rename ner: fields to :ner-kw

backend/addcorpus/models.py

@@ -279,7 +279,7 @@ def has_named_entities(self):
         try:
             # we check if any fields exist for filtering named entities
             ner_exists = client.search(
-                index=self.es_index, query={"exists": {"field": "ner:*"}}, size=0
+                index=self.es_index, query={"exists": {"field": "*:ner-kw"}}, size=0
             )
             if total_hits(ner_exists):
                 return True

backend/addcorpus/tests/test_validators.py

@@ -31,8 +31,8 @@ def test_validate_ner_slug():
         validate_ner_slug(keyword_mapping(), "slug:ner")
     validate_ner_slug(annotated_text_mapping(), "slug:ner")
     with pytest.raises(ValidationError):
-        validate_ner_slug(date_mapping(), "ner:slug")
-    validate_ner_slug(keyword_mapping(), "ner:slug")
+        validate_ner_slug(date_mapping(), "slug:ner-kw")
+    validate_ner_slug(keyword_mapping(), "slug:ner-kw")
 
 
 def test_validate_es_mapping():

backend/addcorpus/validation/creation.py

@@ -146,19 +146,19 @@ def validate_field_name_permissible_characters(slug: str):
 def validate_ner_slug(es_mapping: dict, name: str):
     """
     Checks if colons are in field name, will raise ValidationError if the field does not meet the following requirements:
-    - starts with `ner:` prefix and is a keyword field
     - ends with `:ner` suffix and is an annotated_text field
+    - ends with `:ner-kw` suffix and is a keyword field
     """
     if ":" in name:
         if name.endswith(":ner"):
             if primary_mapping_type(es_mapping) != MappingType.ANNOTATED_TEXT.value:
                 raise ValidationError(
                     f"{name} cannot be used as a field name: the suffix `:ner` is reserved for annotated_text fields"
         )
-        elif name.startswith("ner:"):
+        elif name.endswith(":ner-kw"):
             if primary_mapping_type(es_mapping) != MappingType.KEYWORD.value:
                 raise ValidationError(
-                    f"{name} cannot be used as a field name: the prefix `ner:` is reserved for Named Entity keyword fields"
+                    f"{name} cannot be used as a field name: the suffix `:ner-kw` is reserved for Named Entity keyword fields"
                 )
         else:
             raise ValidationError(

backend/corpora/parliament/conftest.py

@@ -284,10 +284,10 @@ def parliament_corpora_settings(settings):
                 "page": None,
                 "url": None,
                 "sequence": 1,
-                "ner:location": [],
-                "ner:miscellaneous": [],
-                "ner:organization": ["Economische Zaken"],
-                "ner:person": [],
+                "location:ner-kw": [],
+                "miscellaneous:ner-kw": [],
+                "organization:ner-kw": ["Economische Zaken"],
+                "person:ner-kw": [],
             }
         ],
         "n_documents": 2,

backend/corpora/parliament/utils/parlamint.py

@@ -144,7 +144,7 @@ def get_entity_list(extracted_data: tuple[str, dict], entity: str) -> list[str]:
 
 def ner_keyword_field(entity: str):
     return FieldDefinition(
-        name=f"ner:{entity}",
+        name=f"{entity}:ner-kw",
         display_name=f"Named Entity: {entity.capitalize()}",
         searchable=True,
         es_mapping=keyword_mapping(enable_full_text_search=True),

documentation/Named-entities.md

@@ -15,10 +15,10 @@ If the main content field is called `speech`, the field containing named entity
 ```
 
 Moreover, an enriched corpus should contain the following keyword fields:
-- `ner:person`
-- `ner:location`
-- `ner:organization`
-- `ner:miscellaneous`
+- `person:ner-kw`
+- `location:ner-kw`
+- `organization:ner-kw`
+- `miscellaneous:ner-kw`
 These can be used to search or filter (to be implemented).
 
 ## Enriching a corpus with named entities