Change lower_case_sort to normalizer, adjust snowflakes search

4dn-dcic · Oct 16, 2019 · 83aad22 · 83aad22
1 parent 8756395
commit 83aad22
Show file tree

Hide file tree

Showing 2 changed files with 26 additions and 56 deletions.
diff --git a/src/snovault/elasticsearch/create_mapping.py b/src/snovault/elasticsearch/create_mapping.py
@@ -113,14 +113,9 @@ def schema_mapping(field, schema, top_level=False):
                     'ignore_above': KW_IGNORE_ABOVE
                 },
                 'lower_case_sort': {
-                    'type': 'text',
-                    'analyzer': 'case_insensistive_sort',
-                    'fields': {
-                        'keyword': {
-                            'type': 'keyword',
-                            'ignore_above': KW_IGNORE_ABOVE
-                        }
-                    }
+                    'type': 'keyword',
+                    'normalizer': 'case_insensistive_sort',
+                    'ignore_above': KW_IGNORE_ABOVE
                 }
             }
         }
@@ -138,14 +133,9 @@ def schema_mapping(field, schema, top_level=False):
                     'ignore_above': KW_IGNORE_ABOVE
                 },
                 'lower_case_sort': {
-                    'type': 'text',
-                    'analyzer': 'case_insensistive_sort',
-                    'fields': {
-                        'keyword': {
-                            'type': 'keyword',
-                            'ignore_above': KW_IGNORE_ABOVE
-                        }
-                    }
+                    'type': 'keyword',
+                    'normalizer': 'case_insensistive_sort',
+                    'ignore_above': KW_IGNORE_ABOVE
                 }
             }
         }
@@ -159,14 +149,9 @@ def schema_mapping(field, schema, top_level=False):
                     'ignore_above': KW_IGNORE_ABOVE
                 },
                 'lower_case_sort': {
-                    'type': 'text',
-                    'analyzer': 'case_insensistive_sort',
-                    'fields': {
-                        'keyword': {
-                            'type': 'keyword',
-                            'ignore_above': KW_IGNORE_ABOVE
-                        }
-                    }
+                    'type': 'keyword',
+                    'normalizer': 'case_insensistive_sort',
+                    'ignore_above': KW_IGNORE_ABOVE
                 }
             }
         }
@@ -184,14 +169,9 @@ def schema_mapping(field, schema, top_level=False):
                     'ignore_above': KW_IGNORE_ABOVE
                 },
                 'lower_case_sort': {
-                    'type': 'text',
-                    'analyzer': 'case_insensistive_sort',
-                    'fields': {
-                        'keyword': {
-                            'type': 'keyword',
-                            'ignore_above': KW_IGNORE_ABOVE
-                        }
-                    }
+                    'type': 'keyword',
+                    'normalizer': 'case_insensistive_sort',
+                    'ignore_above': KW_IGNORE_ABOVE
                 }
             }
         }
@@ -210,14 +190,9 @@ def schema_mapping(field, schema, top_level=False):
                     'ignore_above': KW_IGNORE_ABOVE
                 },
                 'lower_case_sort': {
-                    'type': 'text',
-                    'analyzer': 'case_insensistive_sort',
-                    'fields': {
-                        'keyword': {
-                            'type': 'keyword',
-                            'ignore_above': KW_IGNORE_ABOVE
-                        }
-                    }
+                    'type': 'keyword',
+                    'normalizer': 'case_insensistive_sort',
+                    'ignore_above': KW_IGNORE_ABOVE
                 }
             }
         }
@@ -231,14 +206,9 @@ def schema_mapping(field, schema, top_level=False):
                     'ignore_above': KW_IGNORE_ABOVE
                 },
                 'lower_case_sort': {
-                    'type': 'text',
-                    'analyzer': 'case_insensistive_sort',
-                    'fields': {
-                        'keyword': {
-                            'type': 'keyword',
-                            'ignore_above': KW_IGNORE_ABOVE
-                        }
-                    }
+                    'type': 'keyword',
+                    'normalizer': 'case_insensistive_sort',
+                    'ignore_above': KW_IGNORE_ABOVE
                 }
             }
         }
@@ -296,12 +266,6 @@ def index_settings():
                             'asciifolding'
                         ]
                     },
-                    'case_insensistive_sort': {
-                        'tokenizer': 'keyword',
-                        'filter': [
-                            'lowercase',
-                        ]
-                    },
                     'snovault_path_analyzer': {
                         'type': 'custom',
                         'tokenizer': 'snovault_path_tokenizer',
@@ -313,6 +277,12 @@ def index_settings():
                         'type': 'path_hierarchy',
                         'reverse': True
                     }
+                },
+                'normalizer': {
+                    'case_insensistive_sort': {
+                        'type': 'custom',
+                        'filter': ['lowercase']
+                    }
                 }
             }
         }

diff --git a/src/snowflakes/search.py b/src/snowflakes/search.py
@@ -552,7 +552,7 @@ def add_to_sort_dict(requested_sort):
             }
         else:
             # fallback case, applies to all string type:string fields
-            sort['embedded.' + name + '.lower_case_sort.keyword'] = result_sort[name] = {
+            sort['embedded.' + name + '.lower_case_sort'] = result_sort[name] = {
                 'order': order,
                 'unmapped_type': 'keyword',
                 'missing': '_last'
@@ -574,7 +574,7 @@ def add_to_sort_dict(requested_sort):
                 for k, v in type_schema['sort_by'].items():
                     # Should always sort on raw field rather than analyzed field
                     # OR search on lower_case_sort for case insensitive results
-                    sort['embedded.' + k + '.lower_case_sort.keyword'] = result_sort[k] = v
+                    sort['embedded.' + k + '.lower_case_sort'] = result_sort[k] = v
         # Default is most recent first, then alphabetical by label
         if not sort:
             sort['embedded.date_created.raw'] = result_sort['date_created'] = {