degbu

generic_ner.py

@@ -238,7 +238,6 @@ def attach_comp_to_closest(entities):
 
 def postprocess_entities(entities):
     # Step 1: Filter entities with the same text, keeping those with the more specific label (contains a dot)
-    filtered_entities = []
     entity_map = {}
 
     # Loop over the entities and prioritize the more specific ones
@@ -248,7 +247,13 @@ def postprocess_entities(entities):
         if entity_text not in entity_map or "." in entity["entity"]:
             entity_map[entity_text] = entity
 
-    return entity_map
+    # Collect the filtered entities from the map
+    filtered_entities = list(entity_map.values())
+
+    # Step 2: Attach "comp.function" entities to the closest other entities
+    filtered_entities = attach_comp_to_closest(filtered_entities)
+
+    return filtered_entities
 
 
 class MultitaskTokenClassificationPipeline(Pipeline):