that alias hadn't been defined before
-        # TODO: check that entity is already in this KB (entity_index is OK)
         # TODO: check sum(probabilities) <= 1
         # TODO: check len(entities) == len(probabilities)
 

spacy/sandbox_test_sofie/testing_el.py

@@ -7,6 +7,10 @@ def create_kb():
     print("kb size", len(mykb), mykb.get_size_entities(), mykb.get_size_aliases())
 
     # adding entities
+    entity_0 = "Q0"  # douglas adams
+    mykb.add_entity(entity_id=entity_0, prob=0.5)
+    print(" adding entity", entity_0)
+
     entity_42 = "Q42"   # douglas adams
     mykb.add_entity(entity_id=entity_42, prob=0.5)
     print(" adding entity", entity_42)
@@ -18,7 +22,7 @@ def create_kb():
     print("kb size", len(mykb), mykb.get_size_entities(), mykb.get_size_aliases())
 
     # adding aliases
-    alias = "douglas"
+    alias = "douglassss"
     print(" adding alias", alias)
     mykb.add_alias(alias=alias, entities=["Q42", "Q5301561"], probabilities=[0.8, 0.2])
 

spacy/tests/pipeline/test_el.py

@@ -0,0 +1,29 @@
+import pytest
+
+from spacy.kb import KnowledgeBase
+
+
+def test_kb_valid_entities():
+    mykb = KnowledgeBase()
+
+    # adding entities
+    mykb.add_entity(entity_id="Q1", prob=0.5)
+    mykb.add_entity(entity_id="Q2", prob=0.5)
+    mykb.add_entity(entity_id="Q3", prob=0.5)
+
+    # adding aliases
+    mykb.add_alias(alias="douglassss", entities=["Q2", "Q3"], probabilities=[0.8, 0.2])
+
+
+def test_kb_invalid_entities():
+    mykb = KnowledgeBase()
+
+    # adding entities
+    mykb.add_entity(entity_id="Q1", prob=0.5)
+    mykb.add_entity(entity_id="Q2", prob=0.5)
+    mykb.add_entity(entity_id="Q3", prob=0.5)
+
+    # adding aliases - should fail because one of the given IDs is not valid
+    with pytest.raises(ValueError):
+        mykb.add_alias(alias="douglassss", entities=["Q2", "Q342"], probabilities=[0.8, 0.2])
+