handle analyzers from Document model

quintuslabs · May 4, 2020 · 10d852d · 10d852d
1 parent 1552f23
commit 10d852d
Show file tree

Hide file tree

Showing 2 changed files with 23 additions and 20 deletions.
diff --git a/shop/search/analyzers.py b/shop/search/analyzers.py
@@ -1,19 +1,20 @@
 from elasticsearch_dsl.analysis import analyzer, token_filter
 
-html_strip = analyzer('html_strip',
-    tokenizer='standard',
-    filter=['lowercase', 'stop', 'snowball'],
-    char_filter=['html_strip'],
-)
-
-german_analyzer = analyzer('german_analyzer',
-    type='custom',
-    tokenizer='standard',
-    filter=[
-        'lowercase',
-        token_filter('asciifolding', type='asciifolding', preserve_original=False),
-        token_filter('german_stop', type='stop', language='german'),
-        token_filter('german_stemmer', type='snowball', language='german'),
-    ],
-    char_filter=['html_strip'],
-)
+body_analyzers = {
+    'default': analyzer('default_analyzer',
+        tokenizer='standard',
+        filter=['lowercase', 'stop', 'snowball'],
+        char_filter=['html_strip'],
+    ),
+    'de': analyzer('german_analyzer',
+        type='custom',
+        tokenizer='standard',
+        filter=[
+            'lowercase',
+            token_filter('asciifolding', type='asciifolding', preserve_original=False),
+            token_filter('german_stop', type='stop', language='german'),
+            token_filter('german_stemmer', type='snowball', language='german'),
+        ],
+        char_filter=['html_strip'],
+    ),
+}
diff --git a/shop/search/documents.py b/shop/search/documents.py
@@ -7,6 +7,7 @@
 from django_elasticsearch_dsl import fields, Document, Index
 
 from shop.models.product import ProductModel
+from shop.search.analyzers import body_analyzers
 
 
 class _ProductDocument(Document):
@@ -71,18 +72,19 @@ class ProductDocument:
     """
     Factory for building an elasticsearch-dsl Document class. This class
     """
-    def __new__(cls, language=None, settings=None, analyzer=None):
+    def __new__(cls, language=None, settings=None):
         if language:
             index_name = 'products-{}'.format(language.lower())
             doc_name = 'ProductDocument{}'.format(language.title())
+            analyzer = body_analyzers.get(language, body_analyzers['default'])
         else:
             index_name = 'products'
             doc_name = 'ProductDocument'
+            analyzer = body_analyzers['default']
         products_index = Index(index_name)
         if settings:
             products_index.settings(**settings)
-        kwargs = {'analyzer': analyzer} if analyzer else {}
-        attrs = {'_language': language, 'body': fields.TextField(**kwargs)}
+        attrs = {'_language': language, 'body': fields.TextField(analyzer=analyzer)}
         doc_class = type(doc_name, (_ProductDocument,), attrs)
         products_index.document(doc_class)
         return doc_class